Subset (5)

F1000Journal · 8.11k rows

Split (1)

data · 8.11k rows

link stringlengths 41 45	date stringlengths 9 9	paper dict	reviews listlengths 1 6	version int64 1 5	main stringlengths 38 42
https://f1000research.com/articles/7-1974/v1	24 Dec 18	{ "type": "Opinion Article", "title": "Hackathon-driven tutorial development", "authors": [ "Bruno M. Grande", "Arjun Baghela", "Anna Cavalla", "Florian Privé", "Peter Zhang", "Yisong Zhen", "Arjun Baghela", "Anna Cavalla", "Florian Privé", "Peter Zhang", "Yisong Zhen" ], "abstract": "Software is essential for data science. However, several software tools remain out of reach for many users due to a lack of documentation, thus limiting progress in the field. Tutorial development by authors and users can greatly improve a tool's accessibility and accelerate its adoption. In this article, we explore hackathons such as hackseq as a venue for authors and users to develop tutorials to address the lack of documented software. We describe four advantages of hackathon-driven tutorial development as well as three challenges that we faced. We also discuss our experience with remote participation. In short, if properly prepared, hackathons can provide a productive venue for assembling a group of passionate people, including remote participants, to develop a suite of related tutorials and address the growing need for accessible software.", "keywords": [ "software", "hackathon", "documentation", "tutorial", "vignettes", "programming", "data science" ], "content": "\n\nData science is an interdisciplinary field that relies heavily on the use of software tools. These tools require advanced domain-specific knowledge, which is often difficult to acquire and keep up-to-date considering the rate at which new methods become available and best practices evolve. This difficulty primarily stems from incomplete or unclear documentation. In the case of software packages (e.g. in R and Python), minimal documentation consists of describing inputs and outputs for individual functions. The Comprehensive R Archive Network (CRAN) is the de facto package repository for the R programming language and requires that submitted packages at least include this degree of documentation. CRAN also offers a framework for package developers to include additional documentation in the form of vignettes, which typically demonstrate real-world use cases. However, it appears that the majority of CRAN packages do not have vignettes1. On the other hand, vignettes are required for submission to Bioconductor, a package repository geared towards computational biology. This requirement has made Bioconductor and its numerous packages much more accessible to both new and experienced users. Unfortunately, the practice of including user-friendly vignettes or tutorials remains uncommon. To address this problem, the authors experimented with tutorial development in a hackathon project at hackseq 20172. We describe our experience in this article.\n\nThe aim of our project was to organize the collective knowledge of a group of computational biologists into modular tutorials that leveraged the same dataset. Tutorial topics were proposed by and then assigned to members of the team. The tutorials were designed to be independent from each other, but they can readily be combined to form workshop lessons that use the same dataset. In this paper, we explore the benefits and challenges associated with hackathon-driven tutorial development, including the trade-offs of remote hackathon participation. Briefly, we believe that hackathons are an excellent venue for tutorial development and are particularly suitable for remote participation.\n\nWe found at least four benefits of collaboratively developing tutorials in a hackathon setting. First, interest in a given topic can be assessed based on the voluntary participation of hackathon attendees. The assembly of people interested in a topic can further motivate tutorial development. Second, once a common dataset is selected and processed, team members can efficiently work in parallel. Third, although team members do not have to rely on one another, they may draw on the collective knowledge of the team. The various perspectives and ideas from different research specialities can guide tutorial design, resulting in higher-quality material. In more practical terms, developing tutorials during a hackathon allows problems to be more readily resolved, and team members can perform peer review, leading to more polished tutorials. Fourth, these hackathon projects often bring together community members that have yet to work together and can thus catalyze new collaborations.\n\nThat said, there are some challenges or considerations to keep in mind when developing tutorials in the context of hackathons. First, the hackathon project should feature a theme or topic that is focused in scope so that team members can assist one another. The skill level of the target audience should also be determined beforehand. Second, once a theme is decided, any existing tutorials should be identified beforehand to avoid repetition. As mentioned previously, vignettes exist for many software tools, and time is best invested in developing new material that is not yet available. Alternatively, one could build on the work of others by adapting or improving existing open-source tutorials. Third, we suggest that you identify a dataset that can be used in all proposed tutorials and meets the following criteria: openly accessible; properly formatted (i.e. little to no missing or malformed data); relevant to the target audience; and ideally sized (i.e. large enough to be interesting but small enough to fit on a personal computer). For example, in the case of tutorials geared towards computational biologists, there are several interesting human genomic datasets, but access is often restricted for privacy. Accordingly, it may be more practical to select a dataset first and then determine a theme and set of tutorial topics that can be developed using this dataset.\n\nIt is often cost-prohibitive to travel for short conferences, especially when travel awards do not cover non-traditional meetings such as hackathons. Fortunately, remote participation is not only possible for hackathons, but relatively straightforward. Several tools exist to support collaborative projects while eliminating the need for collocation. For instance, GitHub offers decentralized code sharing, Skype enables face-to-face team discussions, and Slack is a popular platform for asynchronous communication, which is essential when team members inhabit distant time zones. For example, we successfully managed our project despite being located in Vancouver, BC with remote participants in China and France. We argue that remote participation is especially straightforward for tutorial development because material can be developed independently and thus asynchronously. On the other hand, virtual attendance precludes any participation in social or networking events. There is also additional work involved in ensuring that every local and remote team member have assigned tasks at any given time. Overall though, we believe that allowing remote participation is a net benefit for a hackathon project, especially for tutorial development.\n\nIn conclusion, our experience with developing tutorials at a hackathon with remote participants was positive. We believe that we were able to achieve more together than separately, mostly because we gain access to immediate peer review of our tutorials. We also learned lessons that will help ensure success for future hackathon projects geared toward tutorial development. We believe this approach to be generalizable to other fields and a model for assembling passionate data scientists with similar interests and organizing their collective knowledge into modular tutorials. In turn, these tutorials can greatly benefit the field of data science by facilitating the adoption of powerful tools and accelerating the training of future data scientists.\n\n\nData availability\n\nNo data are associated with this article.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nThe authors acknowledge the hackseq 2017 steering committee for organizing the hackathon.\n\n\nReferences\n\nSilge J: Mining CRAN DESCRIPTION Files. 2017; (Accessed: 26th September 2018). Reference Source\n\nhackseq Organizing Committee 2016: hackseq: Catalyzing collaboration between biological and computational scientists via hackathon [version 2; referees: 2 approved]. F1000Res. 2017; 6: 197. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "42213", "date": "11 Jan 2019", "name": "Ming Tang", "expertise": [ "Reviewer Expertise bioinformatics", "data science" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nSoftware is the driving force for data science. However, lacking documentation hinders the adoption and usage of certain software tools. In this article, Bruno et.al described the advantages and challenges of hackathon-based tutorial development to promote software usage. Additionally, the authors shared their experience with remote participation of hackathons. Overall, the laid-out points are practical and useful to the community. I have several comments below.\n\nI just learned that documentation has four different functions: tutorials, how-to-guides, explanation and technical references 1. They are distinct from each other and good documentation needs to be structured around them and be separated from each other. Specifically, a tutorial is learning-oriented; a how-to-guide is goal-oriented; an explanation is understanding-oriented and a reference guide is information-oriented. That being said, it would be good for the authors to make it clear that a tutorial is only one part of the function of documentation.\nThe authors did not provide any details of the hackathon. It will be good to have examples. E.g. what tutorials were developed for what tools? If hackathon-driven tutorial development is successful, the authors may need to provide what they have accomplished by developing the tutorials and provide some feedback on the user-end of that tool. Are there any other challenges besides choosing topics, identifying existing tutorials and selecting a dataset? Are there any challenges in organizing such a hackathon? Are there any challenges for remote attendance? E.g. if the attendants are in a different time-zone, how the peer-review was conducted in a timely manner? The sentence: “minimal documentation consists of describing inputs and outputs for individual functions” should contain “arguments/parameters” as well. If one shortens this sentence “These tools require advanced domain-specific knowledge, which is often difficult to acquire and keep up-to-date considering the rate at which new methods become available and best practices evolve. This difficulty primarily stems from incomplete or unclear documentation.” to “Difficulties in acquiring domain-specific knowledge primary stems from incomplete or unclear documentation.” This is not true. Please re-structure the sentence.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Partly", "responses": [] }, { "id": "43146", "date": "04 Feb 2019", "name": "Brad A. Chapman", "expertise": [ "Reviewer Expertise bioinformatics", "community development" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors describe their experience with tutorial development in a collaborative event. These events bring together volunteer community participants, either locally or remotely, for a defined period of time to accomplish specific tasks. In this case, the motivation is to help develop and improve user facing documentation for existing software packages.\nThis is an important problem since documentation writing is an undervalued but important part of building useful scientific software. Some suggestions to help improve the manuscript:\nCould you provide specific examples of materials developed during your collaborative events? Seeing outcomes would help motivate readers of the paper. Do you have specific approaches for identifying existing tutorials and datasets, and connecting with other communities doing similar work? One downside of time-bounded events is that it can be hard to have enough time to become familiar with previous work and finish tutorials in the time frame of the event. How do you help make it easier to get started, and ensure that work is in a completed state at the end of the event? How do you maintain documentation following the event? Is it integrated back into the projects? Do subsequent events follow up and continue the work? Out of date and obsolete documentation can be as problematic as absent or terse documentation. Strategies to help mitigate this problem and ensure the produced documentation evolves over time would be helpful. What are your strategies for motivating attendees to attend a documentation based event? It's always challenging to get researchers to write docs, so it might also be hard to convince people to devote dedicated time over multiple days for this. It would be worth mentioning and coordinating with Bérénice Batut, who organizes similar events for the Galaxy Training Materials.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1974
https://f1000research.com/articles/7-1973/v1	24 Dec 18	{ "type": "Research Article", "title": "Clinical characteristics of migraine: A prospective cross-sectional study over nine years", "authors": [ "Khalid Obiad Mohsin Almohammadawi", "Haider Saadoon Qasim Alhilfi", "Rafid Adil Abood Alkhalidy", "Haider Saadoon Qasim Alhilfi", "Rafid Adil Abood Alkhalidy" ], "abstract": "Background: Migraine is the most common primary headache. This study aimed to describe clinical observations about migraine in outpatients in Iraq, including migraine types and subtypes, duration and frequency of acute attacks, severity, disability, effects on the quality of life, and complications. Methods: This is an outpatient-based prospective cross-sectional study, conducted in the Misan province, Iraq over nine years, and included 1412 patients aged 12 to 50 years. The data was collected from clinical records of patients who attended outpatient clinics. Results: The study included 1100 women (77.9%) and 312 men (22.1%); the women/men ratio being 3.5:1. The median age and standard deviation (SD) was 21 ± 5.42 years. The mean age at first attack of migraine was 17 ± 4.91 years. Migraine without aura was the most common type, accounting for 68% of the cases. The mean frequency of the attacks was (2 ± 4.63) days/month. In general, acute attacks were moderate to severe. Conclusions: In our study, we observed that migraine causes a headache resulting in episodes of temporary functional disability and women suffered more than men (ratio of 3.5:1). The mean age at first attack was a young age, and a family history of migraine highly altered distribution. Migraine without aura was the most common type, and symptoms including nausea and vomiting and photophobia were experienced by patients, which were used to diagnose migraines. Experienced aura was the most common migraine with aura, but also aura without a headache and aura with migraine were prevalent; therefore, it is important to differentiate between migraine subtypes. Visual aura was the most common aura, while motor symptoms were very rare. Chronic persistent headaches were a common complication recorded. These features provide evidence for the creation of screening tools in migraine prevention migraine.", "keywords": [ "Headache", "Primary headache", "Migraine", "Misan" ], "content": "Introduction\n\nMigraine is the most common chronic inherited neurovascular disorder. The onset of migraine is typically between 15 and 24 years of age1, and the prevalence is highest in patients aged 35–45 years, 75% of whom have moderate to severe headaches2,3. Diagnosis is based on clinical features, together with radiology images4. The majority of migraine patients experience temporary disability that affects their work and daily activities and, thus, their productivity and quality of life5–7. Regarding pathophysiology, the constriction and then dilation of cerebral blood vessels was believed, until 25 years ago, to cause the neurologic symptoms associated with a migraine8. In approximately 60 - 70% of patients with a migraine, the onset of the headache is preceded by a non-specific malaise and irritable feelings, such as euphoria, depression, food cravings, fatigue, hypomania, cognitive slowing, dizziness, or asthenia. These symptoms are called the migraine prodromes and may occur as early as 24 hours before the actual migraine9–11, followed by the “aura”, which is a focal neurological sign, and then a severe, throbbing headache with photophobia and vomiting12. About 15 - 20% of migraine patients experience aura within one hour of, or simultaneously with a headache13. The migraine aura consists of neurologic abnormalities, including visual loss, hallucinations, numbness, tingling, weakness, or confusion. The aura is due to the cortical spreading depression, a wave of abnormal electrical discharges that travels across the brain's surface and short-circuits the brain14. Furthermore, migraine is best conceptualized as a triad of a paroxysmal headache, nausea and/or vomiting, and an aura of focal neurological events (visual events)15. Patients with these three signs have migraine with aura (or “classical migraine”), while those with a paroxysmal headache (with or without vomiting) but do not have aura are classified as migraine without aura (or “common migraine”)16,17.\n\nThe aim of this study was to observe migraines clinically, leading to infer causality behind this disease, risk factors and triggers. Evidence gathered from this study will enable diagnosis of migraine and its probability to occur in persons who have similarity to patients observed in this study. We believe that early detection of these manifestations correlates with time and money saving on unnecessary investigations and medications used in management and treatment.\n\n\nMethods\n\nOutpatient-based prospective cross-sectional study, which was conducted in the Al-Sadder Teaching Hospital, Misan Province, Iraq, over nine years from 23rd January 2010 to 14th July 2018.\n\nThe total number of patients included in this study was 1510 and included all those that attended to outpatient clinics.\n\nThe cases were aged 12–50 years, of both genders, suffering from migraine headaches according to the criteria of the International Classification of Headache Disorders (ICHD-III b version)4,5.\n\nInclusion criteria were: 1–8 attacks over four weeks, attacks fulfilling the International classification of Headache Disorders migraine diagnostic criteria, and absence of secondary causes of headaches.\n\nExclusion criteria: migraine onset at age >50, headaches attributable to underlying organic disorders, or no migraine attacks during the four weeks of assessment (this is made case by case for each patient from the point of diagnosis till the end of the study).\n\nAll data were collected from participant records.\n\nFor all patients a full medical history and family history of migraine headaches was obtained, and a thorough clinical examination performed, including general examination, assessment of vital signs, Glasgow Coma Scale (GCS), neurological and physical examinations (as per the Seattle Children’s Hospital Research Foundation migraine general assessment pathway).\n\nAll participants were assessed by a standard questionnaire from the Migraine Relief Center and a 4-week headache diagnostic diary procedure (as per the Migraine Trust guidelines)4,18. Disability due to acute migraine attacks was determined using the Migraine Disability Assessment Scale (MIDAS) questionnaire.\n\nWe implemented standard descriptive statistics and data analysis using IBM SPSS Statistics Software (version 20.0, SPSS, Inc., Chicago, Illinois, USA). All p-values < 0.05 were considered statistically significant for on-sample t-test. Mean and standard deviation were used to present data.\n\nWritten informed consent was obtained from the patients or the parents/guardians of minors for those below age of 18 years, for participating in this study, and was conducted according to the ethical standards established by the 1964 Declaration of Helsinki. The Medical Ethical Committee of Misan University approved this study (code:270000425).\n\n\nResults\n\nA total of 1,412 patients with a migraine headache were included, including 1,100 women (77.9%) and 312 men (22.1%); women/men ratio of 3.5:1. Median age ±SD was 21 ± 5.42 years. The mean age at first attack was 17 ± 4.91 years. About 30 ± 15.79 mean±SD of the patients reported a family history of migraine (Table 1).\n\nInternational classification of headache disorder-3 version beta 2013\n\n1.1= Migraine without aura\n\n1.2= Migraine with aura\n\n1.2.1= Migraine with typical aura\n\n1.2.1.2 =Typical aura without headache\n\nA.1.3.2 Chronic migraine with continuous pain\n\nA.1.3 Chronic migraine (alternative criteria)\n\n8.2 Medication-overuse headache (MOH)\n\nA.1.1.1 Pure menstrual migraine without aura\n\nA.1.6 Episodic syndromes-Childhood periodic syndromes\n\nOther subtypes (Ophthalmoplegic ‘migraine’ Retinal migraine, Familial hemiplegic migraine (FHM), Sporadic hemiplegic migraine, Basilar-type migraine, Abdominal migraine, Benign paroxysmal vertigo of childhood)\n\nMigraine without aura was the most common (69.4%) subtype. The mean frequency of attacks was 2 ± 4.63 days per month. The mean duration of attack was 24 hours. Nausea and vomiting, photophobia and other nonspecific symptoms were experienced by 15%, 20%, and 12.5% of the patients respectively, while the remaining patients experienced non-specific prodromes 1–1.5 hours before the attacks (Table 1). About 27% of the patients in this study experienced aura during the period of study, the most common being migraine with aura, but also aura without a headache and aura with migraine (Table 1).\n\nMigraine prevalence rates per year in this study are shown in Table 2; migraine without aura was the highest recorded in 2016 as 73%, which is common subtype with the mean 67.6 ± 2.934. Migraine with aura, in 2012 recorded 10.2%. Chronic migraine with continuous pain presented in 7.5% in 2013, whereas prevalence of chronic migraine (alternative criteria) in 2014 was 3.4%. In 2013, migraine with typical aura recorded a high rate as 4.3%, but in 2010, it was reported 1.2% had typical aura without headache. The medication-overuse headache reported a high rate in 2013 as 3.7%. Pure menstrual migraine without aura, and episodic syndromes-childhood periodic syndromes reported high rates in 2018 as 5.5% and 1.8%, respectively. Finally, others subtypes of migraine present in 2017 with a high prevalence rate 2.3%.\n\nInternational classification of headache disorder-3 version beta 2013\n\n1.1= Migraine without aura\n\n1.2= Migraine with aura\n\n1.2.1= Migraine with typical aura\n\n1.2.1.2 =Typical aura without headache\n\nA.1.3.2 Chronic migraine with continuous pain\n\nA.1.3 Chronic migraine (alternative criteria)\n\n8.2 Medication-overuse headache (MOH)\n\nA.1.1.1 Pure menstrual migraine without aura\n\nA.1.6 Episodic syndromes-Childhood periodic syndromes\n\n*Others subtypes (Ophthalmoplegic ‘migraine’ Retinal migraine, Familial hemiplegic migraine (FHM), Sporadic hemiplegic migraine, Basilar-type migraine, Abdominal migraine, Benign paroxysmal vertigo of childhood)\n\nVisual aura was the most common (50%), while unilateral sensory symptoms, being second in frequency (42.17%). The transient dysphasic speech disturbance was the third most frequent (4.82%). Motor symptoms were very rare (0.6%), especially with a hemiplegic migraine (Table 3).\n\nThe duration (hours) and frequency (days per month) for migraine without aura, migraine with aura and chronic migraine with continuous pain exhibited are shown in Table 4. We also considered disabling symptoms, systemic blood pressure, changes in consciousness level (assessed by GCS in adult and pediatric groups, and trigger factors in relation to migraine without aura, migraine with aura and chronic migraine with continuous pain (Table 4).\n\nModified from, Stewart WF, et al. Reliability of the Migraine Disability Assessment score in a population-based sample of headache sufferers. Cephalalgia 1999;19:107-14\n\nMIGRAINE DISABILITY ASSESSMENT SCALE (MIDAS) QUESTIONNAIRE,GCS =Glasgow coma Scale\n\nOut of 1,412 patients with a migraine headache, enrolled in this study from 2010 to 2018, only a minority reported serious complications, such as chronic persistent headaches in 6.5% especially in migraine without aura and migraine with aura events (Table 5). The medication overuse headache 2.6% and thromboembolic stroke 0.7%, also recorded (Table 5).\n\n (International classification of headache disorder-3 version beta 2013 )\n\n*Mean age=10 years\n\n1.1= Migraine without aura\n\n1.2= Migraine with aura\n\n1.2.1= Migraine with typical aura\n\n1.2.1.2 =Typical aura without headache\n\nA1.3.2 a Chronic migraine with continuous pain\n\nA1.3 Chronic migraine (alternative criteria)\n\n8.2 Medication-overuse headache (MOH)\n\nA1.1.1 Pure menstrual migraine without aura\n\nA1.6 Episodic syndromes-Childhood periodic syndromes, Other subtype of migraine\n\n***Others subtypes (Ophthalmoplegic ‘migraine’ Retinal migraine, Familial hemiplegic migraine (FHM), Sporadic hemiplegic migraine, Basilar-type migraine, Abdominal migraine, Benign paroxysmal vertigo of childhood)\n\n\nDiscussion\n\nIn this study, 1,412 patients diagnosed with migraine headaches, according to established criteria4,19,20 were analyzed. Women were more affected (77.9%) than men (22.1%), and such a 3.5/1 female/male ratio is consistent with the results of large-scale studies6,10,12. This skewed sex ratio is mostly due to hormonal variation during menstruation and pregnancy, and to genetic predisposition1. The median age of first onset in this study was 21 ± 5.42 years, with range 12–45 years. Migraine without aura was the most common subtype (69.4%) in the sample. Childhood migraine prevalence was 0.8%, including migraine with aura and episodic syndromes/childhood periodic syndromes. The pediatric Glasgow Coma Scale (GCS) score in this subgroup ranged between 3 and 14, but this result was not significant. Patients experienced non-specific prodromes 1-1.5 hours before the attacks, including nausea, vomiting, and photophobia. About 27% of the patients in this study experienced transient aura during the study period. Visual aura was the most common (50%), while unilateral sensory symptoms, tingling and numbness was the second most frequent type of aura and transient dysphasic speech disturbance was the third, while motor symptoms were very rare. Aura occurs because of the spreading of a wave of depolarization (cortical spreading depression)20, and is associated first with a reduction, and then an increase in blood flow, and affects the parieto-occipital cortex10. The mean frequency of acute migraine attacks was 2 ± 4.63 days per month; in very few patients (0.5%) the frequency of the attacks was 14–16 days per month, especially in patients suffering from migraine with aura and chronic migraine. The mean duration of acute attacks was 12–24 hours in 60% of the patients. The severity of acute attacks depends on their frequency, duration and disabling symptoms; in general, most of the acute attacks were moderate to severe12. In our study about 8% of migraine patients suffered from debilitating, disabling and incapacitating symptoms. Symptoms were considered disabling, if they reduced by half or more the patient’s ability to work, or more generally to do what needs to be done, or at least 24 hours, thus impairing quality of life2,11,18. In this study about 10% of the patients had acute attacks associated with systemic hypotension (systolic blood pressure (BP) < 90 mmHg), especially in a migraine with aura, and women. Furthermore, systolic hypertension (≥ 150 mmHg) was found in 1.5% of the patients, especially in women. We found variation in the estimates of migraine prevalence and clinical characteristics symptoms, depending on the stage of the migraine (prodrome, aura, acute attack, and postdrome). One third of patients had a family history of migraine, showing that migraine is an inherited condition accompanied by episodic symptoms arising in the brain10,19–21. In our study chronic migraine with continuous pain and chronic migraine (according to alternative criteria) represented about 9% of the total sample, and had a detrimental influence on the patients’ lives, impacting socioeconomic functioning and quality of life. It usually develops from an episode of migraine, with or without aura, which turns into a continuum, with an undetermined annual conversion rate20,22. Another important subtype of a migraine is medication-overuse headache (MOH), which represented about 2.6% of the whole sample.\n\nThe limitation of this study were that it was a single center, local, regional, monophasic study and there was high financial cost involved. Therefore, for future studies we recommend a multi-centric, national, and diphasic study to obtain more information and data to help advance management of migraine.\n\n\nConclusions\n\nIn our study, we observed that migraine causes a headache resulting in episodes of temporary functional disability and women suffered more than men (ratio of 3.5:1). The mean age at first attack was a young age, and a family history of migraine highly altered distribution. Migraine without aura was the most common type, and symptoms including nausea and vomiting and photophobia were experienced by patients, which were used to diagnose migraines. Experienced aura was the most common migraine with aura, but also aura without a headache and aura with migraine were prevalent; therefore, it is important to differentiate between migraine subtypes. Visual aura was the most common aura, while motor symptoms were very rare. Chronic persistent headaches were a common complication recorded. These features provide evidence for the creation of screening tools in migraine prevention migraine.\n\n\nData availability\n\nF1000Research: Dataset 1. Excel sheet file of 1412 citizens from Misan province, Iraq whom suffer from migraine from 2010 to 2018., https://doi.org/10.5256/f1000research.16854.d22986123", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nStewart WF, Wood C, Reed ML, et al.: Cumulative lifetime migraine incidence in women and men. Cephalalgia. 2008; 28(11): 1170–8. PubMed Abstract \| Publisher Full Text\n\nStovner LJ, Hagen K, Jensen R, et al.: The global burden of headache: a documentation of headache prevalence and disability worldwide. Cephalalgia. 2007; 27(3): 193–210. PubMed Abstract \| Publisher Full Text\n\nGBD, 2015 Disease, and Injury Incidence and Prevalence Collaborators: Global, regional, and national incidence, prevalence, and years lived with disability for 310 diseases and injuries, 1990-2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet. 2016; 388(10053): 1545–1602. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nThe International Classification of Headache Disorders, 3rd edition (beta version) (ICHD-3b). Cephalalgia. 2013; 33(9): 629–808. PubMed Abstract \| Publisher Full Text\n\nThe first revision (ICHD-IIR1) (with changes affecting only section 8.2). Cephalalgia. 2005; 25(1): 460–465. Reference Source\n\nTepper SJ, Dahlôf CG, Dowson A, et al.: Prevalence and diagnosis of migraine in patients consulting their physician with a complaint of headache: data from the Landmark Study. Headache. 2004; 44(9): 856–864. PubMed Abstract \| Publisher Full Text\n\nLipton RB, Diamond S, Reed M, et al.: Migraine diagnosis and treatment: results from the American Migraine Study II. Headache. 2001; 41(7): 638–645. PubMed Abstract \| Publisher Full Text\n\nBahra A, Matharu MS, Buchel C, et al.: Brainstem activation specific to a migraine headache. Lancet. 2001; 357(9261): 1016–1017. PubMed Abstract \| Publisher Full Text\n\nMay A, Goadsby PJ: The trigeminovascular system in humans: pathophysiologic implications for primary headache syndromes of the neural influences on the cerebral circulation. J Cereb Blood Flow Metab. 1999; 19(2): 115–127. PubMed Abstract \| Publisher Full Text\n\nPeter JG, Richard BL, Michel DF: Migraine--current understanding and treatment. N Engl J Med. 2002; 346(4): 257–270. PubMed Abstract \| Publisher Full Text\n\nKelman L: The premonitory symptoms (prodrome): a tertiary care study of 893 migraineurs. Headache. 2004; 44(9): 865–872. PubMed Abstract \| Publisher Full Text\n\nGoadsby PJ: Recent advances in the diagnosis and management of migraine. BMJ. 2006; 332(7532): 25–29. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEriksen MK, Thomsen LL, Olesen J: Implications of clinical subtypes of migraine with aura. Headache. 2006; 46(2): 286–297. PubMed Abstract \| Publisher Full Text\n\nSalhofer-Polanyi S, Frantal S, Brannath W, et al.: Prospective analysis of factors related to migraine aura--the PAMINA study. Headache. 2012; 52(8): 1236–1245. PubMed Abstract \| Publisher Full Text\n\nHadjikhani N, Sanchez del Rio M, Wu O, et al.: Mechanisms of migraine aura revealed by functional MRI in human visual cortex. Proc Natl Acad Sci U S A. 2001; 98(8): 4687–4692. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDowson AJ, Massiou H, Aurora SK: Managing migraine headaches experienced by patients who self-report with menstrually related migraine: a prospective, placebo-controlled study with oral sumatriptan. J Headache Pain. 2005; 6(2): 81–87. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStewart WF, Lipton RB, Kolodner K, et al.: Reliability of the migraine disability assessment score in a population-based sample of headache sufferers. Cephalalgia. 1999; 19(2): 107–114; discussion 74. PubMed Abstract \| Publisher Full Text\n\nSteiner TJ; World Headache Alliance.: Lifting the burden: The global campaign against headache. Lancet Neurol. 2004; 3(4): 204–205. PubMed Abstract \| Publisher Full Text\n\nCharles A, Brennan K: Cortical spreading depression-new insights and persistent questions. Cephalalgia. 2009; 29(10): 1115–1124. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIizuka T, Takahashi Y, Sato M, et al.: Neurovascular changes in prolonged migraine aura in FHM with a novel ATP1A2 gene mutation. J Neurol Neurosurg Psychiatry. 2012; 83(2): 205–212. PubMed Abstract \| Publisher Full Text\n\nHansen JM, Goadsby PJ, Charles AC: Variability of clinical features in attacks of migraine with aura. Cephalalgia. 2016; 36(3): 216–224. PubMed Abstract \| Publisher Full Text\n\nMay A, Schulte LH: Chronic migraine: risk factors, mechanisms and treatment. Nat Rev Neurol. 2016; 12(8): 455–464. PubMed Abstract \| Publisher Full Text\n\nAlmohammadawi KOM, Alhilfi HSQ, Alkhalidy RAA: Dataset 1 in: Clinical characteristics of migraine: A prospective cross-sectional study over nine years. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16854.d229861" }	[ { "id": "69143", "date": "11 Aug 2020", "name": "Parisa Gazerani", "expertise": [ "Reviewer Expertise Neuroscience", "Neuropharmacology", "pain", "headache" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a published manuscript from 2018 dealing with clinical characteristics of migraine patients during a duration of 9 years of observation. This study has been conducted in outpatients in Iraq and has looked into several characteristics, including migraine types and subtypes, duration and frequency of acute attacks, severity, disability, effects on the quality of life, and complications.\nSuggested revisions:\n\nThe first sentence of the abstract background must be revised to: TTH is the most common primary headache disorder (ref: https://www.who.int/news-room/fact-sheets/detail/headache-disorders). Migraine might be the most common “disabling”, but the word is missing.\n\nThere’s a point here to consider that if this is a longitudinal or cross sectional? Over 9 years have the patients been followed up? Or at one time point they have been seen in a period of 9 years? i.e. each patient has been only seen once or more during 9 years?\n\nThe age range of 12 to 50 years includes puberty and pediatric migraine, the authors need to separate data presentation of children and adolescence from adults. It seems here teat pooled data have been presented.\n\nAnother point is the classification and diagnosis of migraine. The authors have mentioned that they have done this study with the aim of understanding migraine and to come in with better diagnose. This is a bit complicated, because we have the IHC (that authors have also followed the beta version) that is published and one must follow in diagnosis. Therefore, the suggestion is that perhaps authors wanted to know what are the specific characteristics of migraine patients in this geographical location, ethnic background, or racial effects on pattern and characteristics if we consider that most patients have been local with same ethnic and racial background? This must be clarified so that the purpose of the study can become clear. Besides, authors might have proposed that it might differences from rest of population or other parts of the world, or other racial or ethnic backgrounds, and that is why they wanted to characterize there for better clarification of some specific factors that might influence special populations around the word differently.\n\nWhat about the menstrual versus non-menstrual migraine. Did the authors follow that?\n\nWhat about the drugs (any drug, including analgesics) and comorbidities? Hormone therapy is only for women or?\n\nConclusion needs revision, because the authors need to compare their data with general data available or some other studies around the word to identify the commonalities and differences that might help in understanding differences of racial and ethnic differences in clinical characteristics of migraine patients.\n\nBesides, management means prevention, treatment, or both strategies here, is that correct? Authors need to discuss how based on their findings they may consider different strategies because this study does not really concern management so that cannot be concluded here.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] }, { "id": "69145", "date": "24 Aug 2020", "name": "Raffaele Ornello", "expertise": [ "Reviewer Expertise Migraine observational studies", "cardiovascular risk in migraine", "stroke epidemiology" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn their study, Authors assessed the characteristics of patients with migraine referring to a single center over 9 years. In my opinion, several points of the paper are worth clarifying. Please find my observations below.\nSome terms should be clarified as they are rather unconventional. For example, the term \"experienced aura\" mentioned in the Abstract should be better defined.\n\nAuthors state that they assessed \"migraine causality\" while in fact the cause of migraine is a research field.\n\nI suggest better explaining the rationale of the study and its novelty points as compared with previous studies. Comparisons with the available literature should be performed. Authors should also underline the differences between clinic-based and population-based studies.\n\nSome variables should be clarified. For example, Authors reported the number of monthly headache days without specifying over how many months they assessed the variable.\n\nI suggest reporting the absolute numbers together with proportions in the Tables.\n\nIn my opinion, some variables assessed by the Authors, such as the Glasgow Coma Scale score, have a poor rationale.\n\nWas the present study cross-sectional or longitudinal?\n\nI suggest adding some information about the patients' treatments in their medical history and about those prescribed during the visits.\n\nAuthors state that they assessed the level of disability associated with headache attacks. However, they did not assess the scores of standardized questionnaires such as the MIDAS or HIT-6, which constitutes a limitation.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] }, { "id": "70784", "date": "23 Sep 2020", "name": "Soodeh Razeghi Jahromi", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn methods you mentioned that 4-week headache diagnostic diary was used. As for assessing MIDS you need to record headache characteristics for 3 months, please explain how you assessed MIDAS based on one month headache dairy.\n\nPlease check the potential significant differences of confounding factors like age, sex, and prophylactic and abortive medication. For better comparison of headache characteristic, please adjust for the confounding factors which are significantly different between studied groups. If you have the data about abortive and prophylactic medication and the patients' comorbidities, it would be better to bring them in your article.\nIn table 4, by \"migraine with/without aura\" you meant episodic migraine with/without aura? If yes please revise it in the manuscript.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/7-1973
https://f1000research.com/articles/7-1641/v1	15 Oct 18	{ "type": "Opinion Article", "title": "Sharing health research data – the role of funders in improving the impact", "authors": [ "Robert F. Terry", "Katherine Littler", "Piero L. Olliaro", "Katherine Littler", "Piero L. Olliaro" ], "abstract": "Recent public health emergencies with outbreaks of influenza, Ebola and Zika revealed that the mechanisms for sharing research data are neither being used, or adequate for the purpose, particularly where data needs to be shared rapidly.\n\nA review of research papers, including completed clinical trials related to priority pathogens, found only 31% (98 out of 319 published papers, excluding case studies) provided access to all the data underlying the paper - 65% of these papers give no information on how to find or access the data. Only two clinical trials out of 58 on interventions for WHO priority pathogens provided any link in their registry entry to the background data.\n\nInterviews with researchers revealed a reluctance to share data included a lack of confidence in the utility of the data; an absence of academic-incentives for rapid dissemination that prevents subsequent publication and a disconnect between those who are collecting the data and those who wish to use it quickly. The role of the funders of research needs to change to address this. Funders need to engage early with the researchers and related stakeholders to understand their concerns and work harder to define the more explicitly the benefits to all stakeholders. Secondly, there needs to be a direct benefit to sharing data that is directly relevant to those people that collect and curate the data. Thirdly more work needs to be done to realise the intent of making data sharing resources more equitable, ethical and efficient. Finally, a checklist of the issues that need to be addressed when designing new or revising existing data sharing resources should be created. This checklist would highlight the technical, cultural and ethical issues that need to be considered and point to examples of emerging good practice that can be used to address them.", "keywords": [ "Health research", "data sharing", "public health emergencies", "data standards", "data infrastructure", "pandemics", "curation" ], "content": "Introduction\n\nThe benefits of sharing health research data to improve public health have been promoted by international research funders for over a decade but the reality is that the quality and volume of health research data shared, even in emergency situations, remains low1,2. This lack of progress seems to reflect a cultural reluctance among researchers to ‘give up their data’ without any clear benefits returning to them. This concern is heightened among researchers in low resource settings who feel that the requirements to share data, from funders and journals, risk turning them into data exporters unless greater efforts are made to ensure a fairer distribution of benefits. In this paper we draw on our experience of supporting data sharing initiatives and some commissioned research to highlight the barriers to sharing research data and the role research funders might play to improve this situation.\n\n\nA decade of progress?\n\nIn January 2011, a group of research funding organizations published a joint statement on sharing health research data with the aim to promote the efficient use of those data to accelerate improvements in public health. The funders recognized that for data sharing to be most effective, a combination of technical and cultural issues need to be addressed. They framed this approach around three principles which required any data sharing mechanism they supported to be equitable, ethical and efficient (See Wellcome Trust page on sharing research data). (See Box 1).\n\nEquitable: any approach to the sharing of data should recognise and balance the needs of researchers who generate and use data, other analysts who might want to reuse those data and the communities and funders who expect health benefits to arise from research.\n\nEthical: all data sharing should protect the privacy of individuals and the dignity of communities, while simultaneously respecting the imperative to improve public health through the most productive use of data.\n\nEfficient: any approach to data sharing should improve the quality and value of research and increase its contribution to improving public health. Approaches should be proportionate and build on existing practice and reduce unnecessary duplication and competition.\n\nProgress on encouraging the sharing of research data has been made over the subsequent decade and it is now common for research grants and journals to require the data underlying a paper or clinical trial to be shared (see PLOS editorial and publishing policies, AllTrials, and NIH data sharing policy.) However, recent public health emergencies with outbreaks of influenza, Ebola and Zika have brought into sharp focus the realization that the mechanisms for sharing data are neither being used or adequate for the purpose, particularly where data needs to be shared rapidly3–5.\n\nIn addition, researchers working in low- and middle-income countries highlight an inequity created by the disadvantage as they see it by the blanket requirements to share their data. Their concern is that sharing their data too soon, or without any restrictions will lead to their data being analysed by others with greater capacity, and no benefit will return to the researchers themselves or the populations they work with. In effect they become data exporters rather than partners. So while there is a lot of emphasis placed on data being Findable, Accessible, Inter-operable and Reusable, known as the FAIR approach, many researchers in developing countries fear the reality for them will be far from fair6–8.\n\n\nThe findings of two surveys and a workshop\n\nTo explore this further, we commissioned two surveys to review the governance arrangements and standards within existing data sharing resources. The findings of those studies informed a workshop held in October 2017 with a set of stakeholders representing researchers and funding organizations. All the reports and supporting files are published as open access under a Creative Commons licence and in free-to-access repositories. Readers are strongly encouraged to read that material as the primary source of reference1,2,9,10.\n\nThe first survey – Data Sharing in Public Health Emergencies - focussed on data sharing in public health emergencies concerned with the pathogens named by the World Health Organization as of priority concern because of their epidemic or pandemic potential (see WHO list of Blueprint priority diseases). A review of academic papers published since 2003 relating to these diseases was undertaken and attempts were then made to access the data underlying those publications via the web and through a direct survey of the corresponding authors. Interviews were undertaken with a range of people either conducting or supporting research in these areas and this was supplemented with a review of institutional policies, discussion documents and academic commentaries about standards and norms in data sharing1,2.\n\nThe second survey - Development of International Standards for Online Repositories - was designed to identify which ‘standards’ were being used in data sharing relating to the neglected diseases. Standards were identified following a review of publically accessible information (via the web or publication) relating to three main areas each with a set of elements describing the standards under those areas9.\n\nA third report combined the findings of these two surveys and was used to shape thinking at a workshop held in Antwerp, Belgium in October 201710.\n\nThe workshop brought together 26 experts representing agencies that included those that provide data sharing resources for diseases prevalent in low and middle income countries.\n\n\nWhat does this tell us?\n\nSharing health research data currently remains the exception rather than the norm. The review of research papers, including completed clinical trials related to priority pathogens, found only 31% (98 out of 319 published papers, excluding case studies) provided access to all the data underlying the paper. While a few authors will provide the data on request, 65% of these papers give no information on how to find or access the data. And the review of clinical trial registries, for trials on interventions for priority pathogens, reported an even worse picture. Only two trials out of 58 provided any link in their registry entry to the background data1,2.\n\nInterviews with researchers revealed the reasons for a reluctance to share data included a lack of confidence in the utility of the data and therefore unwillingness to invest resources to prepare it to be shared; absence of academic-incentives for rapid dissemination that prevents subsequent publication (as opposed to the public health need) and a disconnect between those who are collecting the data and those who wish to use it quickly. A similar scepticism about how data might be used or misused, the potential harms to patients and the risks to the researcher sharing data that might reveal errors in their work, have been reported elsewhere8,11.\n\nTable 1 summarises the survey findings that identified which standards are used to share research data for neglected diseases and what those standards cover with respect to data curation, governance, security and longevity. Whilst there is clearly no universal or single standard to cover all the three areas and the elements under them, technical guidance is available across all the areas when those standards are combined. The standards created by the Clinical Data Interchange Standards Consortium (CDISC) were included as the United States Food and Drug Administration (FDA) has required the use CDISC standards in a clinical trial data submission since 2017. Hence these are widely used in industry and CDISC is fast becoming the de facto standard for data labelling and meta data.\n\n+ As stated in publicly available information\n\n- Information was not mentioned in the publicly available information.\n\nTRAC (Trustworthy Repositories Audit & Certification), ISO 16363 (International Standards for Clinical Trial Registries -Space data and information transfer systems, Audit and certification of trustworthy digital repositories), WHO (World Health Organization), ICSU (International Council of Scientific Unions World Data System), H3Africa (The Human Heredity and Health in Africa Initiative), CDISC (Clinical Data Interchange Standards Consortium).\n\n\nWhat are the next steps: the role for funders in support of data sharing\n\nMany health research funders have a generic policy requiring research data to be shared in a manner that maximises health and societal benefit. While some biomedical areas, like genomics, have forged ahead in maximizing data sharing, across health research more generally there is very low compliance with these policies. In part this might reflect the limited guidance offered by the same funders in supporting their researchers to understand and undertake data sharing to implement and monitor these policies in practice.\n\nIt appears the main barrier to sharing is not technical but cultural, with researchers remaining sceptical about the benefits to them of sharing data. For researchers in low-resource setting data sharing can even be seen as a threat that their data will be exported and exploited by others with little benefit returning to them.\n\nTherefore, research funders should take stock and revise data sharing policies to provide incentive structures for researchers. One clear first step would be to engage early with the researchers and related stakeholders to understand their concerns and work harder to define the benefit of sharing beyond a general sense that sharing data is in the public interest. The overall purpose of sharing the data needs to be clear and ideally developed with input from data suppliers, secondary data users, potential end-users and beneficiaries, and if possible with input from the participants that are the source of those data. Concerns regarding privacy versus the secondary use of the data need to be explored and mechanisms put in place to balance the public benefit against potential risks to privacy and confidentiality.\n\nSecondly, there needs to be a direct benefit to sharing data that is directly relevant to those people that collect and curate the data. For example academics require citation of their work, including a data set. The generation of data and its subsequent citation for reuse needs to be integrated into research assessment – an idea captured in the Declaration on Research Assessment (see San Francisco Declaration on Research Assessment). So if the purpose of the data sharing mechanism is clear and all stakeholders buy into that purpose and if they feel their inputs will be recognised in research assessment together this will create a strong incentive to share. This was certainly our experience when working with Schistosomiasis researchers8.\n\nThirdly, whilst there are a myriad of data standards to work with to meet the general principles of making data FAIR, more work needs to be done to realise the intent of making data sharing resources more equitable, ethical and efficient. As evident in the surveys summarized here good practice is starting to emerge so what is needed is better ways to share that practice. Funders need to work with the researchers and their networks to support the technical work required to develop standards that enable inter-operability.\n\nFor example one contributory role for funders would be to collect more systematically the data management plans that they have requested as part of funding grants and make them publicly accessible. In line with good practice these should be standardized where possible and ideally have clear, machine-readable metadata. An online resource that brings together the reference material and policies that are exemplars in each of the categories that cover governance, data curation, security and longevity would provide the basis for a framework to guide the future development of new sharing resources.\n\nFinally, a checklist of the issues that need to be addressed when designing new or revising existing data sharing resources should be created. In addition to defining the purpose of data sharing this would highlight the technical, cultural and ethical issues that need to be considered and point to examples of emerging good practice that can be used to address them. The authors are working on this next stage and hope that with this type of planning and support in place the data sharing long desired by research funders will start to become the norm.\n\n\nData availability\n\nNo data are associated with this article.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nPisani E, Ghataure A, Merson L: Data sharing in public health emergencies: A study of current policies, practices and infrastructure supporting the sharing of data to prevent and respond to epidemic and pandemic threats. Wellcome Trust. 2018. Publisher Full Text\n\nPisani E, Ghataure A, Merson L: Supporting Data for: Data sharing in public health emergencies: A study of current policies, practices and infrastructure supporting the sharing of data to prevent and respond to epidemic and pandemic threats. Harvard Dataverse, V1. 2017. Publisher Full Text\n\nYozwiak NL, Schaffner SF, Sabeti PC: Data sharing: Make outbreak research open access. Nature. 2015; 518(7540): 477–9. PubMed Abstract \| Publisher Full Text\n\nPisani E, Aaby P, Breugelmans JG, et al.: Beyond open data: realising the health benefits of sharing data. BMJ. 2016; 355: i5295. PubMed Abstract \| Publisher Full Text\n\nOlliaro PL: Initiation and publication time-lags of treatment trials for Ebola virus disease. Lancet Infect Dis. 2016; 18(1): 28–29. PubMed Abstract \| Publisher Full Text\n\nWilkinson MD, Dumontier M, Aalbersberg IJ, et al.: The FAIR Guiding Principles for scientific data management and stewardship. Sci Data. [Internet]. 2016; 3: 160018, [cited 2017 Mar 13]. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHate K, Meherally S, Shah More N, et al.: Sweat, Skepticism, and Uncharted Territory: A Qualitative Study of Opinions on Data Sharing Among Public Health Researchers and Research Participants in Mumbai, India. J Empir Res Hum Res Ethics. 2015; 10(3): 239–250. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchistosomiasis Data Platform Stakeholder Meeting Report. World Health Organization, Geneva. 2015. Reference Source\n\nCastillon G, Castilloux AM, Moride Y: Development of Standards for Online Repositories. Wellcome Trust, Version 3. 2017. Reference Source\n\nSharing health research data in low resource settings: Supporting necessary infrastructure and building on good practice. Pisani E. Report prepared for Wellcome Trust and TDR the Special Programme for Research and Training in Tropical Diseases. 2018. Publisher Full Text\n\nCommittee on Population; Division of Behavioral and Social Sciences and Education; The National Academies of Sciences, Engineering, and Medicine: Sharing Research Data to Improve Public Health in Africa: A Workshop Summary. Washington, DC: The National Academies Press, 2015. PubMed Abstract \| Publisher Full Text" }	[ { "id": "39449", "date": "25 Oct 2018", "name": "Phaik-Yeong Cheah", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nHere are my comments:\n\nFindings of two surveys and a workshop\n\nFirst paragraph, first sentence - Who is “we”? The three authors of this paper?\n\nSecond paragraph - The authors talked about a review of academic literature. Is this part of the survey or workshop? Perhaps this should be in the Introduction.\n\nThird paragraph – Suggest briefly stating the three main areas found in the survey.\n\nFourth and fifth paragraph - Please tell us a little bit more about the third report and the workshop.\n\nWhat does this tell us? Does this refer to the findings of the two surveys and workshop?\n\nParagraphs 2 and 3 – It is not clear what the objective of these paragraphs are. Do they summarise the barriers to data sharing? The reason I mentioned barriers is that in the Introduction, the authors said that this paper does two things – (1) to highlight the barriers to sharing research data and (2) the role research funders might play to improve this situation.\n\nIf indeed they are barriers, are these barriers based on the two surveys and workshop or do they include research conducted by other groups? If the latter, I believe there are more barriers such as costs and ownership issues.\n\nParagraph 2, first sentence. “Interviews with researchers”. Please provide reference(s). Did the authors conduct these interviews?\n\nWhat are the next steps: the role for funders in support of data sharing\n\nParagraph 1 – “…there is very low compliance with these policies”. How did the authors get this information? Please provide references if appropriate.\n\nParagraph 2 - “It appears the main barrier to sharing is not technical but cultural,…” How did the authors get this information? Please provide references if appropriate.\n\nParagraph 4 – “There needs to be direct benefit….” This is a general statement. What is the funders’ specific role?\n\nSuggest including a “Conclusions” section.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Partly\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Partly", "responses": [] }, { "id": "40329", "date": "19 Nov 2018", "name": "Amanda Blatch-Jones", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nSharing research data is important and the authors provide some valid points for consideration by funders and researchers. The article could do with some revisions in order to better flow and understand more about the survey and interviews.\n\nThe abstract doesn't state that two surveys and interviews were conducted and when - this would help bring paragraph 1 and 2 together. Also, the review of papers, was this a literature review/systematic review? The third paragraph could do with some revisions especially the first sentence. I wasn't sure what \"Funders need to engage early with the researchers and related stakeholders to understand their concerns and work harder to define the more explicitly the benefits to all stakeholders.\" sentence means? What stakeholders?\nIntroduction - improve what situation? I think this needs teasing out more in the first paragraph. Improving what - the quality of data to be shared, data sharing policies? What group of funders - who was involved? They framed this approach....Who framed? Highlighting that data collected specifically to answer a specific question in a trial for example, may not be sufficient to answer a different question, is important. Understanding why and how the data were collected, analysed and interpreted is important when reviewing secondary data. If this is misinterpreted, there are risks associated to the use of that data.\n\nFindings of the two surveys and workshop - Who was involved in the survey, how many people involved and what was the response rate? Were funders involved in this? (For both of the surveys) Why was the review from 2003, given that data sharing para from 2011? As expected the results of the review are low, so are these numbers from recent research? What is the time period of those that made access to data? Did you see a change over time, those identified are from 2015 onwards for example.\n\nWho participated in the workshop? Agencies from where? I think this would be important given the focus on low to middle income countries.\n\nHow many people interviewed? And who were they? Was this semi-structured/structured interviews, and how did you interpret the transcripts? Providing some context will help determine the value of the comments and confidence about the interpretation of the interviews.\n\nConcluding section - Ethical issues may also need to be reviewed, are participants made aware at time of consent that the data collected for the specific study can later be used [secondary data] for other studies? Data sharing needs to be made aware to those who participate in the research. Transparency of use of data is needed at all levels, participants, researchers and funders. Funders have a role to implement and provide policies to ensure proper use of and sharing of data, and researchers have a role to ensure that they comply, acknowledge the use of the data and provide sufficient information about the data set and its intended purpose.\n\nIn terms of publicly accessible data management plans, is this the role of the funder and/or the researchers? If research protocols are 'meant' to be in the public domain then data sharing and statistical analysis plans should be as well? All funders approach these areas differently, so caveats around this may need to be mentioned. Possibly further consideration of these policies and how they could be standardised across funders, although consideration of early phase research to phase three studies may be required.\n\nIs this applicable to industry and pharma? How does this relate to them considering the sensitivity of data? How does this relate to AllTrials? (http://www.alltrials.net/)\n\nHow does data sharing fit into the wider agenda of transparency, accountability and credibility of research? Stating the larger viewpoint may strengthen the paper.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Partly\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes", "responses": [] }, { "id": "40330", "date": "03 Dec 2018", "name": "Keith C. Norris", "expertise": [ "Reviewer Expertise Clinical trials", "community engagement", "workforce diversity", "health disparities" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is well thought through and timely opinion piece on the barriers to research data sharing, especially for issues of public health emergencies.\n\nThe authors report on the findings from two related surveys and a workshop, and of note reported out of 319 published papers, excluding case studies, only 98 (31%) provided access to all the data underlying the paper.\n\nThey also reported 65% of these papers gave no information on how to find or access the data (does the 65% refer to the 319 papers or the 98?).\n\nThey provide several recommendations, but outside of the checklist the recommendations seem more aspirational and more about why one would want to or should share data. But as framed are not easily actionable.\nFirst - Funders need to engage early with the researchers and related stakeholders to understand their concerns and work harder to define the more explicitly the benefits to all stakeholders. Second - there needs to be a direct benefit to sharing data that is directly relevant to those people that collect and curate the data. Third - more work needs to be done to realize the intent of making data sharing resources more equitable, ethical and efficient. Finally - a checklist of the issues that need to be addressed when designing new or revising existing data sharing resources should be created. This checklist would highlight the technical, cultural and ethical issues that need to be considered and point to examples of emerging good practice that can be used to address them.\n\nThe discussion needs to capture more of the following: There are legitimate concerns about early or premature data sharing that are not discussed, especially focusing on low resource research entities and the power dynamics that leverage the early release of data to the public as an opportunity for resource intensive researchers to leverage the data not only for “public good” but for their gain with no need for attribution to or sharing of downstream resources with the low resource entities that conducted the work. While many health research funders have a generic policy requiring research data to be shared in a manner that maximize health and societal benefit, these are often unfunded mandates, especially for research conducted in low resource countries or by low resource institutions.\n\nReleasing data in the absence of the context of much of the data can lead to significant misinterpretation of the data. Also, for certain data there needs to be time to clean and collapse the data to ensure participant privacy before it can be made publicly available. This can be a laborious process to do correctly and in many instances there may not be funding for this provided by the funder and low resource institutions may not have resources to conduct this unfunded work in an expeditious manner.\n\nFor WHO pathogen work it seems funders might consider an a priori explicit plan of a cooperative agreement/partnership for early shared data for studies led by low resource “partners” doing the hands-on work and a more resource intensive partner (or funder) with capacity to do intense data analysis, with shared credit and equity in leveraging the data to get more funding for both partners. This can help to balance the power dynamics, and support early widespread data dissemination with the resource intensive partner (or funder) co-leading rapid production of publications and the creation of publicly available cleaned data for public benefit.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Partly\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Partly", "responses": [] } ]	1	https://f1000research.com/articles/7-1641
https://f1000research.com/articles/7-594/v1	15 May 18	{ "type": "Method Article", "title": "A method for transplantation of human HSCs into zebrafish, to replace humanised murine transplantation models", "authors": [ "Noémie Hamilton", "Ian Sabroe", "Stephen A. Renshaw", "Ian Sabroe", "Stephen A. Renshaw" ], "abstract": "Haematopoietic stem cell (HSC) transplantation is a critical therapy for haematopoietic malignancies and immune disorders. Incomplete or delayed engraftment of HSCs in the host results in increased risk of infection and morbidity. The mechanisms of HSC engraftment are poorly understood and understanding these processes will increase transplantation success on many levels. Current animal models are immunocompromised 'humanised' mice transplanted with human HSCs. Harmful procedures include genetic manipulations and irradiation to ablate the mouse immune system, and opaque mouse tissues make visualisation of the early steps of HSC engraftment impossible. There is a need for new models to offer alternatives to humanised mice in the study of HSC transplantation. Here we described a detailed method for transplantation of human HSCs into zebrafish, before the onset of adaptive immunity. Human HSCs were purified from whole blood by enrichment of the CD34 cell population using a positive magnetic selection and further purified using an anti-CD34 antibody and cell sorting. Sorted CD34 cells were transplanted into the blood stream of 52 hour old zebrafish larvae. Human HSCs home into the zebrafish haematopoietic niche, where they engage with endothelial cells and undergo cell division. Our model offers the opportunities to image in vivo human HSC engraftment in a transparent organism, without the myeloablative strategies used in mice, and provides a unique system to understand the dynamic process of engraftment and replace current murine models. This technique can be applied to current engraftment protocols to validate the viability and efficiency of cryofrozen HSC grafts. This humanised zebrafish model will be instrumental to develop the 3Rs values in stem cell transplantation research and our detailed protocol will increase the chances of uptake of this zebrafish model by the mouse community.", "keywords": [ "zebrafish", "stem cell transplantation", "xenograft", "humanised zebrafish" ], "content": "\n\n\n\nZebrafish embryos are transparent and represent a tractable system for imaging Zebrafish do not develop an adaptive immunity for the first 2 weeks of life, providing a large time window to perform xenotransplantation\n\nZebrafish larvae can be used to replace mouse models in stem cell transplantation research\n\nUse of zebrafish larvae also avoids subjecting mice to severe irradiation procedures and eliminates the risk of contracting fatal infections\n\nZebrafish are cheaper to raise and host in fish aquariums that can hold thousands of animals\n\nA pair of Zebrafish produces hundreds of small embryos easily transplanted therefore offering the opportunity to perform high-throughput experiments\n\nStudying the engraftment mechanism of human HSCs\n\nDrug screens to identify new drugs to improve engraftment rate of HSCs\n\nIdentifying new human HSC markers to improve the rate and speed of engraftment\n\nAssessing the viability and efficacy of human HSC grafts before transplanting into patient\n\nReplace the use of mouse models during the optimisation phase of HSC transplantation protocols\n\n\nIntroduction\n\nTransplantation of healthy haematopoietic stem cells (HSCs) is a critical therapy for a wide range of malignant haematological and non-malignant disorders and immune dysfunction (Snowden et al., 2012; Sykes & Nikolic, 2005; Thomas et al., 1957). In successful stem cell transplantation (SCT), immune reconstitution following ablation of native immunity leads to the recovery of immune function. Healthy transplanted stem cells home to haematopoietic niches in the host and differentiate into multi-lineage blood cells, providing the patient with a new immune system (Sullivan et al., 2010). Around 2000 people in the UK are in need of SCT every year, and as more hospitals are performing this high-risk life-saving procedure, there is a growing need in improving current protocols. (http://www.anthonynolan.org)\n\nHSCs are collected from: 1) blood harvested from peripheral blood by apheresis following mobilisation by G-CSF cytokine treatment; 2) umbilical cord blood; or 3) bone marrow from donors or patients. HSCs are enriched post-collection by positive selection for the CD34 stem cell marker. Conditioning or myeloablation of the host bone marrow by chemotherapy is necessary to ablate malignant or autoreactive immune populations, adding a considerable risk of infection while engraftment occurs. Incomplete or delayed engraftment results in delayed immune system recovery, increasing considerably the risk of infection and associated morbidity and mortality. The regulating mechanisms of the homing and migration steps of HSC engraftment are poorly understood and understanding these processes will increase transplantation success on many levels. Accelerated and more complete engraftment will reduce morbidity and mortality associated with transplantation, both during engraftment and long-term immune recovery.\n\nThe only models currently employed to study human HSC engraftment are immunocompromised mice transplanted with human HSCs, also called 'humanised' mice (Tanner et al., 2014). Although these mouse models informed current stem cell transplantation protocols, they involve prolonged harmful procedures and it remains difficult to assess and visualise the early steps of engraftment due to the opacity of their tissues. Multiple mouse strains have been generated to create suitable immunocompromised hosts to allow engraftment of a fully developed adaptive immune system (Tanner et al., 2014). In most studies, additional harmful irradiation regimes are used to prevent early rejection of the transplant by the immune system. These immunodepleted mice must be grown to adulthood in order to assess engraftment success, usually performed shortly after birth. These genetically altered mice live their entire lives undergoing severe procedures with high maintenance requirements, since they need to be homed in sterile rooms and fed sterile food to avoid fatal infection due to defective immunity. Published articles test multiple conditions on groups of 5 to 6 mice sacrificed at various time points, resulting in an average of 40 mice per publication. In 2015, 25 publications used immunocompromised mice for HSC transplant studies, representing around 1000 mice each year worldwide – all undergoing severe procedures over a long period of time. As the demand for stem cell transplantation therapy increases, more efficient and less dangerous procedures will be demanded, which will require an even higher mouse usage to optimise current protocols.\n\nZebrafish are already established as a successful model to study the haematopoietic system, with significant homology with mammals (de Jong & Zon, 2005; Gering & Patient, 2005; Kissa & Herbomel, 2010; Renshaw & Trede, 2012; Traver et al., 2003; White et al., 2008). Imaging of zebrafish transparent embryos remains a powerful tool and has been critical to confirm that the zebrafish Caudal Haematopoietic Tissue (CHT) is comparable to the mammalian foetal haematopoietic niche (Gering & Patient, 2005; Kissa & Herbomel, 2010; Tamplin et al., 2015). Xenotransplantation in zebrafish embryos has revealed highly conserved mechanisms between zebrafish and mammals. Recently, murine bone marrow cells were successfully transplanted into zebrafish embryos, revealing highly conserved mechanism of haematopoiesis between zebrafish and mammals (Parada-Kusz et al., 2017). Additionally, CD34 enriched human cells transplanted into zebrafish were shown to home to the CHT and respond to zebrafish stromal-cell derived factors (Staal et al., 2016).\n\nWe propose that transplanting human HSCs into zebrafish larvae, before the onset of adaptive immunity, will offer unprecedented in vivo opportunities to understand stem cell engraftment and help to shift current research towards a 3Rs approach to reduce and refine, and finally replace the usage of mice in HSC transplant studies. Here we describe a detailed transplantation protocol of pure human HSCs into zebrafish larvae. Human PBMCs were enriched for CD34 cells and further purified by cell sorting using the HSC marker CD34. Transplanted of human HSCs into 52hpf larvae was achieved by injection into the Duct of Cuvier. We have evidence that human HSCs home to the zebrafish CHT, where they interact with endothelial cells and undergo cell division. This conserved engraftment mechanism makes zebrafish a unique model to study HSC engraftment and we wish to highlight the significant opportunities to impact on reductions in mammalian model usage. This could lead to new clinical applications to improve the speed and extent of human HSC engraftment.\n\nHumanised zebrafish could offer a welfare improvement compared to current mouse models, as early zebrafish larvae do not require immunodepletion by irradiation or multiple genetic modifications to avoid graft rejection. Zebrafish do not develop functional adaptive immunity until 2 weeks of age and therefore do not require severe procedures if the transplantation occurs in this time window (Langenau et al., 2004). Using a model with substantially reduced risk of fatal infection and eliminating the need for irradiation significantly refines the current substantial severity protocols.\n\nAdditionally, upon transplantation of human stem cells, mice must be grown for several months to assess engraftment success by analysing the reconstitution of the immune system. The transparency of the zebrafish larvae offers a unique system allowing direct live imaging of transplanted cells to visualise cell behaviour and interactions. This will allow the selection of only successful engrafted animals for further analysis and will therefore improve experimental design and throughput whilst simultaneously reducing animal numbers.\n\nMoreover, humanised mice are still being used to optimise protocols of HSCs transplantation, such as source, type and number of cells transplanted and testing different expansion protocol (Tanner et al., 2014). High-throughput assays can easily be performed using zebrafish larvae, with the scientific advantage of generating a broader range of outputs at lower cost. This, combined with transparency of the larvae to quickly assess engraftment, we expect that our model may be used to replace the use of mouse models during these optimisation phases. Mammalian models could then be reserved purely for the analysis required by law before clinical trials are performed. This could replace all the murine models currently used to optimise protocols with zebrafish larvae before the onset of independent feeding – considered in law and ethics as a model of significantly lower neurophysiological sensitivity.\n\nFinally, humanised mice used to study human stem cell transplantation require an average of 1×105 CD34+ cells per animal. 100ml of blood yields approximately 1×108 peripheral blood mononuclear cells (PBMCs) from which 0.1%, or 1×105 CD34+ cells are routinely isolated, enough for just one mouse. However, we have demonstrated that zebrafish larvae only need 20 to 50 cells transplanted to successfully engraft. Thus, replacing mouse models by a smaller non-protected vertebrate will allow testing of multiple conditions into multiple animals from the same human donor. This will decrease natural variations therefore improving reproducibility and reliability of research performed in this field.\n\n\nMethods\n\nA detailed protocol of the procedure is available in Supplementary File 1.\n\nZebrafish (Danio Rerio) were raised and maintained under the Animal [Scientific Procedures] Act 1986 (Home Office Project Licence 70/8178 used to raise and maintain transgenic lines) using standard protocols (Nüsslein-Volhard & Dham, 2002). Zebrafish adults were hosted in UK Home Office-approved aquaria at the Bateson Centre, University of Sheffield, and kept under a 14/10 light/dark regime at 28 degrees. The endothelial cells transgenic reporter line Tg(kdrl:HRAS-mCherry-CAAX) (allele code s916) (Chi et al., 2008) is referred to as kdrl:mCherry in the manuscript.\n\nBlood was taken from healthy volunteers (pool of 9 donors, males and females, aged between 18 and 40) with written informed consent and ethical approval from the South Sheffield Research Ethics Committee (STH18729). Human PBMCs were collected after routine neutrophil preparation by dextran sedimentation followed by plasma-Percoll gradient centrifugation from whole blood (Prince et al., 2017). PBMCs were further purified by positive selection using the CD34 MicroBead kit (Miltenyl Biotech, Bergisch Gladbach, Germany).\n\nCD34 enriched PBMCs were labelled with anti-Human CD34-eFluor450 antibody (eBioscience- RRID:AB_10734946) and positive cells were sorted using Fluorescence Activated Cell Sorting (FACS). Sorted cells were labelled with Fluorescein and injected into the Duct of Cuvier in 52hpf zebrafish.\n\nZebrafish larvae were sedated in Tricaine and embedded in 0.8% low melting point agarose. High-resolution imaging was performed using a Spinning Disk confocal microscope.\n\nSample size (n=10) is represented by number of larvae used to count human HSCs present in the CHT or within perivascular pockets. Paired T-test was used to assess significance using free GraphPad software online.\n\n\nResults\n\nIn human stem transplantation therapy, successful transplantation correlates with high number of injected CD34 positive cells (Zaucha et al., 2001). To produce a population of human cells that would mimic a human HSC graft, we used the CD34 marker to purify HSC from whole blood. The method, summarised in Figure 1 and detailed in Supplementary File 1, consists of multiple steps in order to ensure robust and consistent purification of CD34 cells. Whole blood was collected from healthy volunteers and immediately processed to extract neutrophils and PBMCs. The PBMC fraction was then enriched for CD34 cells using a positive selection magnetic column. These cells were then further labelled with a human anti-CD34eFluor450 antibody and sorted using FACS. Only cells positive for the eFluor450 fluorophore were sorted; therefore ensuring a pure CD34 cell population (Figure 2).\n\nWhole blood preparation by Percoll gradient allowed us to separate peripheral blood mononuclear cells (PBMCs) from neutrophils and red blood cells (RBCs). PBMCs were enriched for CD34 cells using a positive selection magnetic column. A pure CD34 cell population was sorted using a human anti-CD34eFLuor450 antibody by FACS. CD34 cells were labelled using fluorescein and injected into the Duct of Cuvier of 52 hour post fertilisation zebrafish larvae. Animal with human cells in their CHT were selected for further high-resolution imaging.\n\n(A) Healthy cells only were gated to analyse fluorescence. (B) A clear cell population (black rectangle) of small cells was positive for the Violet450 fluorophore as determined by the no antibody control (C) where that same cell population is shifted to the left of the X-axis.\n\nCD34 cells represent a small fraction of PBMCs. During each experiment, cells were counted at each specific point of the protocol and expected ranges of cells have also been noted on the protocol. The volume of blood taken varied between 50ml and 180ml (left axis Figure 3). Cell number was counted on a haemocytometer after each important step of the protocol. Number of cells after PBMCs isolation varied between 83 and 162.5 millions, and after red blood cell (RBC) lysis numbers ranged from 50.6 and 149.6 millions. Of note, our results show no significant difference in PBMC number after RBC lysis (Figure 3, n=14). After CD34 enrichment, cells were counted again and varied between 0.152 and 6.15 millions. Finally, after cell sorting, we recorded a range of pure CD34 cells between 3000 and 100,000. As expected, as the purity of CD34 cells increased, the cell number dramatically decreased (Figure 3). On average, CD34 positive cells represented 0.033% of total PBMCs recovered from the cell preparation (n=10). Moreover, paired Pearson correlation analysis was performed between the blood volume taken and the final number of sorted CD34 cells and no correlation was found (p= 0.115, n=14, Pearson r=0.441). This may be due to the high variability in the pool of CD34 cells between donors.\n\nLeft scale represent the blood volume taken per donors. Paired T-test was used to analyse statistical significance between ‘after blood prep group’ and ‘after red blood cell (RBC) lysis group’ (n=10).\n\nInjected human CD34 cells adhere to the zebrafish CHT. Purified human CD34 cells were labelled with fluorescein and injected into the blood circulation by targeting the Duct of Cuvier in 52hpf zebrafish larvae (Figure 1). We first observed that human CD34 cells are visible in the zebrafish CHT immediately after injection (Figure 4A) where they appeared to adhere to the endothelial wall of the blood vessels forming the CHT. Subsequently, instead of being washed away from the CHT by the blood flow, human CD34 cells were seen to roll and tether along the caudal vein. This behaviour has previously been reported for endogenous zebrafish HSCs, known to emerge from the ventral wall of aorta to join the circulation and roll along the endothelium of the caudal vein to reach their haematopoietic niche (Gering & Patient, 2005; Kissa & Herbomel, 2010; Tamplin et al., 2015). We observed that 100% of injected human CD34 cells initially adhere to the wall of the caudal vein in the CHT within one hour after injection (Figure 4B). Imaging 12 hours later showed that only 50% of these cells are still present in the CHT (Figure 4C).\n\n(A) Stitched Z-stack of whole Zebrafish larvae trunk highlighting the CHT (white rectangle). (B) Representative Z-Stack images of fluorescein labelled human CD34 cells present at the CHT at 1hour post transplantation (hpt), 5hpt, 9hpt and 13hpt. Scale bar=80μm. (C) Quantification of cells in the CHT versus circulating cells in 5 larvae injected with fluorescein labelled human CD34 cells (n=5).\n\nHuman CD34 cells engraft in zebrafish haematopoietic niches. It is known that zebrafish and mouse HSCs, once adhered to the caudal vein, enter perivascular pockets proximally to the caudal vein (Figure 5A, white arrowheads) (Kiel et al., 2005; Nombela-Arrieta et al., 2013; Tamplin et al., 2015). To assess whether human CD34 cells interact with zebrafish endothelial cells, we transplanted CD34 human cells into the Tg(kdrl:mCherry) reporter line labelling endothelial cells in red (Chi et al., 2008). Spinning disk confocal images focusing on the zebrafish CHT showed that at 2 hours post transplantation (hpt) human CD34 cells are still in the caudal vein (Figure 5A). At 9hpt, we observed co-localisation of human CD34 cells with endothelial cells, with some human CD34 cells already inside the perivascular pockets (Figure 5A). Once inside the perivascular pocket, we observed interactions with endothelial cells within the pocket, this process has already been termed ‘cuddling’ when imaging endogenous zebrafish HSCs: endothelial cells surround and embrace the incoming stem cell (Tamplin et al., 2015). Clear extensions, positive for the endothelial cell marker, were observed surrounding a human CD34 cell from within the perivascular pocket (Figure 5B, Supplementary Movie 1). Moreover, we imaged an instance where human CD34 cells divide within the zebrafish haematopoietic niche (Figure 5C, Supplementary Movie 2). These observations showed that these human CD34 cells have engrafted into the zebrafish CHT, therefore confirming that the zebrafish native CHT provides a human-compatible environment to allow human cells to engraft.\n\nSpinning disk confocal stills of timelapses zebrafish Caudal Haematopoietic Tissue (white lines) from the endothelial cell (red) reporter line Tg(kdrl:mCherry) zebrafish larvae transplanted with human CD34 cells (green). (A) At 2hpt, human CD34 cells are still in the vessels and empty perivascular pockets (white arrowheads) do not contain human CD34 cells. At 9hpt, human CD34 cells co-localised with endothelial cells and human CD34 cells appears inside the perivascular pockets (white arrows). DA: Dorsal aorta, CV: Caudal vein. Scale bar= 50μm. (B) High magnification of a human CD34 cell being ‘cuddled’ by surrounding endothelial cells, note the endothelial cell protrusion acting like arms (white arrows). Scale bar= 50μm. (C) Division of a human CD34 cells (white arrow) within a perivascular pocket, with hpt displayed. Note the equal distribution of fluorescence between the two daughter cells in the last frame. Scale bar=10μm.\n\n\nDiscussion\n\nOur protocol of human HSCs transplantation in zebrafish resulted in similar events observed in the numerous studies on zebrafish haematopoiesis. Zebrafish haematopoiesis has been well described and it is known that endogenous HSCs emerge from the ventral wall of aorta to join the circulation at around 36 hours post fertilisation (Gering & Patient, 2005; Kissa & Herbomel, 2010). HSCs roll along the endothelium of the caudal vein in the zebrafish CHT, where they exit the blood vessel to reach perivascular pockets of endothelial cells (Gering & Patient, 2005; Kissa & Herbomel, 2010; Tamplin et al., 2015). This haematopoietic niche protects the stem cell and allows it to divide and colonise other haematopoietic tissues. The well-described engraftment process of zebrafish HSCs in the CHT has provided us with key cell engraftment behaviour to look for.\n\nIndeed, we have observed that human CD34 cells home into the zebrafish CHT, where they engage with endothelial cells. By using high resolution confocal microscopy on transplanted zebrafish from the Tg(kdrl:mCherry) endothelial cells reporter line, we observed that human CD34 cells exit the caudal vein to reach perivascular pockets of endothelial cells within 9 hours after injection. Once in perivascular pockets, human CD34+ cells are ‘cuddled’ by endothelial cells and even divide, processes already described for zebrafish endogenous stem cells.\n\nFuture validation assays: Injections of 20 to 50 labelled human CD34+ cells in the circulation of zebrafish larvae is enough to observe interactions with endothelial cells in perivascular pockets and subsequent division within the zebrafish CHT. To validate the extent to which this interaction can represent engraftment, stem cell colonisation of the definitive haematopoietic organs and further cell differentiation must be studied. Stem cell colonisation of the definitive haematopoietic organs starts from day 5, the thymus is colonised by HSCs and later on the kidney bone marrow, both organs which will contribute to the development of the adaptive immune system (Kissa et al., 2008). Future validation assays looking at stem cell migration to the thymus and kidneys would be useful to confirm the extent of engraftment. Once engrafted, HSCs will produce lineage-committed progenitors that will give rise to blood cells, including immune cells. To further validate the model, populations of human blood cells present in adult zebrafish could be measured. It was shown that enriched human CD34 cells injected into zebrafish larvae resulted in the presence of myeloid lineage human cells only (Staal et al., 2016). Our protocol using pure CD34 cells may provide a better graft and differentiate into multiple lineages.\n\nLimitations of our current protocol: Currently our protocol detailed a method for purifying CD34 cells from whole blood, known to contain a small number of circulating HSCs. Indeed we have shown that blood samples used for cell preparation contained on average 0.03% of CD34 cells. Our current protocol allows the transplantation of 20–50 cells per fish from a pool of minimum 10×103 cells, allowing transplantation of maximum 100 fish. Although this would be sufficient to assess efficiency and viability of a single graft, it would be limiting to use as a high-throughput assay. This protocol could be scaled up to transplant thousands of zebrafish larvae to perform drug screens, or even to apply different HSC markers to study engraftment properties of human HSCs. However, a larger pool of HSCs will be required. Parada-Kusz et al described a high-throughput transplantation assay of murine CD34 cells from bone marrow in zebrafish (Parada-Kusz et al., 2017). Bone marrow from crushed femurs contains considerably more CD34 cells than whole blood, but this source cannot be use for human donors. To obtain a larger pool of HSCs, cord blood and enriched whole blood after cytokine G-CSF treatment would be suitable sources. These samples can be obtained through necessary ethical approvals and will provide enough graft material to scale up our protocol and perform high-throughput assays.\n\nTranslatability: Alongside offering a new model to continue research on stem cell transplantation, our zebrafish assay offers a scientific advantage that could revolutionise how stem cell graft are being tested before transplantation. Currently, most patients in need of transplantation receive a graft that has been cryopreserved and stored at suprafreezing temperatures (Stockschläder et al., 1997). Although these grafts are being tested for viability using the Trypan Blue staining, there is currently no assay quick enough to test for efficacy of the graft to engraft (Fleming & Hubel, 2006). Our data show the potential of human CD34+ cells to colonise the haematopoietic niche in zebrafish, engraft and proliferate within 9 hours after injection. Therefore, our zebrafish assay detailed in this study could provide a quick, cheap and efficient assay to test graft efficiency and even viability in less than 12h.\n\nTransferability: Our zebrafish system to study stem cell transplantation research will also advance the 3Rs components with a major impact on animal welfare. The replacement of humanised mice by humanised zebrafish larvae will represent a giant step for the 3Rs, allowing zebrafish embryos to be a host for human stem cells without any myeloablative procedures. The refinement of the harmful procedures without the need of myeloablation, by using zebrafish before the development of adaptive immunity, represents a powerful alternative to mice. Moreover, using a transparent organism before the onset of independent feeding to visualise stem cell engraftment will allow selection of engrafted animal within 12h. Zebrafish are widely used and for communities without zebrafish facilities, a small zebrafish system to host the few adults needed for this experiment is easy and cheap to start. These critical advantages of this assay using zebrafish as a model system will, we hope, increase the chances of wide uptake of this system.\n\n\nData availability\n\nDataset 1: FACS output files for Figure 2. DOI: 10.5256/f1000research.14507.d200844 (Hamilton et al., 2018a)\n\nDataset 2: Raw values file for Figure 3. DOI: 10.5256/f1000research.14507.d200845 (Hamilton et al., 2018b)\n\nDataset 3: Raw values file and image file for Figure 4. DOI: 10.5256/f1000research.14507.d200847 (Hamilton et al., 2018c)\n\nDataset 4: Raw image file for Figure 5. Images should be opened with Velocity software. DOI: 10.5256/f1000research.14507.d200848 (Hamilton et al., 2018d)", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis study was supported by a NC3Rs grant to SAR and IS (NC/M001490/1). SAR is supported by a MRC Programme Grant (MR/M004864/1) and NH is supported by a European Leukodystrophy Fellowship (ELA 2016-012F4). Imaging was carried out in the Wolfson Light Microscopy Facility, supported by an MRC grant (G0700091) and a Wellcome Trust grant (GR077544AIA).\n\n\nAcknowledgements\n\nWe thank the staff of the flow cytometry facility of the Infection, Immunity and Cardiovascular Disease department and the Bateson Centre Light Microscopy Facility and the aquarium staff at the University of Sheffield for their help.\n\nThis article has been completed according to the ARRIVE checklist (Supplementary File 2).\n\n\nSupplementary material\n\nSupplementary File 1: Full protocol for transplantation of human HSCs into zebrafish.\n\nClick here to access the data.\n\nSupplementary File 2: NC3Rs ARRIVE checklist.\n\nClick here to access the data.\n\nSupplementary Movie 1: Human stem cell being cuddled by zebrafish endothelial cells.\n\nClick here to access the data.\n\nSupplementary Movie 2: Human stem cell dividing within zebrafish perivascular pocket.\n\nClick here to access the data.\n\n\nReferences\n\nChi NC, Shaw RM, De Val S, et al.: Foxn4 directly regulates tbx2b expression and atrioventricular canal formation. Genes Dev. 2008; 22(6): 734–739. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nde Jong JL, Zon LI: Use of the zebrafish system to study primitive and definitive hematopoiesis. Annu Rev Genet. 2005; 39: 481–501. PubMed Abstract \| Publisher Full Text\n\nFleming KK, Hubel A: Cryopreservation of hematopoietic and non-hematopoietic stem cells. Transfus Apher Sci. 2006; 34(3): 309–315. PubMed Abstract \| Publisher Full Text\n\nGering M, Patient R: Hedgehog signaling is required for adult blood stem cell formation in zebrafish embryos. Dev Cell. 2005; 8(3): 389–400. PubMed Abstract \| Publisher Full Text\n\nHamilton N, Sabroe I, Renshaw SA: Dataset 1 in: A method for transplantation of human HSCs into zebrafish, to replace humanised murine transplantation models. F1000Research. 2018a. Data Source\n\nHamilton N, Sabroe I, Renshaw SA: Dataset 2 in: A method for transplantation of human HSCs into zebrafish, to replace humanised murine transplantation models. F1000Research. 2018b. Data Source\n\nHamilton N, Sabroe I, Renshaw SA: Dataset 3 in: A method for transplantation of human HSCs into zebrafish, to replace humanised murine transplantation models. F1000Research. 2018c. Data Source\n\nHamilton N, Sabroe I, Renshaw SA: Dataset 4 in: A method for transplantation of human HSCs into zebrafish, to replace humanised murine transplantation models. F1000Research. 2018d. Data Source\n\nKiel MJ, Yilmaz OH, Iwashita T, et al.: SLAM family receptors distinguish hematopoietic stem and progenitor cells and reveal endothelial niches for stem cells. Cell. 2005; 121(7): 1109–1121. PubMed Abstract \| Publisher Full Text\n\nKissa K, Herbomel P: Blood stem cells emerge from aortic endothelium by a novel type of cell transition. Nature. 2010; 464(7285): 112–115. PubMed Abstract \| Publisher Full Text\n\nKissa K, Murayama E, Zapata A, et al.: Live imaging of emerging hematopoietic stem cells and early thymus colonization. Blood. 2008; 111(3): 1147–1156. PubMed Abstract \| Publisher Full Text\n\nLangenau DM, Ferrando AA, Traver D, et al.: In vivo tracking of T cell development, ablation, and engraftment in transgenic zebrafish. Proc Natl Acad Sci U S A. 2004; 101(19): 7369–7374. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNombela-Arrieta C, Pivarnik G, Winkel B, et al.: Quantitative imaging of haematopoietic stem and progenitor cell localization and hypoxic status in the bone marrow microenvironment. Nat Cell Biol. 2013; 15(5): 533–543. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNüsslein-Volhard C, Dham R: Zebrafish: A practical approach. New York Oxford Univ. Press. 2002. Reference Source\n\nParada-Kusz MM, Clatworthy A, Hagedorn EJ, et al.: Generation of Mouse-Zebrafish Hematopoietic Tissue Chimeric Embryos for Hematopoiesis and Host-Pathogen Interaction Studies. bioRxiv. 2017. Publisher Full Text\n\nPrince LR, Prosseda SD, Higgins K, et al.: NR4A orphan nuclear receptor family members, NR4A2 and NR4A3, regulate neutrophil number and survival. Blood. 2017; 130(8): 1014–1025. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRenshaw SA, Trede NS: A model 450 million years in the making: zebrafish and vertebrate immunity. Dis Model Mech. 2012; 5(1): 38–47. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSnowden JA, Saccardi R, Allez M, et al.: Haematopoietic SCT in severe autoimmune diseases: updated guidelines of the European Group for Blood and Marrow Transplantation. Bone Marrow Transplant. 2012; 47(6): 770–790. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStaal FJ, Spaink HP, Fibbe WE: Visualizing Human Hematopoietic Stem Cell Trafficking In Vivo Using a Zebrafish Xenograft Model. Stem Cells Dev. 2016; 25(4): 360–365. PubMed Abstract \| Publisher Full Text\n\nStockschläder M, Hassan HT, Krog C, et al.: Long-term follow-up of leukaemia patients after related cryopreserved allogeneic bone marrow transplantation. Br J Haematol. 1997; 96(2): 382–6. PubMed Abstract \| Publisher Full Text\n\nSullivan KM, Muraro P, Tyndall A: Hematopoietic cell transplantation for autoimmune disease: updates from Europe and the United States. Biol Blood Marrow Transplant. 2010; 16(1 Suppl): S48–56. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSykes M, Nikolic B: Treatment of severe autoimmune disease by stem-cell transplantation. Nature. 2005; 435(7042): 620–627. PubMed Abstract \| Publisher Full Text\n\nTamplin OJ, Durand EM, Carr LA, et al.: Hematopoietic stem cell arrival triggers dynamic remodeling of the perivascular niche. Cell. 2015; 160(1–2): 241–252. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTanner A, Taylor SE, Decottignies W, et al.: Humanized Mice as a Model to Study Human Hematopoietic Stem Cell Transplantation. Stem Cells Dev. 2014; 23(1): 76–82. PubMed Abstract \| Publisher Full Text\n\nThomas ED, Lochte HL Jr, Lu WC, et al.: Intravenous infusion of bone marrow in patients receiving radiation and chemotherapy. N Engl J Med. 1957; 257(11): 491–496. PubMed Abstract \| Publisher Full Text\n\nTraver D, Paw BH, Poss KD, et al.: Transplantation and in vivo imaging of multilineage engraftment in zebrafish bloodless mutants. Nat Immunol. 2003; 4(12): 1238–1246. PubMed Abstract \| Publisher Full Text\n\nWhite RM, Sessa A, Burke C, et al.: Transparent adult zebrafish as a tool for in vivo transplantation analysis. Cell Stem Cell. 2008; 2(2): 183–189. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZaucha JM, Gooley T, Bensinger WI, et al.: CD34 cell dose in granulocyte colony-stimulating factor-mobilized peripheral blood mononuclear cell grafts affects engraftment kinetics and development of extensive chronic graft-versus-host disease after human leukocyte antigen-identical sibling transplantation. Blood. 2001; 98(12): 3221–3227. PubMed Abstract \| Publisher Full Text" }	[ { "id": "34036", "date": "04 Jun 2018", "name": "Owen J. Tamplin", "expertise": [ "Reviewer Expertise Hematopoietic stem cell niche" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis method article by Hamilton et al. describes in extensive detail the procedures involved in transplantation of human CD34+ hematopoietic stem/progenitor cells (HSPCs) directly into the circulation of 2 day-old zebrafish embryos. Although there have been studies that transplanted mouse1 and human2 HSPCs into zebrafish embryos as a model for homing and lodgement, as the authors discuss, this article provides the technical details necessary to successfully reproduce these experiments. The authors explain clearly how this approach can reduce the number of animal models needed in research (i.e., the 3Rs). Zebrafish embryos as transplant recipients can be collected in the hundreds each week, without the need to sacrifice adult animals. The efficiency of the approach is shown by the small number of CD34+ HSPCs that are needed per recipient, about 50 per embryo, compared to about 10,000 per recipient mouse. The authors perform time-lapse live imaging that reveals transplanted human HSPCs are “cuddled” by zebrafish endothelial cells in the niche, a dynamic cellular behavior that is observed during endogenous HSPC lodgement3. Overall, the data presented in this method article is of high quality, is sufficient to reproduce the technique, and provides a strong rationale for the 3Rs. However, as other similar studies have been performed2, it would be helpful if the authors could further extend their results to highlight improvements to the method.\nSome suggestions for revisions are listed below:\nWhat are the different options for dyes that could be used to label donor cells? Has toxicity been assessed in dosage curves? Have alternatives been explored? How does PKH262compare to fluorescein (this study)? How long do the donor cells survive in the transplant recipient? Staal et al. track the cells until 6 dpf, so 4 days after transplant at 2 dpf. How long were recipients followed in these experiments? How does temperature affect survival of donor cells, given the different optimal temperatures between human and zebrafish (i.e., 37C vs 28C). From the existing data collected for the study, what percentage of lodged HSPCs are found in pockets? Have other injection methods been tried? RO injection in embryos is also possible, and may prove easier and more efficient than injection into the Duct of Cuvier. How were transplants controlled? Is there a way to distinguish between lodged viable cells vs stuck debris? If sorted adult zebrafish HSPCs were transplanted in similar numbers (e.g., kidney cd41:gfp low cells), would the lodged number of cells be comparable to human CD34+ cells? Staal et al. tested chemokine responses ex vivo—could the authors inject human cytokines into zebrafish recipient embryos to test their effects in vivo? For example, would the transplant results change in the presence of human G-CSF?\n\nPossible additional “Discussion” points:\nHow could the method be scaled up for higher throughput studies (e.g., automated injection)? What are the limitations of the system compared with mice? How could the non-isogenic background of zebrafish impact a study, compared to using isogenic mouse recipients? Is ‘engraftment’ appropriate terminology, given that the cells can only be tracked short-term? Would ‘lodgement’ be a better description of the processed that is modeled? Replacing the adult mouse with the zebrafish embryo as a transplant recipient for human HSPCs switches the system from an adult bone marrow niche to a developmental “fetal-like” niche (i.e., CHT). How should this be considered when interpreting the results?\n\nMinor points related to the manuscript:\nCould the authors provide higher resolution images in figure 4A-B? The absolute numbers should also be shown in Figure 4C. What is the number inside and outside of the CHT at 1 hpt? Does 100% mean there are no circulating cells in the embryo at 1 hpt? If there are 50% fewer cells in the CHT 12 hours later, where are the cells going? Are they migrating to a different tissue or dying? How long is the image in Figure 4A taken after injection? Text says 100% after injection are in the caudal vein, but the image shows cells around the embryo. Are the images in Figure 5 maximum projections or single slices? Single Z planes more clearly show the endothelial projections. In Figure 5B, it should be clear that the second frame is only endothelial cells. Single channels could be shown together with the merge. This text on page 3 is not clear: “These immunodepleted mice must be grown to adulthood in order to assess engraftment success, usually performed shortly after birth.” Does this refer to a neonatal transplant? How were the numbers collected for this statement on page 3: “In 2015, 25 publications used immunocompromised mice for HSC transplant studies, representing around 1000 mice each year worldwide – all undergoing severe procedures over a long period of time.” Is there a reference that could be cited?\nAre a suitable application and appropriate end-users identified? Yes\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\nAre the 3Rs implications of the work described accurately? Yes\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4317", "date": "23 Dec 2018", "name": "Noemie Hamilton", "role": "Author Response", "response": "We thank Professor Owen Tamplin for his positive review and encouraging comments for revisions. We have uploaded a revised manuscripts and updated figures addressing all of his minor comments. We have responded to his suggestions and additional discussion points below:Some suggestions for revisions are listed below: We have not tested different dyes for this study and we have stuck to using a green dye to go alongside the red Tg(kdrl:mcherry) line. What happens to the model after 5dpf is a crucial experiment to validate the assay and ensure the human cells have truly engrafted. This is already part of the discussion, mentioning other studies looking at the different lineages emerging from the engrafted human HSCs. However, our current experiments end at 5dpf and raising chimera animals past this age would require an amendment of our project licence which is not within the scope of this study. A very recent study (Cabezas-Sainz et al., 2018) has explored the effect of raising the incubation temperature of zebrafish to accommodate for human cells and we have added to the ‘Limitations’ section of our discussion. We have not tried other routes of injection and we are grateful for the RO injection suggestion which will be tested. We have not transplanted zebrafish CD41 cells and we would expect them to engraft better than human cells. This is a great suggestion and this is how we hope this model will be used in the future: To find new molecules that could improve engraftment success. Possible additional “Discussion” points: We have added the automated injection to our ‘Limitations of our current protocol’ section to go alongside sourcing CD34 from different sources to scale up the pt The limitations of using a non-isogenic recipient and an ‘foetal-like niche’ are very valid points. We hope that our zebrafish embryonic model will be used instead of mice to optimise protocols initially. We fully expect that mice will still be used to corroborate results from zebrafish studies before being used in patients. We agree that ‘lodge’ would be a more suitable description for our model until we show long term engraftment of human cells. This was replaced in the result section. Replacing the adult mouse with the zebrafish embryo as a transplant recipient for human HSPCs switches the system from an adult bone marrow niche to a developmental “fetal-like” niche (i.e., CHT). How should this be considered when interpreting the results? Minor points related to the manuscript: 1+2+3: Figure 4: we apologise for the confusion in the quantification from Figure 4C. We have changed the text and figure legend to be more consistent and emphasise that the absolute number of cells was the number of cells lodged within the CHT at 1hpt. We subsequently quantified the number of cells at 5hpt, 9hpt and 13hpt and plotted them in percentage. We do not find GFP positive cells in any other tissues after transplantation, so we can conclude that the cells initially visualised in the CHT could be debris or healthy cells that have died and be removed by innate immune cells. Figure 5b and 5C are all single plane, we have specified this in the figure legend We have modified the text on page 3 to make sure readers do not think we are referring to neonatal transplant. ‘These immunodepleted mice must be grown to adulthood in order to assess engraftment success. They live their entire lives undergoing severe procedures with high maintenance requirements, since they need to be homed in sterile rooms and fed sterile food to avoid fatal infection due to defective immunity.’ Publications using immune compromised mice were counted using a PubMed search, reading articles and averaging how many mice were used per study. We have added this to the text to make it clearer." } ] }, { "id": "34964", "date": "19 Jun 2018", "name": "Siobhan Crilly", "expertise": [ "Reviewer Expertise Referee suggested by the NC3Rs for their scientific expertise and experience in assessing 3Rs impact." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe methods article from Hamilton et al provides a strong rationale for an improved and accessible model for HSC engraftment in a larval zebrafish. Humanised mouse models require severe procedure methodology and limitations in visualising the engraftment process essential for understanding human mechanisms and zebrafish offer an alternative, complementary model in which to do so.\n\nThe methodology provided is detailed enough to be translated and repeated accurately. As the authors highlight, this model has potential for be up-scaled for high-throughput methodology for drug screening and offers a necessary complementary model to reduce the number of rodents needed for pre-clinical study. Further investigation in this model would prove interesting to see how engrafted HSCs develop into various lineages and employing the porous nature of the zebrafish embryo to investigate drug treatment to encourage engraftment/proliferation/lineage development.\n\nPoints to consider for revision:\nWhen isolating the CD34+ cells using both magnetic beads and antibodies, do the authors remove these markers before transplantation? Have the authors investigated an innate immune cellular response to the engraftment of human cells? How do cells react to the changes in temperature (37 -28 degrees) for zebrafish incubation? Have the authors experimented using different incubation temperatures for fish comparable to some tumour lines? How does the engraftment impact the development of native zebrafish HSCs (from 30hpf) and production of haematopoietic markers? How does the changing productivity of the CHT tissue (in the developing organism) affect the longevity of the engraftment? What happens to the model after the development of the mature adaptive immune system in the zebrafish after 2 weeks – is this the end of the experiment, and engraftment and lineage cannot be investigated further than this? How many of the cells that ‘lodged’ into the tissue went on to proliferate at later timepoints 2-13hpt? Please ensure full details of statistical analyses that were performed are provided in the methods and also that the results of these analyses are described in the main text. Also the figure legends should include details of the statistical comparisons made with any significant results being indicated on the relevant figures.\n\nMinor corrections in the manuscript:\n‘Transplanted’ should be ‘transplantation’ on page 4 left column line 5 Can the authors label figure 4 as done in figure 5, the dorsal aorta, caudal vein and stem cell tissue. In the ‘future validation assay’ section the authors say kidney bone marrow instead of kidney marrow What time point are the images in 5C taken?\n\nAre a suitable application and appropriate end-users identified? Yes\nIf applicable, is the statistical analysis and its interpretation appropriate? Partly\nAre the 3Rs implications of the work described accurately? Yes\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4318", "date": "23 Dec 2018", "name": "Noemie Hamilton", "role": "Author Response", "response": "We thank Prof Allan and Miss Crilly for their positive comments. We have uploaded an improved version of the manuscript including all the minor corrections they suggested. Their recommended revisions have also been added to the manuscript and detailed below points by points: We have added this to the methods. The antibody and the beads were not been removed, however we have been advised by the company selling the CD34 MicroBead kit that the beads naturally detach within 24 hours. We thank you for this suggestion and we hope that the innate immune cellular response to the engraftment of transplanted cells can be part of another study. We have added your suggestion as part of the ‘future validation assays’ of our discussion. A very recent study (Cabezas-Sainz et al., 2018) has explored the effect of raising the incubation temperature of zebrafish to accommodate for human cells and we have added to the ‘Limitations’ section of our discussion. We have not investigated whether the engraftment of new cells had an impact on the development of endogenous zebrafish HSCs, nor the integrity of the CHT (point 5). However, we have added this to our discussion as an important future validation assay. Included in point 4 above. What happens to the model after the development of the mature adaptive immune system in the zebrafish after 2 weeks has already been mentioned in the validation assay part of the discussion. We agree that this is a crucial experiment to perform to investigate the presence of different lineages emerging from the engrafted human HSCs. However, our current experiments end at 5dpf and raising chimera animals past this age is not within the scope of this study. We have added the full details of statistical analyses to our figure legends, results sections and have uploaded updated figures." } ] } ]	1	https://f1000research.com/articles/7-594
https://f1000research.com/articles/7-1962/v1	20 Dec 18	{ "type": "Research Article", "title": "Human activities and persistent coral reef degradation in Gaspar Strait, Bangka Belitung Islands, Indonesia", "authors": [ "Singgih Afifa Putra", "Indra Ambalika Syari", "Helmy Akbar", "Iwan Suyatna", "Davin H. E. Setiamarga", "Indra Ambalika Syari", "Helmy Akbar", "Iwan Suyatna", "Davin H. E. Setiamarga" ], "abstract": "Background: The aim of the study was to describe the coral reef condition in Bangka Belitung Islands, particularly from Gaspar Strait. This research location is well known for its underwater archaeological discovery and shipwreck sites. Recent increases in mining, fishing and tourism activities in the surrounding islands might have affected the condition of the coral reef. Methods: Nine islands inside the strait were visited (i.e. Langer, Kembung, Piling, Aur, Salma, Pongok, Celagen, Kelapan, and Lepar Island), and a line transect was used to observed coral reef conditions. Results: Coral cover was found to be predominantly in fair conditions (25-50%). Coral mortality index also tended to be high, which indicated that the coral reef ecosystem was in threatened conditions. Previous and recent reports also reported the same condition as found by this study. Conclusion: Degradation of the coral community in Bangka Belitung Islands is likely caused by human activities. This suggests that increasing human activities significantly affects the coral reef condition. Protection of coral reefs with sustainable management for mining activity, tourism and fishing practices are needed.", "keywords": [ "Coral Reef", "Degradation", "Human Activity", "Bangka Belitung", "Indonesia" ], "content": "Introduction\n\nThe Research Center for Oceanography (RCO)-LIPI recently reported that Indonesia’s coral reef condition in 2017 was predominantly in fair and poor conditions1. Coral reefs are very vulnerable to damage, particularly from human activity2,3, and their degradation is caused by many reasons among them sedimentation, pollutions from industrial domestic waste, coral mining, over exploitation, and unsustainable fishing activity1,2,4. Human activity (e.g. destructive fishing, uncontrolled tourism) can cause dramatic damage in a short period of time to the coral reef ecosystem. Mechanical damage from destructive fishing practices (e.g. anchoring damage and blast-fishing) could dramatically reduce coral cover, and also reduce coral reef resilience to natural perturbations4,5. The higher the decrease in live coral cover in an area, the larger the decrease in coral species diversity4. The development of uncontrolled tourism in some areas has already caused substantial damage near reef areas6–9. Sedimentation from mining activity has also impacted coral reefs, and sedimentation is a major controlling factor in reef development10. The coral community in high sedimentation sites are relatively similar over a period of time, i.e. stable concerning cover and diversity11.\n\nThe province of Bangka Belitung Islands is one region of Indonesia that has developed high tourism activity. Tourists visiting the Bangka Belitung Islands are increasing12, and has been declared as one tourist area that should be a priority for development in Indonesia. There have been some previous studies on the coral reefs in Bangka Belitung Islands13–15, which have revealed that this coral reef is at threat of a high rate of sedimentation. The sedimentation come from tin mining activities in surrounding islands15, which have existed since 185016. The present study describes the coral reef condition in the Gaspar Strait between Belitung and Bangka Island, which is also suffering from the impact of anthropogenic activities. Information on the coral reef condition and distribution in the Gaspar Strait is important for sustainable use of marine resources in this region.\n\n\nMethods\n\nThe Gaspar Strait is the strait separating Belitung Island and Bangka Island and connecting the Natuna Sea (Karimata Strait) to the Java Sea. The sampling sites were located at reef formations in nine islands inside the Gaspar Strait, Bangka Belitung Islands. Those islands are Langer Island (2°48'18.00\" S and 107°22'14.00\" E), Kembung Island (2°51'36.27\" S and 107°20'22.01\" E), Piling Island (2°55'17.44\" S and 107°21'12.00\" E), Aur Island (02°59'20,7\" S and 107°13'50,1\" E), Salma Island (02°59'21,6\" S and 107°06'20,8\" E), Pongok Island (2°52'30.12\" S and 107° 3'58.51\" E), Celagen Island (2°52'23.36\" S and 107°0'53.36\" E), Kelapan Island (02°51'065\" S and 106°49'899\" E), Lepar Island (02°53'853\" S and 106°47'475\" E) (see Figure 1). The fieldwork was conducted between September and October 2014. This research was approved by Marine and Fisheries Agency (Dinas Kelautan dan Perikanan) of Bangka Belitung Province.\n\n(Source: Google Earth 2018).\n\nData was collected using Line Intercept Transect17,18. 3000 centimetres of transect line was placed parallel to the coastline in each site at 5–7 m depth. Coral reefs are identified based on lifeform category19. Coral lifeform was counted on every centimetre in line where the lifeform found. Percent cover of living corals (hard corals and soft corals) found along the transect are identified and used to define if the coral reef condition is poor (0–25% coral cover), fair (26–50%), good (51–75%), and excellent (76–100%)1,19,20. A coral mortality index was also calculated, which is a simple ratio of dead coral cover to the sum of dead and hard coral cover. Theoretically, this index would scale the life coral cover values to the amount of space corals could occupy20. Statistical values represent means ±SD, with probabilities calculated by one-way ANOVA using Data Analysis function of Microsoft Excel for MAC Version 16.16.2 (180910).\n\n\nResults and discussion\n\nPercentage live coral cover was calculated from identification of hard and soft corals. Hard corals were presented as a total and broken down to 11 lifeforms, (i.e. acropora branching, acropora digitate, acropora encrusting, acropora sub-massive, acropora tabulate, coral branching, coral encrusting, coral foliose, coral massive, coral mushroom, coral sub-massive) (see Dataset 1). An ANOVA revealed that the lifeforms coral cover between sampling sites were not varied significantly (Supplementary Table 1). On average, coral cover in Gaspar Strait was found in fair conditions (33.1±17.2%), with the lowest coral cover found in Salma Island in poor condition (<25%) and the highest coral cover found in Piling and Lepar Island in good condition (50–75%). Coral cover in the other islands was found in fair and poor conditions (see Table 1).\n\nIn total, 89 species of hard corals have been found in Bangka Island, with the most dominant species from the Poritiid and Faviid group15. The highest coral cover found in Piling and Lepar Island in the present study was formed from aggregate coverage of Acroporiid coral, i.e. tabulate and digitate (Figure 2). Non Acroporiid coral found in these two islands commonly from foliose and massive corals. Foliose and massive corals category generally represent Agariciid, Pectiniid, Montiporiid, Poritiid corals, which also commonly found in Bangka Island.\n\nOthers represent soft corals, sponges and other animals found. Algae represent algal assemblage, macroalgae, and turf algae. Abiotic represent silt, rubble, rock, and sand substrate.\n\nDead coral cover was calculated as recently dead coral and dead coral with algae20. On average, dead coral cover in Gaspar Strait found about 33.0±22.0% (see Table 1). The highest dead coral cover was found in Kembung Island (70.8%), while Aur Island has the lowest dead coral cover (3.9%). The higher the coral’s mortality index value, the worse the condition of the coral reefs (see Figure 3). The coral mortality index was used to describe coral degradation in the research area. The high value of mortality index indicated that the coral reef ecosystem in the Gaspar Strait is in threatened conditions.\n\nGraphic shows high value of coefficient of determination (R squared) between dead coral cover and coral mortality index value.\n\nThe most recent status review of Indonesia’s coral reef was published by Research Center for Oceanography (RCO)-LIPI. Coral reef condition in 2016 from Bangka Belitung Islands was predominantly found in fair conditions1. In this research, we found that the coral reef condition in Gaspar Strait was also predominantly in fair conditions, which is two years on from the RCO-LIPI report. Previously, a report from reef formation in Bangka Island (in 2010) also reported fair condition15. This is in contrast with the condition in an earlier report from 2005 and 200813,21, where the reported coral condition from western and eastern parts of the Gaspar Strait was good condition (see Table 2). This showed us that there have been no significant changes between 2010, 2014 and 2016 of coral cover in Gaspar Strait after coral degradation in 2005–2010.\n\nThe causes of coral reef degradation in Indonesia has been explained in various papers and reports1,4,22,23. Since the 1980’s, it has been reported that coral reefs in Indonesia have been severely damaged from sediment and organic pollution as well as excessive exploitation of fishery stocks, with destructive fishing practices24,25. Tin mining activity in the marine waters of Bangka Belitung is legally permitted by the regional government (Supplementary Figure 1), and the coral reef in Bangka Belitung Islands has been allegedly exposed to sedimentation from mining activity15. In the Gaspar Strait, in the present study we found silt substrate under the line transect from four sampling sites. At the ecosystem level, a reef zone with heavy sedimentation will cause lower species diversity with some species absent, greater abundance of forms and species with high resistance to sediment, and lower growth rate10.\n\nBangka Belitung Islands is one of province in Indonesia that encouraged tourism for regional income12. Tourism activities should have positive impacts on the coral reef in terms of support and conservation awareness. But, some reports showed direct negative impacts for the reef7,9. Since it was established as one of ten priority tourism destination in Indonesia, tourism development in Bangka Belitung Islands has grown rapidly, seen from growth of tourists which reaches up to 20.5% each year, and also an increase in the number of hotels and restaurants in the tourism area12.\n\nSimilarly, destructive fishing practices are also one major threat to coral reef degradation4. We found fishing practices such as trap fishing with bottom fish pots –made from iron wire – are being placed in the reef flat with the help of corals for anchoring. Fisherman also are still using potassium nitrate to stun fish around Langer Island. Destructive fishing practices are one of major threats for coral development, which in particular may reduce resilience to natural perturbations, leading to assembly of small, sparse corals and reduced patchiness4,5.\n\nNatural perturbation such as predation may also cause coral reef degradation in Gaspar Strait. Acanthaster planci (crown-of-thorns starfish, COTS) is recognized as a major cause of coral reef degradation throughout much of the Western Pacific26. We found at least two individual COTS around the transect line during fieldwork, in Pongok and Salma Island (see Figure 4). COTS outbreaks in Indonesia have been reported since the early 1980s27. During outbreaks, 90% of coral can be killed in a large area28. Another factor that may cause degradation is coral disease29–32. Recently, Nirwanda et al.32 reported that there are three coral diseases found in Bangka Island; brown-band disease, dark spot disease, and skeletal eroding band are commonly found and generally infect massive coral lifeforms. Improvements in water quality, by minimizing sediments and nutrient pollution, and reduction in fishery exploitation will have definite benefits for the resilience of coral reef ecosystems25.\n\n(a) Coral communities in Piling Island; (b) Acanthaster planci (crown-of-thorns starfish) found in Salma Island (arrow).\n\n\nConclusion\n\nThe present study found that the coral reef around Bangka Belitung Islands was persistently found to be degraded in fair condition and the mortality index was predominantly high. This indicates that the coral reef ecosystem in this research location is under threatened conditions.\n\nThis study identified threats from human activity, such as sedimentation from tin mining activity, tourism development and destructive fishing practices. We also summarized threats to coral degradation from natural perturbations such as predation and disease. Interaction between human activity and other factors (e.g. sedimentation, tourism, predation, and disease) may be critical; thus protective management of island resources in Bangka Belitung Islands, particularly in Gaspar Strait, must be implemented to maintain a healthy coral reef ecosystem in the region. Protection of coral reefs with sustainable management for tourism and fishing practice, and also limitation in the area for tin mining, will increase the resiliency and development of the coral reef community.\n\n\nData availability\n\nF1000Research: Dataset 1. Percent cover of benthic components were found in nine reefs of Gaspar Strait. Value was calculated from 3000 cm line transect. https://doi.org/10.5256/f1000research.16519.d22984633", "appendix": "Grant information\n\nThis research was funded by the Regional Government of Bangka Belitung Islands Province through the Marine Coastal and Small Islands Zoning Plan Program.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe thank to Dinas Kelautan dan Perikanan of Bangka Belitung Islands Province for fieldwork permit, and the fisheries students of University of Bangka Belitung for supports during fieldwork.\n\n\nSupplementary material\n\nSupplementary Table 1: ANOVA (Single Factor) Analysis result for variance of benthic cover between nine reefs from Gaspar Strait.\n\nClick here to access the data.\n\nSupplementary Figure 1: Map of legal permitted tin mining in the surrounding marine waters of Bangka Belitung Islands. (Source: WALHI 2018; Retrieved from https://walhi.or.id/walhi-dosen-ubb-indra-ambalika-diskreditkan-nelayan/).\n\nClick here to access the data.\n\n\nReferences\n\nGiyanto, Abrar M, Hadi TA, et al.: Status Terumbu Karang Indoenesia 2017. Jakarta: Puslit Oseanografi – LIPI; 2017. Reference Source\n\nCesar HSJ: Coral Reefs: Their Functions, Threats and Economic Value. In: Collected Essays on the Economics of Coral Reefs. Cesar HSJ. (ed). Sweden: Kalmar University; 2000. Reference Source\n\nCinner JE, Huchery C, Darling ES, et al.: Evaluating social and ecological vulnerability of coral reef fisheries to climate change. PLoS One. 2013; 8(9): e74321. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEdinger EN, Jompa J, Limmon GV, et al.: Reef Degradation and Coral Biodiversity in Indonesia: Effects of Land-based Pollution, Destructive Fishing Practices and Changes Over Time. Mar Pollut Bull. 1998; 36(8): 617–630. Publisher Full Text\n\nMcManus JW, Reyes RB, Nañola CL: Effects of Some Destructive Fishing Methods on Coral Cover and Potential Rates of Recovery. Environ Manage. 1997; 21(1): 69–78. PubMed Abstract \| Publisher Full Text\n\nHawkins JP, Roberts CM: The growth of coastal tourism in the Red Sea: present and future effects on coral reefs. Ambio. 1994; 23(8): 503–508. Reference Source\n\nAllison WR: Snorkeler damage to reef corals in the Maldive Islands. Coral Reefs. 1996; 15(4): 215–218. Publisher Full Text\n\nHawkins JP, Roberts CM, Kooistra D, et al.: Sustainability of Scuba Diving Tourism on Coral Reefs of Saba. Coast Manage. 2005; 33(4): 373–387. Publisher Full Text\n\nDiedrich A: The impacts of tourism on coral reef conservation awareness and support in coastal communities in Belize. Coral Reefs. 2007; 26: 985. Publisher Full Text\n\nRogers CS: Responses of coral reefs and reef organisms to sedimentation. Mar Ecol Prog Ser. 1990; 62(1): 185–202. Publisher Full Text\n\nMcClanahan TR, Obura D: Sedimentation effects on shallow coral communities in Kenya. J Exp Mar Bio Ecol. 1997; 209(1–2): 103–122. Publisher Full Text\n\nValeriani D, Dudetyo D, Robiani D, et al.: Tourism and Economic Growth of Bangka Belitung Islands Province, Indonesia. IOSR Journal of Economics and Finance. 2017; 8(4): 54–59. Reference Source\n\nSjafrie NDM: Kondisi terumbu karang dan biota lainnya di Perairan Kecamatan Selat Nasik Kabupaten Belitung Tahun 2007-2008. Jurnal Perikanan. 2009; XI(2): 150–156. Reference Source\n\nAdibrata S: Evaluasi Kondisi Terumbu Karang di Pulau Ketawai Kabupaten Bangka Tengah. Jurnal Kelautan. 2013; 6(1): 19–28. Reference Source\n\nSiringoringo RM, Hadi TA: Kondisi dan Distribusi Karang Batu (Scleractinia Corals) di Perairan Bangka. Jurnal Ilmu dan Teknologi Kelautan Tropis. 2013; 5(2): 273–285. Reference Source\n\nKaur A, Diehl F: Tin miners and tin mining in Indonesia, 1850–1950. Asian Stud Rev. 1996; 20(2): 95–120. Publisher Full Text\n\nHill J, Wilkinson C: Methods for ecological monitoring of coral reefs. Townsville: Australian Institute of Marine Science. 2004; 117. Reference Source\n\nBeenaerts N, Berghe EV: Comparative Study of Three Transect Methods to Assess Coral Cover, Richness and Diversity. Western Indian Ocean J Mar Sci. 2005; 4(1): 29–37. Reference Source\n\nEnglish S, Wilkinson CR, Baker VJ: Survey Manual for Tropical Marine Resources. Townsville: Australian Institute of Marine Science. 1994; 368.\n\nGomez ED, Alino PM, Yap HT, et al.: A review of the status of Philippine reefs. Marine Pollution Bulletin. 1994; 29(1–3): 62–68. Publisher Full Text\n\nSiringoringo RM, Giyanto, Budiyanto A, et al.: Komposisi Jenis dan Persentase Tutupan Karang Batu di Perairan Lepar-Pongok, Bangka Selatan. Oseanologi dan Limnologi di Indonesia. 2006; 41: 71–84. Reference Source\n\nBrown BE, Suharsono: Damage and recovery of coral reefs affected by El Niño related seawater warming in the Thousand Islands, Indonesia. Coral Reefs. 1990; 8(4): 163–170. Publisher Full Text\n\nWilkinson CR: Status of coral reefs of the world: 2004. Australian Institute of Marine Science (AIMS). 2004. Reference Source\n\nBrown BE: Human induced damage to coral reefs. UNESCO Reports in Marine Science No. 40. 1986; 179. Reference Source\n\nWilkinson CR, Chou LM, Gomez E, et al.: Status of coral reefs in Southeast Asia: threats and responses. RN. 1993; J33–J39.\n\nBaird AH, Pratchett MS, Hoey AS, et al.: Acanthaster planci is a major cause of coral mortality in Indonesia. Coral Reefs. 2013; 32: 803–812. Publisher Full Text\n\nAziz A: Beberapa catatan tentang kehadiran bintang laut jenis Acanthaster planci di perairan Indonesia. J Oseana. 1995; 20(2): 23–31. Reference Source\n\nChesher RH: Destruction of Pacific Corals by the Sea Star Acanthaster planci. Science. 1969; 165(3890): 280–283. PubMed Abstract \| Publisher Full Text\n\nHaapkylä J, Seymour AS, Trebilco J, et al.: Coral disease prevalence and coral health in the Wakatobi Marine Park, south-east Sulawesi, Indonesia. J Mar Biol Ass UK. 2007; 87: 403–414. Publisher Full Text\n\nPutra SA: Kolonisasi komunitas karang di Cagar Alam Laut Krakatau, dan Implikasi Pengelolaanya. [Thesis] Bogor: Institut Pertanian Bogor. 2014. Publisher Full Text\n\nPutra SA, Damar A, Samosir AM: Colonization of Coral Communities in the Krakatau Islands Strict Marine Nature Reserve, Indonesia. Indonesian Journal of Marine Science. 2014; 19(2): 63–74. Publisher Full Text\n\nNirwanda S, Adi W, Syari IA: Inventarisasi Penyakit Karang di Perairan Turun Aban Kabupaten Bangka. Akuatik Jurnal Sumberdaya Perairan. 2017; 11(1): 18–25. Reference Source\n\nPutra SA, Syari IA, Akbar H, et al.: Dataset 1 in: Human activities and persistent coral reef degradation in Gaspar Strait, Bangka Belitung Islands, Indonesia. F1000Research. 2018. https://doi.org/10.5256/f1000research.16519.d229846" }	[ { "id": "42247", "date": "02 Jan 2019", "name": "Bert W. Hoeksema", "expertise": [ "Reviewer Expertise Coral reef research", "marine biodiversity" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe paper on Gaspar Strait coral reef degradation indicates possible threats consisting of tin mining, tourism activities and destructive fisheries but it is not clear how the survey results correspond with the conclusions. I have some comments that may help the authors to clarify their findings:\nIf tin mining has existed in the research area since 1850, how can the effects of destructive fisheries and tourism be recognized as important threats? We may assume that the coral reefs here have been threatened for many years.\n\nThe data were obtained per location from only one 30-m line transect at 5-7 m depth. This is very little and does not tell much about the condition of the reefs over their whole depth profile. Perhaps the authors can explain how the transects could be considered representative for the localities?\n\nA total of 89 species were recorded but there is no information about these species. It is unclear what the meaning of \"Poritiid group\" and \"Faviid group\" is, since these are not proper taxonomic units and they do not correspond with life forms. The same counts for \"Agariciid, Pectiniid, Montiporiid, Poritiid corals\", which does not explain whether this concerns genera or families.\n\nHow was it determined that changes between 2010, 2014 and 2016 were not significant after 2005-2010? This is perhaps possible if the data were taken in the same way from the exact same spots but the differences are only based on overall conclusions.\n\nSediments can be a natural component of reefs. So, if sediment was found in a transect, this does not necessarily imply that there is a threat.\n\nHow should tourism activities have positive impact on the reefs? Tourism does not automatically result in awareness and conservation.\n\nThe authors found \"at least two individual COTS\" around the transects. Why is it not clear how many individuals were counted? Such a low number does not seem to be threatening.\n\nThe authors mention the occurrence of coral diseases reported from a previous survey in Bangka but do not mention any disease found in their own transects. How does this relate to the present results?\n\nConclusion - what kind of management do the authors propose?\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [] }, { "id": "43390", "date": "07 Feb 2019", "name": "Michael John Risk", "expertise": [ "Reviewer Expertise Coral reef ecology/Indonesian reefs" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nReview of Putra et al.:\n\nI am normally delighted to see research coming out of Indonesia, particularly coral research produced by Indonesian scientists. This present paper, however, falls short in a number of areas.\n\nFirst of all, the title is very misleading. Authors have not produced research on \"human activities\", they have reported the results of a limited field study. They mention a variety of probable local stresses, some of which have undoubtedly affected these reefs, but nowhere are they able to connect stress with impact. This is quite disappointing, because there is an enormous amount of work that could have led them in the right direction. They mention the possible impact of sediments, but do not attempt to link their data to this putative stress. There have been documented changes in the nature of coral communities in response to the sediments, such as the decrease in Acropora and increase in massives, but this is not investigated. The authors mention the impacts of tin mining, and this would have been interesting to investigate by, for example, trace element analysis of selected corals – but the authors do not even mention this possibility. I could go on, but I will stop with this observation: the authors can say nothing about possible human impacts. What is more, they seem unfamiliar with research that might have allowed them to tease out some of these stresses. I recognize that some techniques, such as isotopic analysis of the coral tissue, might have been beyond their budget limits (although I note that nitrogen ratios can now be had for 20 bucks a pop), but they should at least be aware of the availability of techniques.\n\nSecondly, I have never liked this arbitrary definition of coral reefs into the good, the bad, and the ugly depending on observed percentage of coral cover. This is a particularly bad way in which to track changes over time. I realize this approach was born out of necessity in the Philippines decades ago, but surely we can get past this. A reef that has recently declined from 75% coral to 50% coral is on the way to extinction, whereas one that has happily maintained 30% coral cover throughout the Holocene is quite healthy.\n\nThe authors mention Acanthaster as a possible source of reef degradation, and I want to see this entire section removed. It is true that some years ago some Australian COTS experts visited Indonesia and produced a paper that said, lo and behold, COTS are a serious problem. I encourage the authors to read the magnificent Tomascik et al. (20001) book on the reefs of Indonesia, both volumes, cover to cover. Then read all of the papers produced by the McMaster coral reef efforts, by the likes of Jompa, Limmon, Edinger and Risk. Search through this voluminous literature, almost none of which is cited in the present manuscript, which represents several person-years' coral reef surveys underwater. Tabulate the number of times Acanthaster is mentioned. I can inform the authors that the answer is zero. It is quite clear that COTS represents no danger to the reefs of Indonesia, especially as compared with the myriad other stresses.\n\nThe fieldwork and the reporting thereof seems to have problems. If I am reading the paper correctly, there is only 30m of transact at each location. This is insufficient for an undergraduate student report, let alone a paper in the primary literature. The authors state that measurements were to the nearest centimeter, and all of us who have done underwater surveys recognize that this is an optimistic overestimate of precision. To then report percent cover to two decimal places betrays a fundamental lack of understanding of number theory. The authors are only legitimately allowed to report to the nearest percent.\n\nThe literature coverage seems very scattershot. There is copious citation of British research of little relevance, and some glaring holes. The omission of any citation of Tomascik’s work is hard to forgive.\n\nIs the work clearly and accurately presented and does it cite the current literature? No\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [] }, { "id": "47202", "date": "08 May 2019", "name": "James Davis Reimer", "expertise": [ "Reviewer Expertise Marine biodiversity", "ecology", "taxonomy." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe idea behind the work as presented is valid, and there is a clear need for more field surveys on many reefs, not only in Indonesia, but around the world, and thus this work is a welcome addition to the literature. However, I feel there are many shortcomings that need to be addressed before any indexing. I have refrained from detailed comments on grammar etc. as any revision will likely greatly change this paper. For now, I list only my major concerns:\nThe M&M need many more details. If I read this correctly, only one transect per site? But yet you discuss SD (although it is not shown in the relevant Figure 2 or Table 1). The field methodologies need many more details in order to be reproducible, including a list of \"lifeforem\" categories - this is listed later in Results incorrectly, images if available, etc. How did you choose sites? Also, why did you merge living coral to include both hard (Scleractinian?) and soft (octocorals?)? I would also like to see a definition of what you clearly mean by hard and soft corals, as there are numerous definitions. Define each category taxonomically. Finally, did you take images while doing LIT? Your work is a single point in time, but after you compare with past work. Detailed information on how comparisons were made is needed.\n\nI am not sure I agree with the lumping of categories' health statuses simply by live coral cover. Could you instead look at bleaching, or ratio of live to recently dead, etc.? Some locations are inherently low in cover, and others high. Low cover does not always mean low health.\n\nThere seems to be some mixing of the different sections. In the Results and Discussion, the first part detailing categories is actually M&M. Also, the statement of 89 spp. of hard coral (hermatypic? Scleractinian?) being listed within your results is a bit misleading, even with the reference, as you did not identify your animals to species-level. The sentence is not even needed here.\n\nAre there any environmental (water-quality) data available for this region? This could help your Discussion. I also do not completely agree with the coral mortality without more details. By Figure 3 inset and your text, it seems many corals may have died recently, yet you state the conditions of the reef are stable compared to 2010, 2012 and 2014. Are these also just 25% bins? In my opinion, Figure 3 is not needed.\n\nThe authors mention many possible causes of degradation in the Discussion, but provide almost no data or observations, except for some anecdotal ones. This section reads less like a paper and more like a textbook. As a scientific work, it would be good to more clearly link these causes with your field observations, which you do to some degree with the fishing practices. Observing 2 COTS during surveys does not really indicate much, unfortunately.\n\nFinally, the English, while very easy to read and generally well-done, could do with a simple brush-up here and there. I imagine a colleague could do this easily for you.\n\nSorry, one more small-ish comment. In Figure 2 - \"abiotic\" includes rubble - isn't this also biotic if coral rubble? I am a bit confused if some sites do not add up to 100%, as by including all these categories, shouldn't each site = 100%?\n\nThe authors are free to contact me with any questions or comments; I am more than happy to help if I can.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/7-1962
https://f1000research.com/articles/7-1961/v1	20 Dec 18	{ "type": "Software Tool Article", "title": "pdb-tools: a swiss army knife for molecular structures", "authors": [ "João P. G. L. M. Rodrigues", "João M.C. Teixeira", "Mikaël Trellet", "Alexandre M. J. J. Bonvin", "João M.C. Teixeira", "Mikaël Trellet" ], "abstract": "The pdb-tools are a collection of Python scripts for working with molecular structure data in the Protein Data Bank (PDB) format. They allow users to edit, convert, and validate PDB files, from the command-line, in a simple but efficient manner. The pdb-tools are implemented in Python, without any external dependencies, and are freely available under the open-source Apache License at https://github.com/haddocking/pdb-tools/ and on PyPI.", "keywords": [ "bioinformatics", "chemistry", "macromolecules", "PDB", "protein structure", "Python", "structural biology" ], "content": "Introduction\n\nObtaining and analyzing three-dimensional structures of biological macromolecules, such as proteins or nucleic acids, is often a key step towards understanding their biological function. Of the many file formats used in structural biology to store three-dimensional coordinate data, the Protein Data Bank (PDB) format remains one of the most widely adopted, despite the introduction of a new standard - the mmCIF format - in recent years1,2. The PDB format encodes 44 possible different record types in a human-readable flat-text format, with each record having a number of fields with stringent spacing rules in a total of 80 characters. This strictness and complexity of the PDB format make manual editing difficult. As a result, researchers use molecular viewers, e.g. PyMOL3 and VMD4, to perform basic editing tasks, such as selecting chains from a structure. More advanced edits, such as changing chain identifiers, deleting specific atoms, or renumbering residues require expertise with scripting languages, whose syntax varies from viewer to viewer. Other edits, such as selecting atomic positions based on occupancy values, are even more challenging to accomplish in molecular viewers, if possible at all. The alternative is for researchers to become proficient in a programming language, such as Python, and use one of its structural bioinformatics libraries, such as BioPython5 or ProDy6.\n\nHere, we present pdb-tools, a modular toolkit providing several everyday tasks when handling PDB files, namely downloading, editing, filtering, merging, sorting, and validation, as well as conversion to and from the more recent PDBx/mmCIF file format2 (Figure 1). In order to shorten the learning curve for new users, all pdb-tools implement a unified command-line interface. Moreover, to support complex editing operations, we designed the toolkit to allow the serial concatenation of several tools in a pipeline without the need to read or write intermediary files. For developers, the coherent architecture simplifies maintenance and development of new tools.\n\nThe pdb-tools are written in Python and are currently developed on GitHub under the open-source Apache License version 2.0. In addition, to simplify the installation procedure for the end-user, the toolkit is also available for download through PyPI. Finally, pdb-tools are also part of the SBGrid initiative7.\n\nA common usage example is shown on the gray box insert.\n\n\nMethods\n\nThe pdb-tools toolkit is implemented in CPython, using only modules of the standard library, which allows us to support a wide range of Python versions (2.7 and all of the 3.x series). The tools make use of Python generator expressions to improve performance and memory usage. This feature, along with support for streaming input and output data, allows users to create complex pipelines by chaining together different tools (e.g. format conversion followed by chain selection followed by editing followed by validation).\n\nAs a result, using ’pdb_reres’ to renumber the residues of a structure with 64,606 atoms takes ~0.17 seconds (on an Intel i7-7500U CPU) and ~10 MB of memory. Merging eight copies of the same structure using ’pdb_merge’ took ~0.74 seconds and ~18 MB of memory. These performance indicators make pdb-tools particularly useful when combined with shell scripting for batch processing entire collections of PDB files.\n\nTo aid the maintenance and development of pdb-tools, we wrote approximately 250 separate unit tests that provide 81% coverage for the tools and their possible usage options. In addition, each commit is checked for coding style against the PEP8 style guide, using flake8. Besides hosting the source code on GitHub, we use setuptools to package and distribute pdb-tools through PyPI. Further, we use semantic versioning to indicate compatibility between releases with the same major version. The code is tested continuously using Travis CI and AppVeyor webhooks on GitHub, which monitor incoming pull requests and any commits to the master branch. We test on virtual instances of Windows 10 and Linux 16.04 LTS, under Python 2.7, 3.6, and 3.7. Nevertheless, the pdb-tools should run on most UNIX derivatives and Windows versions, provided a supported version of Python in installed.\n\nConcerning documentation, each tool contains a self-contained help string, describing its purpose, command-line interface, and usage examples. This help text is available by running the tool without any argument. Additional documentation on the usage and installation of pdb-tools is available online on the project’s web-page and in the README file accompanying the source code.\n\nFinally, we develop pdb-tools as a community open-source project on GitHub and encourage contributions of any kind. To date, the project has more than 30 forks and has been starred by over a dozen users on GitHub. To help and promote these contributions, and based on our experience with building pdb-tools collaboratively, we have documents setting coding best practices, contribution guidelines, and a code of conduct for developers on our GitHub repository. In addition, we make use of and encourage users to contribute to our issue tracker, which documents our discussions and development decisions publicly.\n\n\nUse Cases\n\nThe pdb-tools follow a ’one tool, one job’ design philosophy, where each tool performs a simple operation on an input file, while more complex operations can be achieved by chaining different tools together. Over the years, the pdb-tools have been used in both research and educational contexts, some of which we highlight below.\n\nMolecular viewers, such as PyMOL or VMD, are the tools of choice for making selections of specific chains or residue ranges in PDB files. When handling more than one structure, programming libraries such as Biopython or Prody are better suited for the task. The selection tools included in pdb-tools offer a simple solution to make complex selections.\n\n\n\nImportantly, users can use shell scripting to efficiently process large collections of structures. Those using UNIX-based systems can also make use of the parallel utility to distribute processing over several cores.\n\n\n\nHADDOCK is an integrative modeling software for protein interactions8, whose web server provides a user-friendly interface that oversaw over 200.000 job submissions to date. The simplest of submissions requires two structures, in PDB format, that cannot have more than one chain, cannot have overlapping residue numbers, and cannot have alternate locations for any atom. Often, users submitting jobs to the HADDOCK web server are faced with error messages because of such formatting requirements. Using pdb-tools, producing a HADDOCK-compliant PDB file is straightforward, as we demonstrate below.\n\n\n\nAlthough PDB files may include SEQRES records that store the sequence of the construct used for structure determination, the final structure does not necessarily include all the amino acids or nucleotides. This discrepancy is particularly important when building alignments for homology modeling. To this end, we include in pdb-tools a PDB to FASTA converter that extracts the sequence of the structure directly from the ATOM/HETATM records. The user can specify if the entire structure should be exported as a single FASTA sequence record, or divided by chain, using the -multi option.\n\n\n\nDue to several limitations with the PDB file format, the Protein Data Bank recently introduced the new PDBx/mmCIF format. To support this decision, we include in pdb-tools converters to and from the mmCIF format. These converters allow users to implicitly use pdb-tools on mmCIF files. However, large structures that cannot abide by the PDB format specification (too many chains or residues or atoms) cannot be converted.\n\n\n\nThe pdb-tools provide a command-line interface to edit three-dimensional molecular structures in the PDB format. Since it is written in Python, without any third-party dependencies, the toolkit is available on Linux, Mac OS, and Windows and supported on both Python 2 and 3. The source code is hosted on GitHub and licensed under the open-source Apache License version 2.0, to encourage contributions from the community, and available for installation through PyPI and SBGrid. As such, the pdb-tools are an efficient, portable, and extremely flexible toolkit for both starting and experienced researchers in structural biology.\n\n\nSoftware availability\n\npdb-tools is available from: https://haddocking.github.io/pdb-tools/, https://pypi.org/project/pdb-tools/, and https://sbgrid.org/software/titles/pdb-tools\n\nSource code available from: https://github.com/haddocking/pdb-tools, https://github.com/haddocking/pdb-tools/releases/tag/2.0.0-rc1\n\nArchived source code as at time of publication: http://doi.org/10.5281/zenodo.21680709\n\nLicense: Apache License, version 2.0", "appendix": "Grant information\n\nJ.P.G.L.M.R acknowledges funding from a Niels Stensen Fellowship and NIH grant R35GM122543. M.T. and A.M.J.J.B were supported by the European H2020 e-Infrastructure grant BioExcel (grant no. 675728).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nThe authors acknowledge the members of the Bonvin group for fruitful discussions and suggestions of new tools over the years.\n\n\nReferences\n\nWestbrook JD, Fitzgerald PM: The PDB format, mmCIF, and other data formats. Methods Biochem Anal. 2003; 44: 161–179. PubMed Abstract \| Publisher Full Text\n\nwwPDB consortium: Protein Data Bank: the single global archive for 3D macromolecular structure data. Nucleic Acids Res. 2018; gky949. PubMed Abstract \| Publisher Full Text\n\nSchrödinger LLC: The PyMOL Molecular Graphics System, Version 1.8. 2015.\n\nHumphrey W, Dalke A, Schulten K: VMD: visual molecular dynamics. J Mol Graph. 1996; 14(1): 33–38, 27-8. PubMed Abstract \| Publisher Full Text\n\nCock PJ, Antao T, Chang JT, et al.: Biopython: freely available Python tools for computational molecular biology and bioinformatics. Bioinformatics. 2009; 25(11): 1422–1423. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBakan A, Meireles LM, Bahar I: ProDy: protein dynamics inferred from theory and experiments. Bioinformatics. 2011; 27(11): 1575–1577. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMorin A, Eisenbraun B, Key J, et al.: Collaboration gets the most out of software. eLife. 2013; 2: e01456. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nvan Zundert GCP, Rodrigues JPGLM, Trellet M, et al.: The HADDOCK2.2 Web Server: User-Friendly Integrative Modeling of Biomolecular Complexes. J Mol Biol. 2016; 428(4): 720–725. PubMed Abstract \| Publisher Full Text\n\nRodrigues J, Teixeira JMC, Trellet M, et al.: haddocking/pdb-tools: Release Candidate 1 (Version 2.0.0-rc1). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.2168070" }	[ { "id": "42178", "date": "04 Jan 2019", "name": "Bruno L. Victor", "expertise": [ "Reviewer Expertise Molecular modelling", "computational chemistry", "computational medicinal chemistry", "software development applied to drug discovery projects" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this work, Rodrigues et al, present to the readers a collection of python based tools to edit, convert and validate PDB files, in a very simple and efficient way. The authors clearly describe in the methods section the procedure used to implement the pdb-tools toolkit, and they even provide some simple use cases where the toolkit can be applied. I have installed the toolkit in my workstation and I did not have any kind of problem. I have tested several of the available functionalities and I have found it to be very useful and easy-to-use on day-to-day work (as the authors clearly state in the article). This type of contribution software is very useful for any scientist (with beginner or advanced skills) dealing with structure data in PDB format.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [] }, { "id": "42177", "date": "18 Jan 2019", "name": "Peter W. Rose", "expertise": [ "Reviewer Expertise Structural Bioinformatics", "Scientific Software Development" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nWhile there are many, often complex, tools available to manipulate pdb files, pdb-tools stands out as a light-weight set of command line tools for common use cases. The requirement for each tool to perform one task makes these tools easy to use in shell scripts and to combine into a pipeline for batch processing a set of PDB structures.\n\nOn the software side, this is a well-designed open source tool kit available in GitHub that follows best software development practices, including documentation, checking for consistent coding style, high test coverage, using continuous integration, semantic versioning, and deployment on PyPi. A liberal open source license encourages community contributions. The modular design makes this project easy to maintain and expand compared to many of the legacy tools still in use.\n\nRecommendation:\n\nList all, not just a subset, of the available tools with an example on the documentation page. That way, a Google search will find those tools and examples are helpful for new users.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1961
https://f1000research.com/articles/7-1615/v1	08 Oct 18	{ "type": "Research Article", "title": "Mean deviation based identification of activated voxels from time-series fMRI data of schizophrenia patients", "authors": [ "Indranath Chatterjee" ], "abstract": "Background: Schizophrenia is a serious mental illness affecting different regions of the brain, which causes symptoms such as hallucinations and delusions. Functional magnetic resonance imaging (fMRI) is the most popular technique to study the functional activation patterns of the brain. The fMRI data is four-dimensional, composed of 3D brain images over time. Each voxel of the 3D brain volume is associated with a time series of signal intensity values. This study aimed to identify the distinct voxels from time-series fMRI data that show high functional activation during a task. Methods: In this study, a novel mean-deviation based approach was applied to time-series fMRI data of 34 schizophrenia patients and 34 healthy subjects. The statistical measures such as mean and median were used to find the functional changes in each voxel over time. The voxels that show significant changes for each subject were selected and thus used as the feature set during the classification of schizophrenia patients and healthy controls. Results: The proposed approach identifies a set of relevant voxels that are used to distinguish between healthy and schizophrenia subjects with high classification accuracy. The study shows functional changes in brain regions such as superior frontal gyrus, cuneus, medial frontal gyrus, middle occipital gyrus, and superior temporal gyrus. Conclusions: This work describes a simple yet novel feature selection algorithm for time-series fMRI data to identify the activated brain voxels that are generally affected in schizophrenia. The brain regions identified in this study may further help clinicians to understand the illness for better medical intervention. It may be possible to explore the approach to fMRI data of other psychological disorders.", "keywords": [ "fMRI", "Schizophrenia", "Time-series", "Classification" ], "content": "Introduction\n\nSchizophrenia is a severe mental disorder that affects different regions of the brain, often involving hallucinations and delusions. Functional magnetic resonance imaging (fMRI) data comprising 3D brain scans acquired over time (thus resulting in a 4D set) is often used to study brain regions affected by schizophrenia. Each voxel of the 3D brain volume is associated with a time series of signal intensity values. General linear model (GLM)1 and independent component analysis (ICA)2 are often employed to study the voxel activity by transforming the 4D time-series data to a 3D spatial map.\n\nThe present work involves a novel application of mean deviation on time-series fMRI data to identify the distinct voxels that show high functional activation during a task. The work aims to identify the relevant brain regions that are affected in schizophrenia. Further, the identified voxels (features) are used to distinguish between schizophrenia patients and healthy subjects.\n\n\nMethods\n\nThe time-series fMRI data having 1.5T strength was taken from the FBIRN phase – II data repository3 available at site 0009 and site 0010. From the dataset, four different runs of auditory oddball task data of 34 schizophrenic patients (group G1) and 34 healthy controls (group G2) were extracted. Every run of each subject’s data contains 140 brain volumes acquired in 280 seconds time (TR = 2 seconds). Table 1 shows the dataset details.\n\nPre-processing of the fMRI data was done using SPM8 toolbox in Matlab2014b. The temporal variation was corrected using slice timing correction, followed by the motion correction using realignment. Each of the fMRI scans was spatially normalized into standard Montreal Neurological Institute (MNI) space using an EPI template yielding voxel dimension of 3×3×3 mm3. Finally smoothing was done using a 9×9×9 mm3 full width at half maximum (FWHM) Gaussian kernel, resulting in a 3D brain volume containing 53×63×46, i.e., 1,53,594 voxels.\n\nThe activation pattern of the voxels was analysed in two phases.\n\nPhase I. In the first phase, identification of voxels exhibiting high activation pattern (anytime during its time-course) is carried out for each subject. As the study focused on the variation in the signal intensity of the voxels (V) over time, absolute mean deviation (Vd¯) for each of the 140 time points was computed for each voxel, and the median (M) of the 140 values of V− was found. Mean deviation (V−) values were compared with α times M (α was chosen to be 3, based on experimentation) to identify whether a voxel exhibited high level of activation at any time during the 140 units of time. This voxel-wise analysis was performed for all the voxels of a given subject. Thus, a set of relevant voxels showing high degree of activation was obtained for each subject.\n\nPhase II. In the second phase, a common subset of voxels exhibiting high degree of activation across all the subjects within a group was obtained. Finally, both the subsets belonging to groups G1 (schizophrenia patients) and G2 (healthy controls) were merged to get the set S. The voxels in set S were backtracked to MNI brain space and finally mapped into Talairach’s space4 to identify the brain regions. This procedure has been described in Algorithm 1.\n\nClassification. The set S was used to distinguish between schizophrenia patients and healthy subjects using two classifiers, viz., support vector machine (SVM) with sigmoid kernel5 and extreme learning machine (ELM) classifier6.\n\nExperimental settings. All the implementations were done in MATLAB2014b. Parameter α was varied in the range1,7 in steps of 1 to identify the number of voxels that exhibited a high level of activations during the task. When the value of α was taken as 1 and 2, a large number of voxels showed activation level higher than α times M, resulting in set S having voxels that represents almost the entire brain. However, for α = 3, it was found that set S contained only 1580 distinct voxels that mapped to the brain regions which are generally affected in schizophrenia. When α was taken as more than 3, the number of voxels in the set S were close to zero rendering it too small for any meaningful analysis. Thus, α = 3 was found to be the most suitable value.\n\nFurther, the set S of voxels obtained α = 3 was used to fine-tune the classifiers. The SVM classifier gave the best results for the regularization parameter C = 1.09, and sigmoid kernel based ELM classifier gave best the results with 503 hidden neurons.\n\nTo evaluate the distinguishing capability of the voxels/features in set S, a comparison was done between the classification accuracy obtained using S and the accuracy obtained using the voxels set given by the GLM based approach. In this case, GLM was applied using SPM8 toolbox to convert the 4D time-series fMRI data to 3D contrast map for each subject. The GLM yielded an activation map comprising around 60000 voxels out of 153594 which were activated during the task.\n\n\n\nNotations:\n\nm (=34): the number of subjects in each group\n\nn (=140): the number of observations in a run\n\nVi : time-series of ith voxel\n\ni.e. Vi = [vi,1 vi,2 vi,3 ⋯ vi,n];\n\nμi:meanofVii.e.μi=∑j=1j=nvi,jn\n\nSteps:\n\n1. Calculate absolute mean deviation for each voxel using V−di=\|Vi−μι\|.\n\n2. Find median Mi of V−di.\n\n3. For each subject k ∈{1,2,...,m}, select the set Vsk of voxels that show deviation higher than αMi.\n\n4. Find the group wise intersection of the voxels selected in step 3 for groups G1 and G2\n\ni.e. VsG1=∩k=1mVsk(G1)\n\nVsG2=∩k=1mVsk(G2)\n\n5. Merge the two sets, obtained in step 4 to obtain set S\n\ni.e. S = VSG1 ∪ VSG2\n\n6. Map S into the brain space to identify affected regions.\n\n\nResults\n\nA comparison of the results of the classification accuracies obtained using feature sets given by the GLM and the proposed approach is shown in Table 2. The features selected by the proposed approach when backtracked to Talairach’s space revealed the brain regions that are generally affected in schizophrenia8–10, which validates the efficacy of the approach. The distribution of the selected voxels that distinguish the schizophrenia patients from the healthy subjects is shown in Figure 1 (a–d). Figure 2 (a–c) show the activated voxels when plotted on a sample fMRI image for an axial, coronal and sagittal view of the brain.\n\nIdentified brain regions at different levels of hierarchy, namely, hemisphere level (a), lobes level (b), gyrus level (c), and Brodmann's area level (d).\n\nVoxels identified by the proposed approach plotted over a functional brain image in different views of the brain, i.e., axial (a), coronal (b) and sagittal (c) plane.\n\n\nDiscussion\n\nUnlike other conventional methods such as GLM to select the voxels showing a statistically significant response to the experimental conditions7, the proposed approach identifies the neural activity in a particular voxel over time, irrespective of any experimental condition. The proposed approach does not require any details for the task and conditions. It works on the temporal values of each voxel for each subject's data one by one. Like other multi-voxel pattern analysis (MVPA) methods7,11,12, this approach also tries to find the participation of multiple voxels when selecting the final set of relevant voxels across a particular group of the subjects.\n\nThe classification accuracies, as shown in Table 2, demonstrate the efficacy of the proposed methodology. The reduced set of 1580 voxels achieved a much higher accuracy when compared to the GLM approach. Figures 1a–d show the distribution of the selected voxels for each level of brain regions. These regions show distinct changes in functional activation in schizophrenia patients when compared to healthy controls, and thereby distinguish between schizophrenia and healthy subjects with high classification accuracy. Most of the regions identified in the study comply with the existing literature13–16. The regions such as superior frontal gyrus, cuneus, lingual gyrus, medial frontal, middle occipital gyrus, superior temporal gyrus, anterior cingulate, and declive show the changes in functional activation. Studies showed functional changes in superior frontal gyrus13, superior temporal gyrus15, lingual gyrus15, and cuneus15. Even functional abnormality in anterior cingulate was found in several studies16,17. The literature also suggests functional changes in middle occipital gyrus18. When observed at the cell level of brain regions in Talairach’s space, this study shows distinguishable functional changes in Brodmann’s area (BA) 18, 10, 9, 17, 19, 32, 21, 37, 11, and BA 6. Previous studies14,19 also showed changes in functional activation in these areas of the brain.\n\n\nConclusions\n\nThis work describes a simple and fast feature selection algorithm based on mean deviation for time-series fMRI data to identify the activated brain voxels that are generally affected in schizophrenia. The proposed approach was found to be efficient in selecting a minimal set of relevant voxels directly from time-series 4D fMRI data. The obtained voxel set was capable of distinguishing between healthy and schizophrenic subjects. One may explore the possibility of applying this approach to fMRI data of other psychological disorders.\n\n\nData availability\n\nThe Matlab source codes, a text file containing dataset details including subject ID and their age, and the instructions for the study can be found at: https://github.com/IndraChatterjee/AnomalyDetection_TimeSeries_fMRI_Schizophrenia.\n\nThe complete source codes are archived in a publicly accessible record at: https://doi.org/10.5281/zenodo.143853920\n\nLicense: CC0\n\nThe four runs of auditory oddball task fMRI data from the FBIRN phase II repository can be downloaded from http://schizconnect.org/ querying 1.5T fMRI data for healthy and schizophrenia subjects available at site 0009 and 0010. The list of subjects chosen for this study is mentioned in the ‘DataDetails_FBIRN15T.txt’ file available at the GitHub repository. Users are required to sign-up to SchizConnect to download data and conditions of use are as written in the data use agreement of the FBIRN project.\n\n\nAuthor endorsement\n\nCameron Craddock confirms that the author has an appropriate level of expertise to conduct this research, and confirms that the submission is of an acceptable scientific standard. Cameron Craddock declares the following competing interests: I am the Chair of Brainhack, and this organisation awarded this paper this year's Brainhack poster prize. Affiliation: Associate Professor of Diagnostic Medicine, Dell Medical School, The University of Texas at Austin, Austin, TX, USA.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nThe author would like to thank the organizers and all the attendees of 2018 OHBM Brainhack Singapore.\n\nData used for this study were hosted in the Function BIRN Data Repository (http://fbirnbdr.birncommunity.org:8080/BDR/) using Project Accession Number 2007-BDR-6UHZ1, supported by grants to the Function BIRN (U24-RR021992) Testbed funded by the National Center for Research Resources at the National Institutes of Health, U.S.A.\n\n\nReferences\n\nFriston KJ, Holmes AP, Worsley KJ, et al.: Statistical parametric maps in functional imaging: a general linear approach. Hum Brain Mapp. 1994; 2(4): 189–210. Publisher Full Text\n\nKim DI, Mathalon D, Ford JM, et al.: Auditory oddball deficits in schizophrenia: an independent component analysis of the fMRI multisite function BIRN study. Schizophren Bull. 2009; 35(1): 67–81. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKeator DB, van Erp TG, Turner JA, et al.: The Function Biomedical Informatics Research Network Data Repository. NeuroImage. 2016; 124(Pt B): 1074–1079. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLancaster JL, Woldorff MG, Parsons LM, et al.: Automated Talairach atlas labels for functional brain mapping. Hum Brain Mapp. 2000; 10(3): 120–131. PubMed Abstract \| Publisher Full Text\n\nCortes C, Vapnik V: Support-vector networks. Mach Learn. 1995; 20(3): 273–297. Publisher Full Text\n\nHuang GB, Zhu QY, Siew CK: Extreme learning machine: theory and applications. Neurocomputing. 2006; 70(1–3): 489–501. Publisher Full Text\n\nNorman KA, Polyn SM, Detre GJ, et al.: Beyond mind-reading: multi-voxel pattern analysis of fMRI data. Trends Cogn Sci. 2006; 10(9): 424–430. PubMed Abstract \| Publisher Full Text\n\nCastro E, Gómez-Verdejo V, Martínez-Ramón M, et al.: A multiple kernel learning approach to perform classification of groups from complex-valued fMRI data analysis: application to schizophrenia. NeuroImage. 2014; 87: 1–17. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGarrity AG, Pearlson GD, McKiernan K, et al.: Aberrant “default mode” functional connectivity in schizophrenia. Am J Psychiatry. 2007; 164(3): 450–457. PubMed Abstract \| Publisher Full Text\n\nGur RE, Gur RC: Functional magnetic resonance imaging in schizophrenia. Dialogues Clin Neurosci. 2010; 12(3): 333–343. PubMed Abstract \| Free Full Text\n\nFormisano E, De Martino F, Valente G: Multivariate analysis of fMRI time series: classification and regression of brain responses using machine learning. Magn Reson Imaging. 2008; 26(7): 921–934. PubMed Abstract \| Publisher Full Text\n\nDe Martino F, Valente G, Staeren N, et al.: Combining multivariate voxel selection and support vector machines for mapping and classification of fMRI spatial patterns. NeuroImage. 2008; 43(1): 44–58. PubMed Abstract \| Publisher Full Text\n\nHof PR, Haroutunian V, Friedrich VL Jr, et al.: Loss and altered spatial distribution of oligodendrocytes in the superior frontal gyrus in schizophrenia. Biol Psychiatry. 2003; 53(12): 1075–1085. PubMed Abstract \| Publisher Full Text\n\nChatterjee I, Agarwal M, Rana B, et al.: Bi-objective approach for computer-aided diagnosis of schizophrenia patients using fMRI data. Multimed Tools Appl. 2018; 77(20): 26991–27015. Publisher Full Text\n\nHoptman MJ, Zuo XN, Butler PD, et al.: Amplitude of low-frequency oscillations in schizophrenia: a resting state fMRI study. Schizophr Res. 2010; 117(1): 13–20. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVercammen A, Knegtering H, den Boer JA, et al.: Auditory hallucinations in schizophrenia are associated with reduced functional connectivity of the temporo-parietal area. Biol Psychiatry. 2010; 67(10): 912–918. PubMed Abstract \| Publisher Full Text\n\nCarter CS, Mintun M, Nichols T, et al.: Anterior cingulate gyrus dysfunction and selective attention deficits in schizophrenia: [15O]H2O PET study during single-trial Stroop task performance. Am J Psychiatry. 1997; 154(12): 1670–1675. PubMed Abstract \| Publisher Full Text\n\nBrunet E, Sarfati Y, Hardy-Baylé MC, et al.: Abnormalities of brain function during a nonverbal theory of mind task in schizophrenia. Neuropsychologia. 2003; 41(12): 1574–1582. PubMed Abstract \| Publisher Full Text\n\nCamchong J, Dyckman KA, Austin BP, et al.: Common neural circuitry supporting volitional saccades and its disruption in schizophrenia patients and relatives. Biol Psychiatry. 2008; 64(12): 1042–1050. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChatterjee I: Feature selection technique for time-series fMRI data of schizophrenia patients [Source Code]. Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1438539" }	[ { "id": "39310", "date": "01 Nov 2018", "name": "Sahil Bajaj", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nHere author describes an interesting algorithm to identify the activated brain voxels affected in schizophrenia from time-series fMRI data.\nI think this is an interesting paper and a nice example which can be implemented in more severe cases of mental distress.\nHowever, I have few minor concerns:\n\nHow did the author remove the effect of gender, I noticed that there are way more males in the data than females. I can see some voxels outside the brain and which are at the skull. I am not sure why did the author get those voxels and if the author made any effort to exclude those voxels?\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "39217", "date": "22 Nov 2018", "name": "Sagarika Bhattacharjee", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe present study describes an methodology that classifies schizophrenia patients from healthy controls. The study claims to demonstrate high classification accuracy. The study has significant relevance to the neuroscience community however I have following concerns:\nThe functional significance of the brain regions involved needs to be elaborated so that their activation could be validated. The description of the role of obtained brain regions in schizophrenia patients and healthy individuals will indicate that the obtained regions are actually involved in schizophrenic patients and not a result of Type II error. It will be good to provide some details about the demographics, level of education, duration of disease, medication history of the participants in order to evaluate the role of these confounding factors in the obtained results. These factors might cause some variation in the fMRI signals and just wondering if any of these parameters were taken as covariate in the analysis. The fMRI signals obtained are task state and not resting state. These signals were obtained while doing oddball paradigm. So, I was wondering whether such classification would apply to schizophrenia patients only when they are doing this particular task, or it would apply to all schizophrenic patients irrespective of their state.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4291", "date": "20 Dec 2018", "name": "Indranath Chatterjee", "role": "Author Response", "response": "I am thankful to the respected referee for her insightful comments and valuable suggestions. I have updated the manuscript in accordance with the suggestions and queries. The suggested changes are made in the discussion section of the revised manuscript. I have also included some demographic details of the subjects in the dataset table. As I have not incorporated any covariates in this study, the limitation and the scope of future works are added in the discussion section." } ] } ]	1	https://f1000research.com/articles/7-1615
https://f1000research.com/articles/7-1682/v1	22 Oct 18	{ "type": "Method Article", "title": "Embryonic zebrafish xenograft assay of human cancer metastasis", "authors": [ "David Hill", "Lanpeng Chen", "Ewe Snaar-Jagalska", "Bill Chaudhry", "Lanpeng Chen", "Ewe Snaar-Jagalska", "Bill Chaudhry" ], "abstract": "Cancer metastasis is the most important prognostic factor determining patient survival, but currently there are very few drugs or therapies that specifically inhibit the invasion and metastasis of cancer cells. Currently, human cancer metastasis is largely studied using transgenic and immunocompromised mouse xenograft models, which are useful for analysing end-point tumour growth but are unable to accurately and reliably monitor in vivo invasion, intravasation, extravasation or secondary tumour formation of human cancer cells. Furthermore, limits in our ability to accurately monitor early stages of tumour growth and detect micro-metastases likely results in pain and suffering to the mice used for cancer xenograft experiments. Zebrafish (Danio rerio) embryos, however, offer many advantages as a model system for studying the complex, multi-step processes involved during cancer metastasis. This article describes a detailed method for the analysis of human cancer cell invasion and metastasis in zebrafish embryos before they reach protected status at 5 days post fertilisation. Results demonstrate that human cancer cells actively invade within a zebrafish microenvironment, and form metastatic tumours at secondary tissue sites, suggesting that the mechanisms involved during the different stages of metastasis are conserved between humans and zebrafish, supporting the use of zebrafish embryos as a viable model of human cancer metastasis. We suggest that the embryonic zebrafish xenograft model of human cancer is a tractable laboratory model that can be used to understand cancer biology, and as a direct replacement of mice for the analysis of drugs that target cancer invasion and metastasis.", "keywords": [ "Zebrafish embryo", "xenograft", "cancer", "melanoma", "prostate cancer", "metastasis", "replacement" ], "content": "\n\nOptimal xenotransplantation can be performed in zebrafish embryos at 48 hpf, allowing for a 72-hour period to model key stages of metastatic behaviour.\n\nMetastatic processes can be visualised at the single cell level.\n\nZebrafish embryos can be used to replace mouse xenograft models in early cancer metastasis research.\n\nOnly a small number of cancer cells (100–200 cells per fish) are required.\n\nUse of fluorescent cell markers in conjunction with transgenic zebrafish lines allows for host cells to be distinguished from human cancer cells in real-time/in situ, avoiding the need for post-mortem immunocytochemistry.\n\nThe zebrafish embryo assay is higher-throughput than mouse xenograft models.\n\nStudying tumour invasion and metastatic dissemination of different human cancer cell lines using time-lapse microscopy.\n\nStudying the invasion of human cancer cells into zebrafish blood vessels and in the formation of secondary tumours.\n\nStudying the heterogeneity of tumours and cooperation of different cancer cells from patient-derived tumours.\n\nStudying the remodelling of the extracellular matrix during tumour invasion.\n\nAs a screening assay to identify new agents/drugs that reduce the metastatic behaviour of cancer cells.\n\n\nIntroduction\n\nMetastasis is a clinical term describing the spread of tumour cells from a primary location to distant sites. It is suggested that more than 90% of deaths from cancer are not caused by the primary tumour but by the direct effects of metastatic deposits and from the metabolic burden of a rapidly growing tumour cell mass (Jemal et al., 2011). Traditionally an orderly cascade of cellular behaviours was presumed to underlie the progression from a well circumscribed and localised tumour growth to distant spread, based on initial local invasion, entry into the vascular or lymphatic system, survival in those fluid channels followed by extravasation and colonisation in a distal site (Massagué et al., 2017). However, this orderly progression is not borne out by current research and the mechanisms of metastatic spread remain controversial. The role of epithelial to mesenchymal transformation (EMT) is unclear and plasticity of cells that metastasise and their relationship to the primary tumour cells, e.g. stem cells, remains the object of current research (Pandya et al., 2017). Furthermore, cancers do not appear to disseminate randomly, but exhibit tropism for specific organs, especially lung, liver and bone (Tarin, 2011). This observation, made over 100 years ago by Steven Paget, led to the \"seed and soil\" hypothesis which remains unproven. In clinical practice, surgical resection or local treatment of primary tumours is effective, but metastases remain difficult to treat. This is particularly evident for melanoma, where localised and slow-growing metastatic deposits can appear long after apparent cure (Gershenwald & Scolyer, 2018). Similarly, in prostatic cancer the primary site is rarely a clinical problem in comparison to the pain and pathological fractures from osteolytic vertebral deposits (Akakura et al., 1996).\n\nUnderstanding the multi-step processes that regulate cancer metastasis will likely result in new therapeutics to benefit patients with a wide range of cancers at different stages of progression. Although in vitro systems, e.g. the artificial skin model for melanoma (Hill et al., 2015) can be highly effective for studying primary tumour behaviour, connected organ systems are needed to understand metastasis. The mouse has traditionally been used as a pre-clinical model organism to study cancer under the rationale that they are a mammalian species, with the same organ systems as humans. Although genetically modified animals do spontaneously develop tumours, the introduction of human tumour cells into other species, xenografting, is a vital pre-clinical tool that enables researchers to study tumour metastasis and evaluate drug responses (van Marion et al., 2016). Xenografts provide greater experimental control and can provide a direct translational link to the patient, particularly when the developmental origin of cancer remains unknown. However within a mouse, metastatic spread from xenografts often occurs late, well after the primary deposit has become distressing to the animal, and further pain can also result from the aggressive invasive nature of the metastases (Gómez-Cuadrado et al., 2017). Highly metastatic cell lines are often used to accelerate the development of metastatic tumours, but these may not reflect normal metastasis, and therefore several different lines must be used, requiring many more animals (Cruz-Munoz et al., 2008). It is sometimes possible to surgically remove the primary tumour prior to analysis of metastatic dissemination (Srivastava et al., 2014); however, this is often associated with excessive tissue damage requiring prolonged post-operative analgesia. Direct injection of cancer cells into the tail vein (Elkin & Vlodavsky, 2001; Minn et al., 2005), heart (Kang et al., 2003), illiac artery (Bos et al., 2009; Wang et al., 2015), spleen (Morikawa et al., 1988), peritoneum (Chu et al., 2015) or tibia (Fisher et al., 2002) have all been used to model local metastatic behaviours, but the mouse model is limited since metastasising single cells cannot be tracked and only relatively large metastatic growths can be detected, precluding study of the earliest metastatic events. Furthermore, mouse models have also had limited success when predicting anti-cancer drug efficacy in human trials (Day et al., 2015; Kersten et al., 2017).\n\nThe zebrafish is a tropical bony fish which for over 30 years has been increasingly used in developmental biology and human disease modelling as it contains almost all human organ systems except lungs (Penberthy et al., 2002). The zebrafish genome has been sequenced and there is a high degree of conserved genes and genetic signalling pathways compared to humans (Howe et al., 2013). Importantly for the study of cancer metastasis, embryos are completely transparent, facilitating imaging at single cell level within developing organs whilst also imaging the entire animal. Furthermore, the majority of studies can be carried on early-stage embryos before they are capable of independent feeding (which for the zebrafish is widely considered to be 5 days post fertilisation (dpf)), and protected under the Animals (Scientific Procedures) Act (ASPA) and EU Directive (2010/63/EU). The extra-uterine development of hundreds of eggs also permits a greater number of studies in genetically identical organisms. Since the first reported xenotransplantation of human cells into zebrafish (Lee et al., 2005), many laboratories have shown that zebrafish embryos are useful for the study of other facets of tumour biology including cancer-induced angiogenesis (Haldi et al., 2006); cancer cell invasion and metastasis (de Boeck et al., 2016; Marques et al., 2009); cancer stem cell growth (Bansal et al., 2014; Chen et al., 2017); interaction of cancer cells with the host (Feng & Martin, 2015); and drug screening (Corkery et al., 2011; Gibert et al., 2013). Importantly, the development of human tumours and their response to chemotherapeutic treatment in zebrafish embryos is comparable to that observed in mouse xenograft assays (Fior et al., 2017). Additionally, while mouse xenograft models require immuno-deficient mice to prevent immune-rejection of the human cancer cells, the lack of a mature adaptive immune system within zebrafish embryos up to 14 days post-fertilisation (dpf) allows analysis of human cancers without rejection (Lam et al., 2004).\n\nIn this article we describe the techniques for performing embryonic zebrafish xenograft experiments and demonstrate the utility of using zebrafish embryos as a model system for studying human cancer metastasis, in particular metastatic melanoma and prostate cancer. We highlight the advantages over mouse xenograft models and provide a practical experimental protocol showing how zebrafish embryos can be used as a replacement for mice to conveniently study metastatic tumour behaviour in the laboratory.\n\n\nMethods\n\nA full step-by-step protocol can be found in Supplementary File 1.\n\nTransparent Casper Tg(kdrl;GFP) zebrafish were housed under standard conditions at 28.5°C (Westernfield, 2000). All animals were maintained under UK Home Office project licence 604548 according to the requirements of the Animals (Scientific Procedures) Act 1986 of the UK Government and conformed to Directive 2010/63/EU of the European Parliament. Zebrafish eggs were collected by timed pair mating and incubated in E3 media at 28.5°C in air until 48 hours post fertilisation (hpf). A completed ARRIVE checklist can be found in Supplementary File 2.\n\nHuman melanoma cells A375 (American Type Culture Collection (ATCC), Manassas, USA; RRID, CVCL_0132), as well as C8161 (RRID, CVCL_6813) and WM164 (RRID, CVCL_7928) (generously gifted by Professor Meenhard Herlyn, The Wistar Institute, Philadelphia, USA), or PC-3M-Pro4-mCherry prostate cancer cells (ATCC; RRID, CVCL_D579), were incubated at 33°C for 24 hours to precondition cells prior to staining with 1,1′-Dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate (DiI; Vybrant red fluorescent dye; Invitrogen, Paisley, UK) and injection into zebrafish embryos.\n\nZebrafish embryos at 2 dpf were immobilised using 1.2 mM tricaine methanesulfonate, which is a water soluble, fast-acting anaesthetic agent. Zebrafish embryos were then embedded in a thin film of low-melting-point agarose to stabilise the fish in a lateral position. To investigate invasion of cancer cells from the extravascular compartment into the vasculature, approximately 250 Dil-labelled melanoma cells in a volume of 5 nl were injected into the inferior section of the yolk sac. Similarly to investigate tissue tropism of cancer cells, 250 DiI-labelled prostate cancer cells in a volume of 5 nl were injected into the vein of Cuvier. Following injection, fish were carefully removed from the agarose/tricaine solution using Dumont No5 fine forceps and transferred individually into 96-well plate imaging chambers created from 1% agarose using 3D printed pins (Wittbrodt et al., 2014). Minor orientation was required and embryos were suitable for microscopic analysis within 2 hours of injection.\n\nConfocal images (250 μm total z-depth) of each fish were captured at 0, 24 and 72 hour time points, or every 15 mins for 5 hours for time-lapse imaging, using an inverted Leica SP8 confocal microscope (Leica Microsystems, GmbH Heidelberg, Germany) at 405 nm (blue FluoSpheres), 488 nm (green blood vessels) and 564 nm (red cells). The movement of DiI-positive melanoma cells was analysed using Volocity 3D Image Analysis Software (Volocity 6.3; PerkinElmer, Waltham, Massachusetts, USA) by manually measuring the two-dimensional distance moved by individual melanoma cells from site of injection. The number of RFP-expressing prostate cancer cells was analysed using ImageJ software version 1.8.0_112 (https://imagej.nih.gov/ij) to quantify the total area and intensity of RFP fluorescence.\n\nFor the analysis of tissue-specific homing of prostate cancer cells, 2 dpf zebrafish embryos from a pool of embryos derived from several mated adult zebrafish pairs were randomly assigned to receive an injection of PBS, or an injection of cancer cells, into the vein of Cuvier. The experimental unit is the individual zebrafish embryo, and a sample size of 4 embryos per group was selected on the basis of a normal standard deviation set at 95% confidence level (z = 1.96), a confidence interval (c) of 0.05 and assuming an effect size of 90% (p = 0.9) based on pilot experiments, according to the formula: n = (2z(p)(1-p))/2c. Measurement of total RFP-fluorescence within confocal images was performed and analysed using two-tailed Student’s t-test by a second researcher using GraphPad Prism 7 software (Graph Pad, San Diego, CA USA).\n\n\nResults\n\nThe initial event in metastatic spread is the movement of an individual cancer cell from the tumour niche. This can be modelled in in-vitro systems such as skin organoids or the Dunn chemotactic chamber, but neither of these assays are suitable for measuring metastasis. In our zebrafish embryo xenograft model, we inject small deposits of fluorescently labelled human cancer cells into the yolk sac at 2 days post fertilisation (dpf), and over the next 3 days are able to track individual cells (Figure 1A). By using a zebrafish line with absent pigmentation it is possible to achieve excellent views throughout transgenic embryos with GFP-labelled endothelial blood vessels (green; 510 nm emission), ensuring injection of DiI-labelled A375 melanoma cells (red; 565 nm emission) into the extravascular compartment (Figure 1Bi), which directly migrate to peripheral sites (Figure 1Bii. Although embryos are normally allowed to develop at 28.5°C and human cells at 37°C, a compromise at 33°C works well. Between this time point and the end of the experiment before 120 hpf, the movement of individual melanoma cells from site of injection can be measured using Volocity image analysis software (Figure 1C).\n\nA) Site-specific injection (depicted into the yolk sac) of DiI- or RFP-labelled (Red) cancer cells in 5 nl PBS into 2 dpf zebrafish embryos is followed by incubation of zebrafish for 72 hours at 33°C and subsequent imaging analysis of invasion and metastatic dissemination of cancer cells. B) Approximately 250 DiI-labelled A375 melanoma cells 0 hrs (Bi) and 72 hrs (Bii; white arrows indicate position of melanoma cells) after injection into the yolk sac of Tg(kdrl-GFP) Casper zebrafish (Green blood vessels). C) Confocal z-stack images are used to visualise red DiI fluorescence of melanoma cells within zebrafish (Ci) and the distance from injection site measured using Volocity image analysis software (Cii); Scale bar = 500 μm.\n\nThe ability to carry out time-lapse imaging on embryos affords the opportunity to examine individual cell movement. Injected embryos were lightly anaesthetised using tricaine and orientated in low-melting-point agarose. By focusing on the point of injection, DiI-labelled melanoma cells were visualised moving through the extravesicular compartment within the yolk sac of zebrafish embryos using low-voltage time-lapse confocal microscopy (Figure 2A and Supplementary Movie 1). A 3D-rendering of the confocal image z-stack was rotated to reveal the transverse section of the blood vessel showing a melanoma cell positioned between the zebrafish endothelial cells, indicating that this cell is directly within the blood vessel (Figure 2Ax and Supplementary Movie 2).\n\nAi–ix) Confocal z-stack images taken at 15 minute intervals showing an individual DiI-labelled melanoma cell (white arrows) migrating within the yolk sac of a casper zebrafish embryo and interacting with a GFP-tagged blood vessel. Ax) 3D-render of image Aix rotated to show the transverse section through the GFP-tagged blood vessel with DiI-labelled melanoma cell indicated by white arrows. Bi–v) Confocal z-stack images taken at 15 minute intervals showing an individual DiI-labelled melanoma cell (white arrows) within the GFP-tagged blood vessels of a casper zebrafish embryo. Scale bar = 150 μm.\n\nHaematological or lymphatic metastatic dissemination requires interaction with the endothelium during entry and exit. However, patients can also have cancer cells circulating in their blood that do not necessarily show metastases (Reymond et al., 2013). It is now recognised that metastasising cells exhibit sticking and rolling as they interact with the endothelium, and surface molecules such as selectins and CD44 are implicated. The zebrafish embryo xenograft model shows potential to be an extremely powerful tool in understanding the relationship between the surface biology of tumour and endothelial cells. Time-lapse confocal microscopy at 15-minute intervals readily captured melanoma cells as they demonstrated sticking and rolling behaviours on the surface of vascular endothelium (Figure 2B and Supplementary Movie 3) clearly suggesting a specific interaction of the human melanoma cells with zebrafish endothelial cells.\n\nTumour cells can also be directly injected into the circulation of developing zebrafish via the vein of Cuvier providing a tractable model of metastatic cancer cell-endothelial interaction during vascular exit. This is particularly important in some tumours that do not readily enter the vasculature, but do have tissue tropic exit routes. For example, the prostate cancer cell line PC-3M-Pro4-mCherry does not metastasise from the yolk sac, but when injected into the circulation these cells seed in the caudal hematopoietic tissue of the zebrafish tail where they proliferate, suggesting a specific microenvironmental niche favourable for tumour development (Figure 3A, B).\n\nA) PC-3M-Pro4-mCherry prostate cancer cells injected into the duct of Cuvier form tumours in the caudal hematopoietic tissue of the zebrafish tail; Scale bar = 150 μm. B) Quantification of total mCherry fluorescence by prostate cancer cells after 1 and 3 days post injection; n=4, **p<0.01, 0.05 CI, student’s t-test. Ci–ii) C8161 and Di–ii) WM164 melanoma cells (stained with Red DiI dye) injected alongside FluoSpheres (Blue) into the yolk sac survive and invade throughout the yolk sac; Scale bar = 500 μm.\n\nA specific benefit of embryonic zebrafish over other larger preclinical laboratory models is that several experiments can be carried out in parallel on the same microscope stage. This allows screening of a library of pharmacological candidates, but importantly evaluation of different metastatic cell types, which may be primary cell lines derived directly from patients. This is important as heterogeneity between or within patient tumours may be important in metastatic behaviour. We have seen this in our own melanoma work, where C8161 cells disseminated widely throughout the yolk sac (Figure 3C) while WM164 cells formed a localised tumour-like mass with fewer melanoma cells invading the yolk sac (Figure 3D). Metastatic A375 cells were found in the distal tail vessels, whilst very few C8161 and WM164 cells were found in the tail and other regions of the zebrafish by 72hpf, indicating C8161 and WM164 cells have a reduced capacity to invade blood vessels, which may limit their metastatic potential.\n\nThese vital imaging-based assays used in combination with the ability to genetically modify zebrafish or apply pharmacological agents represent important new tools and approaches to understand these metastatic processes at a cellular level.\n\n\nDiscussion\n\nWe and others have shown that xenotransplantation of human cancer into zebrafish embryos can be optimally carried out from 48-hour post-fertilisation when gastrulation is complete and the main body plan of the animal is established. The next 72 hours provides sufficient time frame to model key stages of metastatic behaviour, including local invasion, vascular entry, circulation and vascular exit. In this paper we have demonstrated how tumour invasion and/or metastatic dissemination by human cancer cells can be monitored through time-lapse microscopy. Most importantly, metastatic processes of single cells can be visualised at the earliest time points, which is not possible in a mouse model. The zebrafish embryonic xenograft of human cancer therefore directly replaces the need for using mouse xenografts and avoids welfare concerns associated with mouse models, including pain and suffering due to unexpected or excessive primary tumour growth. In the UK alone, it is estimated that over 550,000 mice are used each year for cancer research (UK Home Office statistics). On average 50 mice are used per study of cancer metastasis, and over the past 5 years there have been on average 900 publications per year in this area. We therefore estimate that 45,000 mice are used each year for research of cancer metastasis using mouse xenograft models, many of which could be replaced by embryonic zebrafish at unregulated stages of development, using the model described in this paper. To do this several historical concerns need to be addressed.\n\nExperimentally, it is essential that following xenotransplantation human cells can be distinguished from host cells of the zebrafish. Whilst this can be achieved post-mortem by detecting human-specific antigens using immunocytochemistry (Bentley et al., 2015), the use of lipophilic fluorescent cell membrane stains in conjunction with zebrafish transgenic lines allows much more data to be collected from animals of greater immaturity. However, quantification of xenografted cancer cell proliferation is equivalent when measured using either membrane stains or fluorescent proteins (Bentley et al., 2015).\n\nThere are concerns that differences in cell size, microenvironmental niches and molecular signalling pathways between human patients and preclinical models (mice as well as zebrafish) could limit the relevance and translational value of data obtained from animal studies. However, our studies show that human cancer cells are able to invade zebrafish blood vessels and form secondary tumours; while previous studies have shown that VEGF and CXCR4 signalling are conserved between human cancers and zebrafish (He et al., 2012; Tulotta et al., 2016). Nevertheless, further studies to characterise the response of human cells in the zebrafish model organism are required.\n\nThe future direction of research using zebrafish embryos for human xenograft studies will likely focus on strategies and methods to increase assay throughput and improve analysis of large data sets. These objectives will benefit from a number of technical innovations, such as devices to orientate the zebrafish for imaging (Wittbrodt et al., 2014) as well as automated quantification and analysis of tumour cell dissemination (Ghotra et al., 2012; Heilmann et al., 2015).\n\nStable cancer cell lines are often dramatically different from patient tumour cells and by definition have been selected for ease of maintenance in the laboratory environment. However, it is likely that heterogeneity and cooperation of cancer cells in patient tumours drives tumour invasion through remodelling of the extracellular matrix (Chapman et al., 2014). Thus, cancer is represented by cells that vary in their proliferative, invasive and metastatic phenotype, which contributes both to tumour growth and also emergence of drug resistance (Anderson et al., 2011). However, it is often not feasible to investigate the effect of tumour cell heterogeneity in mouse xenograft models as large numbers of patient primary tumour cells are required for successful engraftment. In contrast, a major advantage of the embryonic zebrafish xenograft assay is the capacity to accurately detect and monitor a small number of cells (100–200 cells per fish), including low-number cancer subpopulations such as cancer stem cells, drug-resistant cells or primary patient tumour tissue where only small numbers of cells can be recovered e.g. circulating tumour cells. Patient-derived tumour xenograft models are therefore a potential solution to the problem of limited intratumoural heterogeneity of cell line derived xenografts, which may improve the accuracy of tumour drug-response studies.\n\nIt is becoming increasingly clear that no single pre-clinical model can substitute for actual human trials, and therefore as researchers we must continually reassess and adapt our model assays to improve their relevance, which will likely involve employing an approach that combines multicellular in vitro organoid assays (Hill et al., 2015) with both zebrafish and mouse in vivo studies. We suggest that this combinatorial approach will reduce the reliance on mouse xenograft models for the study of human cancer metastasis and drug screening. However, the challenge for translational cancer research will be to integrate the multitude of data from different model organisms to identify evolutionary conserved drug-tumour interactions between species so that we may select the most appropriate therapeutics that have the highest chance of providing an effective treatment for patients with cancer.\n\n\nData availability\n\nDataset 1. Raw images used to generate figures shown in this study. Shown are images for Figure 1 and Figure 3; images in Figure 2 were obtained from stills of Supplementary Movie 1–Supplementary Movie 3. DOI: https://doi.org/10.5256/f1000research.16659.d221978 (Hill et al., 2018).", "appendix": "Grant information\n\nThis work was supported by grant funding from the National Centre for the Replacement, Reduction and Refinement of Animals in Research (NC3Rs; NC/L002000/1), and from Alpe D'HuZes (AdH)/KWF PROPER (UL2014-7058).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary material\n\nSupplementary File 1. Complete step-by-step protocol for Zebrafish xenograft of human cancer cells.\n\nClick here to access the data\n\nSupplementary File 2. Completed ARRIVE checklist.\n\nClick here to access the data\n\nSupplementary Movie 1. Representative time lapse confocal movie showing active invasion of a DiI-labelled A375 melanoma cell through the yolk sac and into a kdrl-GFP labelled blood vessel of a Casper zebrafish embryo.\n\nImages were taken every 15 minutes for 5 hours.\n\nClick here to access the data\n\nSupplementary Movie 2. 3D rendering of confocal movie showing active invasion of a DiI-labelled A375 melanoma cell through the yolk sac and into a kdrl-GFP labelled blood vessel of a Casper zebrafish embryo.\n\nTransverse field of view through the blood vessel demonstrates that the red melanoma is within the green blood vessel and actively moving against the flow of blood.\n\nClick here to access the data\n\nSupplementary Movie 3. Representative time lapse confocal movie showing movement of a DiI-labelled A375 melanoma cell through a kdrl-GFP labelled blood vessel of a Casper zebrafish embryo.\n\nImages were taken every 15 minutes for 5 hours.\n\nClick here to access the data\n\n\nReferences\n\nAkakura K, Akimoto S, Shimazaki J: Pain caused by bone metastasis in endocrine-therapy-refractory prostate cancer. J Cancer Res Clin Oncol. 1996; 122(10): 633–7. PubMed Abstract \| Publisher Full Text\n\nAnderson AC, Pollastri MP, Schiffer CA, et al.: The challenge of developing robust drugs to overcome resistance. Drug Discov Today. 2011; 16(17–18): 755–61. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBansal N, Davis S, Tereshchenko I, et al.: Enrichment of human prostate cancer cells with tumor initiating properties in mouse and zebrafish xenografts by differential adhesion. Prostate. 2014; 74(2): 187–200. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBentley VL, Veinotte CJ, Corkery DP, et al.: Focused chemical genomics using zebrafish xenotransplantation as a pre-clinical therapeutic platform for T-cell acute lymphoblastic leukemia. Haematologica. 2015; 100(1): 70–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBos PD, Zhang XH, Nadal C, et al.: Genes that mediate breast cancer metastasis to the brain. Nature. 2009; 459(7249): 1005–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChapman A, Fernandez del Ama L, Ferguson J, et al.: Heterogeneous tumor subpopulations cooperate to drive invasion. Cell Rep. 2014; 8(3): 688–95. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChen L, Groenewoud A, Tulotta C, et al.: A zebrafish xenograft model for studying human cancer stem cells in distant metastasis and therapy response. Methods Cell Biol. 2017; 138: 471–96. PubMed Abstract \| Publisher Full Text\n\nChu Y, Tang H, Guo Y, et al.: Adipose-derived mesenchymal stem cells promote cell proliferation and invasion of epithelial ovarian cancer. Exp Cell Res. 2015; 337(1): 16–27. PubMed Abstract \| Publisher Full Text\n\nCorkery DP, Dellaire G, Berman JN: Leukaemia xenotransplantation in zebrafish--chemotherapy response assay in vivo. Br J Haematol. 2011; 153(6): 786–9. PubMed Abstract \| Publisher Full Text\n\nCruz-Munoz W, Man S, Xu P, et al.: Development of a preclinical model of spontaneous human melanoma central nervous system metastasis. Cancer Res. 2008; 68(12): 4500–5. PubMed Abstract \| Publisher Full Text\n\nDay CP, Merlino G, Van Dyke T: Preclinical mouse cancer models: a maze of opportunities and challenges. Cell. 2015; 163(1): 39–53. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nde Boeck M, Cui C, Mulder AA, et al.: Smad6 determines BMP-regulated invasive behaviour of breast cancer cells in a zebrafish xenograft model. Sci Rep. 2016; 6: 24968. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nElkin M, Vlodavsky I: Tail vein assay of cancer metastasis. Curr Protoc Cell Biol. 2001; Chapter 19: Unit 19.2. PubMed Abstract \| Publisher Full Text\n\nFeng Y, Martin P: Imaging innate immune responses at tumour initiation: new insights from fish and flies. Nat Rev Cancer. 2015; 15(9): 556–62. PubMed Abstract \| Publisher Full Text\n\nFior R, Póvoa V, Mendes RV, et al.: Single-cell functional and chemosensitive profiling of combinatorial colorectal therapy in zebrafish xenografts. Proc Natl Acad Sci U S A. 2017; 114(39): E8234–e43. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFisher JL, Schmitt JF, Howard ML, et al.: An in vivo model of prostate carcinoma growth and invasion in bone. Cell Tissue Res. 2002; 307(3): 337–45. PubMed Abstract \| Publisher Full Text\n\nGershenwald JE, Scolyer RA: Melanoma Staging: American Joint Committee on Cancer (AJCC) 8th Edition and Beyond. Ann Surg Oncol. 2018; 25(8): 2105–10. PubMed Abstract \| Publisher Full Text\n\nGhotra VP, He S, de Bont H, et al.: Automated whole animal bio-imaging assay for human cancer dissemination. PLoS One. 2012; 7(2): e31281. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGibert Y, Trengove MC, Ward AC: Zebrafish as a genetic model in pre-clinical drug testing and screening. Curr Med Chem. 2013; 20(19): 2458–66. PubMed Abstract \| Publisher Full Text\n\nGómez-Cuadrado L, Tracey N, Ma R, et al.: Mouse models of metastasis: progress and prospects. Dis Model Mech. 2017; 10(9): 1061–74. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHaldi M, Ton C, Seng WL, et al.: Human melanoma cells transplanted into zebrafish proliferate, migrate, produce melanin, form masses and stimulate angiogenesis in zebrafish. Angiogenesis. 2006; 9(3): 139–51. PubMed Abstract \| Publisher Full Text\n\nHe S, Lamers GE, Beenakker JW, et al.: Neutrophil-mediated experimental metastasis is enhanced by VEGFR inhibition in a zebrafish xenograft model. J Pathol. 2012; 227(4): 431–45. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHeilmann S, Ratnakumar K, Langdon E, et al.: A Quantitative System for Studying Metastasis Using Transparent Zebrafish. Cancer Res. 2015; 75(20): 4272–82. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHill DS, Robinson ND, Caley MP, et al.: A Novel Fully Humanized 3D Skin Equivalent to Model Early Melanoma Invasion. Mol Cancer Ther. 2015; 14(11): 2665–73. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHill D, Chen L, Snaar-Jagalska E, et al.: Dataset 1 in: Embryonic zebrafish xenograft assay of human cancer metastasis. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16659.d221978\n\nHowe K, Clark MD, Torroja CF, et al.: The zebrafish reference genome sequence and its relationship to the human genome. Nature. 2013; 496(7446): 498–503. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJemal A, Bray F, Center MM, et al.: Global cancer statistics. CA Cancer J Clin. 2011; 61(2): 69–90. PubMed Abstract \| Publisher Full Text\n\nKang Y, Siegel PM, Shu W, et al.: A multigenic program mediating breast cancer metastasis to bone. Cancer Cell. 2003; 3(6): 537–49. PubMed Abstract \| Publisher Full Text\n\nKersten K, de Visser KE, van Miltenburg MH, et al.: Genetically engineered mouse models in oncology research and cancer medicine. EMBO Mol Med. 2017; 9(2): 137–53. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLam SH, Chua HL, Gong Z, et al.: Development and maturation of the immune system in zebrafish, Danio rerio: a gene expression profiling, in situ hybridization and immunological study. Dev Comp Immunol. 2004; 28(1): 9–28. PubMed Abstract \| Publisher Full Text\n\nLee LM, Seftor EA, Bonde G, et al.: The fate of human malignant melanoma cells transplanted into zebrafish embryos: assessment of migration and cell division in the absence of tumor formation. Dev Dyn. 2005; 233(4): 1560–70. PubMed Abstract \| Publisher Full Text\n\nMarques IJ, Weiss FU, Vlecken DH, et al.: Metastatic behaviour of primary human tumours in a zebrafish xenotransplantation model. BMC cancer. 2009; 9: 128. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMassagué J, Batlle E, Gomis RR: Understanding the molecular mechanisms driving metastasis. Mol Oncol. 2017; 11(1): 3–4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMinn AJ, Gupta GP, Siegel PM, et al.: Genes that mediate breast cancer metastasis to lung. Nature. 2005; 436(7050): 518–24. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMorikawa K, Walker SM, Nakajima M, et al.: Influence of organ environment on the growth, selection, and metastasis of human colon carcinoma cells in nude mice. Cancer Res. 1988; 48(23): 6863–71. PubMed Abstract\n\nPandya P, Orgaz JL, Sanz-Moreno V: Modes of invasion during tumour dissemination. Mol Oncol. 2017; 11(1): 5–27. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPenberthy WT, Shafizadeh E, Lin S: The zebrafish as a model for human disease. Front Biosci. 2002; 7: d1439–53. PubMed Abstract\n\nReymond N, d'Agua BB, Ridley AJ: Crossing the endothelial barrier during metastasis. Nat Rev Cancer. 2013; 13(12): 858–70. PubMed Abstract \| Publisher Full Text\n\nSrivastava K, Hu J, Korn C, et al.: Postsurgical adjuvant tumor therapy by combining anti-angiopoietin-2 and metronomic chemotherapy limits metastatic growth. Cancer Cell. 2014; 26(6): 880–95. PubMed Abstract \| Publisher Full Text\n\nTarin D: Cell and tissue interactions in carcinogenesis and metastasis and their clinical significance. Semin Cancer Biol. 2011; 21(2): 72–82. PubMed Abstract \| Publisher Full Text\n\nTulotta C, Stefanescu C, Beletkaia E, et al.: Inhibition of signaling between human CXCR4 and zebrafish ligands by the small molecule IT1t impairs the formation of triple-negative breast cancer early metastases in a zebrafish xenograft model. Dis Model Mech. 2016; 9(2): 141–53. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nvan Marion DM, Domanska UM, Timmer-Bosscha H, et al.: Studying cancer metastasis: Existing models, challenges and future perspectives. Crit Rev Oncol Hematol. 2016; 97: 107–17. PubMed Abstract \| Publisher Full Text\n\nWang H, Yu C, Gao X, et al.: The osteogenic niche promotes early-stage bone colonization of disseminated breast cancer cells. Cancer Cell. 2015; 27(2): 193–210. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWesternfield M: A Guide for the Laboratory Use of Zebrafish (Danio Rerio). 4th Edition. 2000. Reference Source\n\nWittbrodt JN, Liebel U, Gehrig J: Generation of orientation tools for automated zebrafish screening assays using desktop 3D printing. BMC Biotechnol. 2014; 14: 36. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "39771", "date": "06 Nov 2018", "name": "Adam Hurlstone", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI enjoyed reading this methodology paper which lays out clearly steps required to achieve successful xenotransplantation in zebrafish embryos and how subsequent growth and dissemination of cancer cells can be monitored.\n\nThe attached method protocol would be improved by inclusion of the below details:\nExplain how to remove embryos from agarose using forceps Indicate an appropriate model of microtitre plate for imaging purposes with an inverted microscope (does imaging require glass bottom plates or a certain grade of plasticware?) Specify where they obtained the plastic pin mold: have they manufactured it, requested it, or purchased it? Comment on whether 72 h incubation in anesthetic is detrimental to embryo health/development Which image analysis modules/tools were used in Velocity and Image J. Mention whether default parameters were selected or otherwise? Why 2D rather than 3D distances were calculated using Velocity? Velocity is relatively expensive proprietary software and may not therefore be widely accessible, whereas Image J is free. Could the whole analysis not be undertaken with Image J? Specify an appropriate method of ensuring destruction of the embryos within 120 hpf\n\nTurning to the rest of the manuscript:\nExplain what metrics would be captured by the analysis depicted in Fig 1B and present a representative graph. Mean/median migration distance? Is there a way of distinguishing between several small clusters of cells or a few larger ones? Specify the cell line used for Fig 2. How efficient is the model for capturing intravasation events? How many cells are captured intravasating per hour per embryo? The presentation and analysis of data in Fig 3B is inappropriate as these are not independent populations of cells. A line graph and linear regression is the appropriate analysis. Does proliferation contribute to the expansion of cells in the caudal hematopoietic tissue? It would be of value to include a statement describing the distribution of fluospheres injected either into the yolk or into the duct of cuvier\n\nAre a suitable application and appropriate end-users identified? Yes\nIf applicable, is the statistical analysis and its interpretation appropriate? Partly\nAre the 3Rs implications of the work described accurately? Yes\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Partly\n\nAre sufficient details provided to allow replication of the method development and its use by others? Partly\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4266", "date": "20 Dec 2018", "name": "David Hill", "role": "Author Response", "response": "The authors would like to thanks Dr Hurlstone for his thorough and inciteful review of our article. We have made suggested changes to clarify methodology and correct typographical errors in our article, including details of consumables and of pin-mould manufacture. Importantly, applying forceps to break the agarose away from the embryo allows its release into surrounding media. Toxicity of tricaine was not observed in our studies, but if there are concerns it should be excluded with a specific experimental control. Embryos were killed using a schedule 1 method (destruction of the brain). However, for post-mortem histological analysis, cooling and fixation in 4% PFA was also used. Whilst we used commercially available software, we also recommend use of the Fiji implementation of imageJ (https://fiji.sc/), which contains tools for measurement and tracking. The cell line used for Figure 2 was the parental A375 human melanoma cell line. We have modified the figure legend to reflect this. We have updated our analysis of Figure 3B to reflect that the same cell populations are measured at two time points by using a paired t-test rather than a student’s t-test, and have updated Figure 3 and the legend for Figure 3 accordingly. We have also included a reference (Verykiou et al., 2018) in the main text. We have used this method of analysis to measure the distance invaded by MEKi-resistant A375 melanoma cells. The use of nuclear-localised fluorescent proteins allow individual cells within a cluster to be distinguished, while the use of membrane dyes are ideal for analysis of primary tumour cells, low-number tumour subpopulations and transient events such as intra/extravasation or interaction of tumour cells with host cells and stroma." } ] }, { "id": "39773", "date": "12 Nov 2018", "name": "Yi Feng", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nZebrafish embryo xenograft model is a widely adopted method to study many aspects of cancer cell behaviour in vivo. A variety of protocols that differ by injection site, incubating temperature and down stream phenotype chosen, have been used in the field. So far, there is still a lack of a publicly available, detailed standard protocol, which new users could follow. The current methodology paper provides a very detailed, easy to follow protocol for xenograft fluorescent labeled cancer cells in the yolk sac and vein of Cuvier in 2 days old embryos and how to monitor cancer cell dissemination following the graft to assess their metastasis potential. This paper would be of benefit to anyone who wants to try out the zebrafish embryo xenograft model. The step by step protocol could include more details that would make it more user friendly for someone who is new to fish models.\nMaterial section:\nIt would help if more details were given for the following reagents such as Cat number because there could be multiple products available under a similar name: Low-melting point agarose, Borosilicate glass capillaries, Ultrafine forceps. What size?, 96-well plate. Are these with a glass bottom? 3D printed mould to create imaging chambers – this is referred to in another paper without any details given. Perhaps they could expand on how they made it or obtained it? The imaging chamber is a key aspect and for this protocol to be “high through-put” more details would really help the reader.\nProtocol steps\nOnce embryos transferred into imaging chamber, the authors indicate that the embryos will be maintained in 150 ml 1X E3 media containing 1.2 mM tricaine for 72 hours. This could have detrimental effects on embryos. The authors should comment on whether any adverse effects were observed and how to avoid them. One of the Optional steps: “Tracer beads can be used to label the original injection site and to distinguish active tumour cell migration from passive development associated movement that occurs when tissues and organs within the yolk sac grow.” To my mind it is very important for the initial set up of the model, as different cancer cell lines display different metastasis capacities and the tracer beads can be found expanding from the original injection site due to various reasons that the authors pointed out. So the use of tracer bead would help to set up a baseline index for passive expansion, true cancer cell invasion and metastasis that can then be evaluated according to the baseline parameter. The authors could perhaps comment more on their experience of using tracer beads in the main text of the paper.\nThe main text of the paper\nIn the main text the authors nicely presented three examples of what biological features can be captured using confocal imaging analysis following graft. It would be very helpful if they could elaborate more on each of the models, presenting more details on what parameters could be established from each model. Model 1 (figure 1 Cii) cancer cell dissemination. It seems that 2D distance is used instead of 3D and there is no mention of the size of each cell cluster that appeared to be metastatic growth. The authors should explain more extensively why they choose such a parameter and whether there are other potential parameters that one could measure to assess dissemination of cancer cells.\nModel 2 (figure 2) intravasation and distal metastasis are extremely rare events according to other publications (Roh-Johnson M, et al), perhaps authors could comment on how frequently they can capture intravasation or cancer cells within blood vessels? Perhaps provide some information on their experience with different cell lines in their intravasation capacity.\nModel 3 (Figure 3 A,B) injecting into the vein of Cuvier is similar to mouse tail vein injection where cancer cells are grafted directly into the blood stream. This allows for study of the capacity of cancer cell extravasion and proliferation in distal tissues to establish metastic growth. Data presented in (Figure 3 A, B) using fluorescent intensity as read out for cancer cell proliferation (same for C, D), which is quite a crude way of quantification. I wonder if it is possible to use more precise methods such as EdU incorporation or pH3 staining, ki67 staining or PCNA staining? Perhaps authors could share their experiences of make some comments on other ways of evaluating the cancer cell proliferation in vivo after xenograft.\n\nThe authors focused on using the zebrafish xenograft models for metastasis analysis. Angiogenesis was one the first assays developed using zebrafish embryo xenograft model (Nicoli S, et al). As a methodology, the protocol presented here can be adapted for angiogenesis analysis. Perhaps they could comment on how their protocol could be adapted for evaluating angiogenesis in vivo. There are new developments of the zebrafish embryo xenograft model for angiogenesis such as (Britto DD, et al) perhaps they could refer to this work in the introduction or discussion, so as to guide the reader to other and more specific examples.\n\nAre a suitable application and appropriate end-users identified? Yes\nIf applicable, is the statistical analysis and its interpretation appropriate? Partly\nAre the 3Rs implications of the work described accurately? Yes\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Partly\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4265", "date": "20 Dec 2018", "name": "David Hill", "role": "Author Response", "response": "The authors would like to thanks Dr Feng for her insightful comments and suggestions. We detail below the additions and changes we have made specifically in response to this review: We have incorporated all changes as suggested including: catalogue numbers of critical reagents and reference to angiogenesis; and mention of alternative post-mortem analyses for cell proliferation. Whilst we have only observed local movement of tracer beads, we have emphasised the importance of this quality control in the main manuscript. For these studies we chose to analyse movements in a single plane and were able to find significant differences. However, with increased imaging time and use of lenses with limited depth of field, cells can be tracked in three dimensions. Caution should be employed as the light exposure in obtaining such image stacks may affect cell behaviour. The capability of cancer cells to invade blood vessels was cell line dependent, for example, more than 80% of embryos injected with A375 cells had cancer cells within the blood vessels by 72 hours, while C8161 and WM164 cells invaded blood vessels less frequently and also showed variable local movement. These preliminary studies indicate the utility of the zebrafish at pre-regulated embryonic stages to study key aspects of metastatic cancer spread." } ] } ]	1	https://f1000research.com/articles/7-1682
https://f1000research.com/articles/7-1817/v1	20 Nov 18	{ "type": "Research Article", "title": "Do age and mating status affect olfactory response of the parasitoid, Microplitis croceipes (Hymenoptera: Braconidae) to host-related plant odors?", "authors": [ "Matthew Burrows", "Tolulope Morawo", "Henry Fadamiro", "Matthew Burrows", "Henry Fadamiro" ], "abstract": "Background: Parasitic wasps (parasitoids) use volatile organic compounds released by herbivore-infested plants to locate their hosts. Response of parasitoids to plant odors may be plastic and dependent on their physiological state. Using Microplitis croceipes (Hymenoptera: Braconidae), a relatively specialized larval endoparasitoid of Heliothis virescens (Lepidoptera: Noctuidae), we asked whether age and mating status of parasitoids affect their olfactory response to host-related odors. Methods: Four odor stimuli of varying complexity were selected based on previous reports of parasitoid response to cotton volatiles: cis-3-hexenol (a green leaf volatile), α-pinene (a constitutive monoterpene), a 50/50 v/v binary mixture (cis-3-hexenol + α-pinene), and H. virescens-infested cotton odors. Female M. croceipes used in Y-tube olfactometer bioassays were either mated or unmated, and grouped 1–3, 4–6, and 7–9 d-old. Female parasitoids used in electroantennogram (EAG) recording were mated and grouped 1–3, 4–6, 7–9 and 10–12 d-old. Results: In Y-tube olfactometer bioassays, neither age nor mating status played a major role in the attraction of parasitoids to test odor stimuli, with two exceptions: 4–6 d-old mated parasitoids showed attraction to the binary mixture, and 1–3 d-old mated parasitoids showed attraction to H. virescens-infested cotton. Age did not affect EAG response of parasitoids to test stimuli. Conclusions: The present results suggest that age and mating status do not play a major role in modulating olfactory responses of M. croceipes to host-related plant odors. Instead, plasticity of olfactory response may be limited in M. croceipes due to strong innate sensitivity to host-related odor cues.", "keywords": [ "cotton", "electroantennogram", "Heliothis virescens", "physiological state", "Y-tube olfactometer" ], "content": "Introduction\n\nParasitic wasps (parasitoids) use volatile organic compounds (VOCs) released by herbivore-infested plants as odor cues to locate their hosts1–4. However, the level of olfactory response to VOCs may not remain consistent throughout the duration of parasitoid adult life. Changes in physiological states can affect response to VOCs in parasitoids and other insects5–9. Olfactory plasticity in insects has been previously attributed to changes in physiological states such as age, mating, and nutritional status6,9–13.\n\nAging has been shown to affect olfaction in insects9,14–16. This effect may be associated with senescence of olfactory structures and/or processing units in insects. For instance, behavioral senescence of responses involved with locomotion, olfaction, and learning has been reported in fruit flies17–19. However, previous studies have reported mixed results regarding age-related plasticity of response to VOCs in parasitoids, thus the need for further studies14,20,21. Mating status of female parasitoids has also been reported to influence their foraging behavior and parasitization potential, with mated females showing a higher parasitization rate than unmated females13,22,23. However, the effect of mating on plasticity of parasitoid response to host-related plant volatiles has gained little attention despite its implications and relevance to host searching and parasitization potential (but see Chen & Fadamiro24).\n\nIn the present study, the relatively specialized larval endoparasitoid, Microplitis croceipes (Hymenoptera: Braconidae) and its caterpillar host, Heliothis virescens (Lepidoptera: Noctuidae) were used as a study system to test the effect of age and mating status on the olfactory response of parasitoids to host-related plant volatiles. Heliothis virescens is a generalist herbivore and a serious pest of cotton, tobacco, and other crops of economic importance25. Based on previous reports on olfactory response in M. croceipes to cotton volatiles26–30, four odor stimuli of varying complexity were chosen. These include: cis-3-hexenol (a green leaf volatile), α-pinene (a constitutive monoterpene), a 50/50 v/v binary mixture (cis-3-hexenol + α-pinene), and headspace volatiles from H. virescens-infested cotton. cis-3-Hexenol and α-pinene have been consistently detected in the headspace of H. virescens-infested cotton1,29,31 and have been reported to elicit antennal and behavioral responses in M. croceipes24,26,29,30.\n\nY-tube olfactometer was used to test attraction (behavioral response) of parasitoids while electroantennogram (EAG) was used to record antennal response of female M. croceipes to select host-related cotton volatiles. To the best of our knowledge, this is one of the few studies that investigated the effects of age and mating status on olfactory responses of parasitoids to host-related plant volatiles. The implications of these findings are discussed.\n\n\nMethods\n\nCocoons of Microplitis croceipes were provided by the USDA-ARS, Insect Biology and Population Management Research Laboratory (Tifton, Georgia, USA) and reared in our laboratory (Auburn University, AL, USA) on 2nd – 3rd instar larvae of H. virescens. Upon emergence, adult wasps were transferred to aerated plastic BugDorm® cages (Megaview Science Co. Taichung, Taiwan) and supplied with 10% sucrose/water solution (w/v). Naive (untrained) parasitoids were used in both EAG and Y-tube olfactometer bioassays to test innate responses of parasitoids to host-related plant odors. Eggs of H. virescens were initially purchased from Benzon Research Inc. (Carlislie, PA, USA) and reared in our laboratory at Auburn University. Larvae of H. virescens were reared on pinto bean artificial diet according to Shorey and Hale32. In total, about 640 female parasitoids were used for bioassays. The general rearing conditions for all insects were 25 ± 1°C, 75 ± 5 % RH and L14:D10 h (L:D) photoperiod.\n\nCotton (Gossypium hirsutum, var. max 9, All-Tex Seed Inc., Levelland, TX, USA) plants were grown in individual pots (9 cm high, 11 cm diameter) in a growth chamber at 26.6°C day, 25.6°C night, 60% RH, L16:D8 h (L:D) photoperiod. Seeds were planted in a top soil/vermiculate mixture. Plants used for headspace volatile collections were 4–6 weeks-old.\n\nUpon emergence, parasitoids were separated based on their sex. Subsequently, an equal number of females were randomly designated to ‘mated’ or ‘unmated’ cages. Male parasitoids were put inside cages (19 × 13 × 10 cm) designated to mated females at a 2:1 (male: female) ratio while cages designated to unmated females contained females only. Female parasitoids were allowed to mate for at least 24 h before use in bioassays. Both mated and unmated females were further designated to separate cages based on their age. Age groups 1–3, 4–6 and 7–9 d-old were used in Y-tube olfactometer bioassays while age groups 1–3, 4–6, 7–9, and 10–12 d-old were used in electroantennogram (EAG) recording. The 10-12 d-old age group was not included in behavioral bioassays due to low numbers of insects surviving beyond 10 days in the laboratory, resulting in a low number of replicates. It should be noted that Y-tube olfactometer bioassays required more replicates than EAG recording.\n\nHeadspace volatiles were collected from H. virescens-infested cotton plants using the protocol described by Ngumbi et al.31. A plant pot with soil was wrapped with aluminum foil to minimize contamination. The plant was then placed in a volatile collection chamber (Analytical Research Systems, Inc., Gainesville, FL, USA) consisting of a 5-L air-tight glass jar. A purified air stream of 500 ml/min was passed through the jar at room temperature using Teflon tubing connected to an air delivery system. To induce volatiles from plants, 30 2nd –3rd instar larvae of H. virescens were allowed to feed on a cotton plant for 24 h during volatile collection from 1100 h one day to 1100 h of the following day. Headspace volatiles were collected with a trap containing 50 mg of Super-Q (CAT#: 2735, Alltech Associates, Deerfield, IL, USA) and eluted with 300 µl of methylene chloride. The resulting extract was stored in a freezer (-20°C) until use.\n\nA Y-tube olfactometer (Analytical Research Systems, Inc., Gainesville, FL, USA) was used to test attraction of female M. croceipes to four odor stimuli of varying complexity. The setup and procedure was similar to that reported by Morawo and Fadamiro33. Parasitoids were introduced individually into the olfactometer and allowed to make a choice between test stimulus and control. Insects were tested once and discarded. Parasitoids that made no choice within 5 min were removed and excluded from the analyses. The number of non-responding parasitoids in the 24 sets of bioassays ranged from 0 to 5 with a mean of 1.2 insects per test.\n\ncis-3-Hexenol (CAT#: W256307) and α-pinene (CAT#: 147524) (purity 95-99%; Sigma-Aldrich®, St. Louis, MO, USA) were individually formulated in hexane (HPLC-grade) at 1 µg/µl concentration. A 50/50 v/v binary mixture (cis-3-Hexenol + α-pinene) of the two compounds was also prepared. A central dose of 10 µg (10 µl sample) was previously determined to be optimal in a related study30. In separate bioassays, each compound or binary mixture was delivered as a 10 µl sample on filter paper strips (40 × 7 mm, CAT#: 1001090, Whatman® No. 1) in the treatment arm while the control arm contained the same volume of hexane (solvent control). Humidified and purified (charcoal filtered) air was pushed into each arm at the rate of 250 ml/min and removed by suction from the central arm of the olfactometer at the rate of 500 ml/min to avoid odor mix-up.\n\nTo test parasitoid response to H. virescens-infested cotton, one arm of the olfactometer was connected to an air-tight glass jar (5-L) containing an infested plant (with host larvae). The other arm was connected to a similar glass jar containing a pot of soil covered with aluminum foil, which served as control. A new plant was used on different days during which bioassays were conducted (5–7 plant replicates). Inlet air was pushed into the olfactometer through each jar at the rate of 300 ml/min and sucked out at the rate of 600 ml/min. Experiments were performed in a randomized complete block design with equal number of insect replicates from each age and mating status groups tested per day (n = 20 per test). All olfactometer bioassays were conducted between 1100 h and 1700 h on different days.\n\nEAG response of female M. croceipes was recorded to measure odor perception in mated parasitoids. The EAG protocol used was previously described by Ngumbi et al.31. Glass capillaries (1.1 mm I.D.) filled with Ringer solution served as reference and recording electrodes. The reference electrode was connected to the back of the head of a female M. croceipes while the recording electrode was connected to the cut tip of the terminal segment of the antenna. The analog signal was detected through a probe (INR-II, Syntech, the Netherlands), and was captured and processed with a data acquisition controller (IDAC-4, Syntech, the Netherlands). EAG 2000 software v2.7 (Syntech, the Netherlands) was used to analyze digital signal readouts. Test stimuli were delivered as 10-µl samples (10 µg dose) on filter paper strips (7 × 40 mm) placed inside 14 cm Pasteur pipettes (Fisher Scientific, Pittsburgh, PA, USA).\n\nFour treatment stimuli and two control stimuli were individually delivered as 0.2-s puffs of air, with 2 min interval between puffs. For each antennal preparation, the following stimuli were presented: hexane (control), methylene chloride (control), cis-3-hexenol, α-pinene, binary mixture, headspace volatile extract, hexane and methylene chloride. Thus, hexane and methylene chloride (solvent controls) were applied at the beginning and end of each recording series while the position of other test stimuli was randomized across replicates (see Morawo et al.34). Recordings were performed in a randomized complete block design with equal number of insect replicates (n = 10) from each age group tested per day.\n\nAttraction of parasitoids to each of four test stimuli in Y-tube olfactometer was modeled as a binary response (stimulus = 1, control = 0) using logistic regression to analyze possible interactions between age and mating status factors. The model adequacy for each set of experiment was confirmed with a likelihood ratio test35. When no significant interaction was recorded, each factor was analyzed separately. For olfactometer data, deviation of parasitoid responses from a 50:50 (stimulus: control) distribution was analyzed using a Chi-square goodness-of-fit test. Absolute EAG responses (EAG response to solvent control deducted from EAG response to test stimuli) of mated parasitoids across age groups were compared using Kruskal-Wallis test. All analyses were performed in SAS v9.2 (SAS Institute Inc., Cary, NC, USA) with P = 0.05 level of significance.\n\n\nResults\n\nOverall, there was no significant interaction between age and mating status factors for any of the four odor stimuli tested (cis-3-hexenol: P = 0.7295, Figure 1A; α-pinene: P = 0.7352, Figure 1B; binary mixture: P = 0.1136, Figure 1C; host-infested cotton: P = 0.7044, Figure 1D; Logistic Regression). However, mated parasitoids, 4-6 d-old were significantly (80/20%, χ2= 7.20, df = 1, P =0.0073) more attracted to the binary mixture (cis-3-hexenol + α-pinene) compared to hexane control (Figure 1C). Similarly, 1-3 d-old mated parasitoids were significantly (75/25%, χ2= 5.00, df = 1, P =0.0253) more attracted to H. virescens-infested cotton compared to control (Figure 1D). Parasitoids did not show significant attraction to test stimuli in other Y-tube olfactometer bioassays.\n\nBars represent percentage of mated and unmated parasitoids of ages 1–3, 4–6, and 7–9 d-old when given a choice between hexane (solvent control) and synthetic compounds cis-3-hexenol (A), α-pinene (B), a 50/50 v/v binary mixture of cis-3-hexenol and α-pinene (C), and a choice between control jar (with no plant) and Heliothis virescens-infested cotton (D). Synthetic compounds were formulated in hexane at 1 µg/ µl and presented as 10 µl samples (10 µg dose). Thirty 2nd–3rd instar larvae of H. virescens were allowed to infest cotton plants for 24 h before bioassays. N = 20 responding parasitoids per choice test. Numbers in the bars indicate actual number of responding individuals that chose each arm of the olfactometer. Asterisk (*) indicates significant deviation of parasitoid responses from a 50:50 (stimulus: control) distribution (χ2 goodness of fit test, P < 0.05).\n\nIn general, the age of mated female M. croceipes did not have a significant effect on their EAG response to test odor stimuli (Figure 2). Both relatively young and older parasitoids showed similar levels of innate antennal sensitivity to single components, binary mixture and headspace volatile extract of host-infested cotton. These results are mostly in agreement with those recorded in Y-tube olfactometer bioassays.\n\nBars represent mean absolute EAG responses (mV± SE, N = 10) of mated parasitoids age 1–3, 4–6, 7–9 and 10–12 d-old to cis-3-hexenol (A), α-pinene (B), a 50/50 v/v binary mixture of cis-3-hexenol and α-pinene (C), and Heliothis virescens-infested cotton headspace volatile extract (D). Absolute EAG for each stimulus is the actual EAG value minus EAG value of solvent control. Synthetic compounds were formulated in hexane at 1 µg/ µl and presented as 10 µl samples (10 µg dose). NS indicates no significant difference.\n\n\nDiscussion\n\nAge and mating status are among several physiological factors that may affect olfactory responses of parasitoids to odor cues used in foraging. In the present study, neither age nor mating status of female M. croceipes played a major role in their olfactory responses to host-related plant volatiles. In general, relatively younger parasitoids showed similar levels of attraction to test stimuli as older parasitoids. Likewise, mated and unmated female parasitoids showed little or no difference in their attraction to host-related odors in Y-tube olfactometer bioassays. Subsequent experiments were conducted to measure the level of odorant perception in mated female parasitoids across age groups. The results showed that EAG response of mated parasitoids was not significantly different across age groups.\n\nAlthough aging may negatively affect the host searching ability of female parasitoids, it may not always be attributed to a decline in odor perception in all species. Previous studies showed that relatively younger parasitoids tend to parasitize at a significantly higher rate than older parasitoids36–39. This may be in part due to senescence of some odorant processing apparatus or due to a decrease in energy levels in older parasitoids, thus affecting foraging activities. In a previous related study, age did not affect EAG response of M. croceipes to single VOCs21. In instances where age played a significant role, plasticity of olfactory response in insects at the peripheral and behavioral levels has been attributed to senescence through physiological and neuronal mechanisms17,19.\n\nA few previous studies have reported that the mating status of female parasitoids may affect their foraging behavior and parasitization potential13,22,23,40. In the present study, mating status of female M. croceipes had no significant effect on their olfactory response to host-related odors. This suggests that both virgin and mated females of M. croceipes are likely to seek hosts using host-related odor cues. The concepts of haplodiploidy and optimal foraging provide a better understanding of the ecological ramifications of these results. Haplodiploid parasitoids such as M. croceipes produce male-only offspring from unfertilized eggs and male/female offspring from fertilized eggs. If the sex ratio of a local population is already at equilibrium, host foraging by unmated females may yield immediate benefits41. Otherwise, the cost may outweigh immediate benefits with the development of a male-biased population. Mated females are expected to optimize host foraging and produce progeny with a more balanced sex ratio, which is a critical fitness benefit for the population.\n\nOverall, age and mating did not significantly affect the attraction of female M. croceipes to four test stimuli of varying complexity in the present study, with a few exceptions: mated parasitoids, 4–6 d-old showed significant attraction to the binary mixture and 1–3 d-old mated parasitoids showed significant attraction to H. virescens-infested cotton. The degree of host specificity in M. croceipes may provide a plausible explanation for the overall pattern and exceptions recorded in the present study. It has been proposed that specialist parasitoids exhibit strong congenitally fixed responses while generalist parasitoids exhibit greater plasticity of response to host-related cues42. This hypothesis is consistent with the present results in which M. croceipes (specialist) showed little olfactory plasticity with changes in physiological state. However, M. croceipes is not a strictly specialized species at the extreme of the spectrum. Instead, M. croceipes is a relatively specialized parasitoid utilizing Heliothis/ Helicoverpa host species43. This may possibly explain the few exceptions in which mated and relatively young parasitoids showed significant attraction to the binary mixture and host-infested cotton.\n\nIn summary, the current findings suggest that age and mating status do not play a major role in modulating olfactory responses of M. croceipes to host-related odors. Instead, plasticity of olfactory response may be limited in M. croceipes due to a strong innate sensitivity to host-related odor cues. This may have an impact on their potential as biological control agents. Other physiological factors such as level of nutrition may also have significant effect on olfactory plasticity in parasitoids11,44. This creates an opportunity for augmentation of parasitoids after field releases. Future studies, especially in the field, should investigate the effect of other physiological conditions that may affect plasticity of behavioral response to host-related odors in natural enemies.\n\n\nData availability\n\nThe work presented here was part of an MS project completed by MB. The results presented here have been previously published as MB’s MS thesis available from Auburn University Electronic Theses and Dissertations repository: http://hdl.handle.net/10415/5108\n\nF1000Research: Dataset 1. Attraction of female Microplitis croceipes to four host-related plant odors in Y-tube olfactometer bioassays, https://doi.org/10.5256/f1000research.16927.d22466945\n\nF1000Research: Dataset 2. Electroantennogram (EAG) response of female Microplitis croceipes to four host-related plant odors, https://doi.org/10.5256/f1000research.16927.d22467046", "appendix": "Grant information\n\nThis research was supported in part by the Alabama Agricultural Experiment Station.\n\n\nAcknowledgements\n\nThe authors would like to thank Lindsay McMillan, Benjamin Reeves, and Savannah Duke for assistance with insect rearing.\n\n\nReferences\n\nDeMoraes CM, Lewis WJ, Pare PW, et al.: Herbivore-infested plants selectively attract parasitoids. Nature. 1998; 393: 570–573. Publisher Full Text\n\nPare PW, Tumlinson JH: Plant volatiles as a defense against insect herbivores. Plant Physiol. 1999; 121(2): 325–332. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMumm R, Hilker M: The significance of background odour for an egg parasitoid to detect plants with host eggs. Chem Senses. 2005; 30(4): 337–343. PubMed Abstract \| Publisher Full Text\n\nWei JN, Kang L: Electrophysiological and behavioral responses of a parasitic wasp to plant volatiles induced by two leaf miner species. Chem Senses. 2006; 31(5): 467–477. PubMed Abstract \| Publisher Full Text\n\nBrowne LB: Physiologically induced changes in resource-oriented behavior. Annu Rev Entomol. 1993; 38: 1–23. Publisher Full Text\n\nTakasu K, Lewis WJ: Host- and food-foraging of the parasitoid Microplitis croceipes: Learning and physiological state effects. Biol Control. 1993; 3(1): 70–74. Publisher Full Text\n\nRivero A, Casas J: Incorporating physiology into parasitoid behavioral ecology: the allocation of nutritional resources. Residual Population Ecology. 1999; 41(1): 39–45. Publisher Full Text\n\nMartel V, Anderson P, Hansson BS, et al.: Peripheral modulation of olfaction by physiological state in the Egyptian leaf worm Spodoptera littoralis (Lepidoptera: Noctuidae). J Insect Physiol. 2009; 55(9): 793–797. PubMed Abstract \| Publisher Full Text\n\nGadenne C, Barrozo RB, Anton S: Plasticity in Insect Olfaction: To Smell or Not to Smell? Annu Rev Entomol. 2016; 61: 317–333. PubMed Abstract \| Publisher Full Text\n\nDing HJ, Guo YY, Wu CH: Olfactory electrophysiological responses of cotton bollworm, to allelochemicals of host plants. Acta Entomologica Sinica. 1997; 40: 66–72. Reference Source\n\nStapel JO, Cortesero AM, De Moraes CM, et al.: Extrafloral nectar, honeydew, and sucrose effects on searching behavior and efficiency of Microplitis croceipes (Hymenoptera: Braconidae) in cotton. Environ Entomol. 1997; 26(3): 617–623. Publisher Full Text\n\nGemeno C, Haynes KF: Periodical and age-related variation in chemical communication system of black cutworm moth, Agrotis ipsilon. J Chem Ecol. 2000; 26(2): 329–342. Publisher Full Text\n\nFauvergue X, Lo Genco A, Lo Pinto M: Virgins in the wild: mating status affects the behavior of a parasitoid foraging in the field. Oecologia. 2008; 156(4): 913–920. PubMed Abstract \| Publisher Full Text\n\nHérard F, Keller MA, Lewis WJ, et al.: Beneficial arthropod behavior mediated by airborne semiochemicals : III. Influence of age and experience on flight chamber responses of Microplitis demolitor wilkinson. J Chem Ecol. 1988; 14(7): 1583–1596. PubMed Abstract \| Publisher Full Text\n\nCrnjar R, Yin CM, Stoffolano JG Jr, et al.: Influence of age on the electroantennogram response of the female blowfly (Phormia regina) (Diptera: Calliphoridae). J Insect Physiol. 1990; 36(12): 917–921. Publisher Full Text\n\nKendra PE, Montgomery WS, Mateo DM, et al.: Effect of age on EAG response and attraction of female Anastrepha suspensa (Diptera: Tephritidae) to ammonia and carbon dioxide. Environ Entomol. 2005; 34(3): 584–590. Publisher Full Text\n\nCook-Wiens E, Grotewiel MS: Dissociation between functional senescence and oxidative stress resistance in Drosophila. Exp Gerontol. 2002; 37(12): 1345–1357. PubMed Abstract \| Publisher Full Text\n\nWang MC, Bohmann D, Jasper H: JNK signaling confers tolerance to oxidative stress and extends lifespan in Drosophila. Dev Cell. 2003; 5(5): 811–816. PubMed Abstract \| Publisher Full Text\n\nGrotewiel MS, Martin I, Bhandari P, et al.: Functional senescence in Drosophila melanogaster. Ageing Res Rev. 2005; 4(3): 372–397. PubMed Abstract \| Publisher Full Text\n\nSteinberg S, Dicke M, Vet LEM, et al.: Response of the braconid parasitoid Cotesia (=Apanteles) glomerata to volatile infochemicals: Effects of bioassay set-up, parasitoid age and experience and barometric flux. Entomol Exp Appl. 1992; 63(2): 163–175. Publisher Full Text\n\nPark KC, Zhu J, Harris J, et al.: Electroantennogram responses of a parasitic wasp, Microplitis croceipes, to host-related volatile and anthropogenic compounds. Physiol Entomol. 2001; 26(1): 69–77. Publisher Full Text\n\nTagawa J, Yoshida C, Hashimoto T, et al.: Effects of mating on the oviposition behaviour of the parasitic wasp, Apanteles glomeratus L. (Hymenoptera:Braconidae). J Ethol. 1987; 5(1): 37–41. Publisher Full Text\n\nMichaud JP, Mackauer M: Oviposition behavior of Monoctonus paulensis (Hymenoptera: Aphidiidae): Factors influencing reproductive allocation to hosts and host patches. Ann Entomol Soc Am. 1995; 88(2): 220–226. Publisher Full Text\n\nChen L, Fadmiro HY: Differential electroantennogram response of females and males of two parasitoid species to host-related green leaf volatiles and inducible compounds. Bull Entomol Res. 2007; 97(5): 515–522. PubMed Abstract \| Publisher Full Text\n\nGraham HM, Robertson OT: Host plants of Heliothis virescens and H. zea (Lepidoptera: Noctuidae) in the Lower Rio Grande Valley, Texas. Ann Entomol Soc Am. 1970; 63(5): 1261–1265. Publisher Full Text\n\nRose USR, Lewis WJ, Tumlinson JH: Specificity of systemically released cotton volatiles as attractants for specialist and generalist parasitic wasps. J Chem Ecol. 1998; 24(2): 303–319. Publisher Full Text\n\nWei J, Wang L, Zhu J, et al.: Plants attract parasitic wasps to defend themselves against insect pests by releasing hexenol. PLoS One. 2007; 2(9): e852. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYu H, Zhang Y, Wyckhuys KA, et al.: Electrophysiological and behavioral responses of Microplitis mediator (Hymenoptera: Braconidae) to caterpillar-induced volatiles from cotton. Environ Entomol. 2010; 39(9): 600–609. PubMed Abstract \| Publisher Full Text\n\nNgumbi E, Fadamiro H: Species and sexual differences in behavioural responses of a specialist and generalist parasitoid species to host-related volatiles. Bull Entomol Res. 2012; 102(6): 710–718. PubMed Abstract \| Publisher Full Text\n\nMorawo T, Fadamiro H: Attraction of two larval parasitoids with varying degree of host specificity to single components and a binary mixture of host-related plant volatiles. Chemoecology. 2014a; 24(4): 127–135. Publisher Full Text\n\nNgumbi E, Chen L, Fadamiro HY: Comparative GC-EAD responses of a specialist (Microplitis croceipes) and a generalist (Cotesia marginiventris) parasitoid to cotton volatiles induced by two caterpillar species. J Chem Ecol. 2009; 35(9): 1009–1020. PubMed Abstract \| Publisher Full Text\n\nShorey HH, Hale RL: Mass-Rearing of the Larvae of Nine Noctuid Species on a Simple Artificial Medium. J Econ Entomol. 1965; 58(3): 522–524. Publisher Full Text\n\nMorawo T, Fadamiro H: Identification of Key Plant-Associated Volatiles Emitted by Heliothis virescens Larvae that Attract the Parasitoid, Microplitis croceipes: Implications for Parasitoid Perception of Odor Blends. J Chem Ecol. 2016; 42(11): 1112–1121. PubMed Abstract \| Publisher Full Text\n\nMorawo T, Burrows M, Fadamiro H: Electroantennogram response of the parasitoid, Microplitis croceipes to host-related odors: The discrepancy between relative abundance and level of antennal responses to volatile compound [version 2; referees: 4 approved]. F1000Res. 2017; 5: 2725. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWajnberg É, Haccou P: Statistical tools for analyzing data on behavioral ecology of insect parasitoids. In: Wajnberg É, Bernstein C, van Alphen J (eds) Behavioral ecology of insect parasitoids: from theoretical approaches to field, Blackwell Publishing Ltd, Oxford, UK, 2008; 402–429. Publisher Full Text\n\nBellows TS Jr: Effects of host and parasitoid age on search behaviour and oviposition rates in Lariophagus distinguendus Förster (Hymenoptera: Pteromalidae). Res Popul Ecol. 1985; 27(1): 65–76. Publisher Full Text\n\nRajapakse RHS, Waddill VH, Ashley TR: Effect of host age, parasitoid age and temperature on interspecific competition between Chelonus insularis Cresson, Cotesia marginiventris Cresson and Microplitis manilae Ashmead. Int J Trop Insect Sci. 1992; 13(1): 87–94. Publisher Full Text\n\nAmalin DM, Pena JE, Duncan RE: Effects of host age, female parasitoid age, and host plant on parasitism of Ceratogramma etiennei (Hymenoptera: Trichogrammatidae). Fla Entomol. 2005; 88(1): 77–82. Publisher Full Text\n\nAyvaz A, Karasu E, Karabörklü S, et al.: Effects of cold storage, rearing temperature, parasitoid age, and irradiation on the performance of Trichogramma evanescens Westwood (Hymenoptera: Trichogrammatidae). J Stored Prod Res. 2008; 44(3): 232–240. Publisher Full Text\n\nMichaud JP: Differences in foraging behaviour between virgin and mated aphid parasitoids (Hymenoptera: Aphidiidae). Can J Zool. 1994; 73(9): 1597–1602. Publisher Full Text\n\nGodfray HCJ: The causes and consequences of constrained sex allocation in haplodiploid animals. Evol Biol. 1990; 3(1–2): 3–17. Publisher Full Text\n\nVet LEM, Dicke M: Ecology of infochemical use by natural enemies in a tritrophic context. Annu Rev Entomol. 1992; 37: 141–172. Publisher Full Text\n\nTillman PG, Laster ML: Parasitization of Heliothis virescens and H. virescens-H. subflexa backcross (Lepidoptera: Noctuidae) by Microplitis croceipes (Hymenoptera: Braconidae). Environ Entomol. 1995; 24(2): 409–411. Publisher Full Text\n\nBurrows M, Morawo T, Fadamiro H: Sugar Diet Affects Odor Reception but Variation in Sugar Concentration Plays Minimal Role in the Response of the Parasitoid, Microplitis croceipes (Hymenoptera: Braconidae), to Host-Related Plant Volatiles. J Econ Entomol. 2017; 110(3): 971–977. PubMed Abstract \| Publisher Full Text\n\nBurrows M, Morawo T, Fadamiro H: Dataset 1 in: Do age and mating status affect olfactory response of the parasitoid, Microplitis croceipes (Hymenoptera: Braconidae) to host-related plant odors? F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16927.d224669\n\nBurrows M, Morawo T, Fadamiro H: Dataset 2 in: Do age and mating status affect olfactory response of the parasitoid, Microplitis croceipes (Hymenoptera: Braconidae) to host-related plant odors? F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16927.d224670" }	[ { "id": "40848", "date": "30 Nov 2018", "name": "Lukasz Stelinski", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis manuscript describes a set of behavioural and electrophysiological experiments testing the hypotheses that: 1) mating status and 2) age affect behavioural and antennal response of the parasitoid wasp, Microplitis croceipes to host-related plant odors. Both synthetic and authentic host related odors were tested. Odors were either a general green leaf volatile or volatile organic compounds induced by host (Heliothis virescens) feeding on plants. Overall, the data tend to reject both hypotheses. The authors postulate that M. croceipes display strong innate sensitivity and response to host-related odors with little plasticity due to their relative ecological host specialization.\n\nThe investigators have made numerous scientific contributions using this parasitoid-host study system over a number of years and thus their methods are well established and proven state of the art. The experiments were adequately replicated and the methods are robust and well designed to test the stated hypotheses. The data analysis is appropriate. The data is clearly presented and the conclusions are consistent with the results obtained. I have no reservations endorsing this manuscript for indexing.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4311", "date": "18 Dec 2018", "name": "Tolulope Morawo", "role": "Author Response", "response": "We thank the reviewer for the comments on the manuscript." } ] }, { "id": "41225", "date": "05 Dec 2018", "name": "Andrea Clavijo McCormick", "expertise": [ "Reviewer Expertise Chemical Ecology", "Plant-Insect Interactions", "Insect Behaviour", "Plant Volatiles" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article by Burrows and coworkers investigates the olfactory responses of mated and unmated Microplitis croceipes females of different age-groups to host-related plant volatiles. The results indicate that age and mating status do not affect the odour-guided behaviour or EAG responses of the parasitoids.\nThe paper is clear and well written, citing relevant literature on the topic from previous studies by the same group and other authors. The methods used are standard in the field of chemical ecology and appropriate to answer the research question. One minor issue regarding the methods section is the inclusion of the conditions in which behavioural trials were conducted.\n\nTemperature and relative humidity are known to affect insect behaviour; therefore it would be important to mention if the assays were conducted under controlled conditions or not. And if the second was true, which were the average values for these parameters on the assayed days. A couple of minor style comments:\nThe use of 'and/or' in the introduction (second paragraph). It is a matter of preference but 'or' only should be sufficient. In the methods section (Insects) 'L14:D10 h (L:D) photoperiod' is repetitive, either use L14:D10 h photoperiod or 14:10h L:D photoperiod.\nThe statistical analyses are correct and appropriate to the data collected. The data sets provided ensure the reproducibility of the study, and the results support the conclusions.\nThis is a high quality manuscript and I recommend it for indexing without hesitation.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4310", "date": "18 Dec 2018", "name": "Tolulope Morawo", "role": "Author Response", "response": "We thank the reviewer for the comments on the manuscript. We have indicated that the same general conditions (25 ± 1°C, 75 ± 5 % RH) used for insect rearing were maintained during bioassays. The other two minor changes have been made as suggested." } ] } ]	1	https://f1000research.com/articles/7-1817
https://f1000research.com/articles/7-1297/v1	15 Aug 18	{ "type": "Research Article", "title": "Comparison of clustering tools in R for medium-sized 10x Genomics single-cell RNA-sequencing data", "authors": [ "Saskia Freytag", "Luyi Tian", "Ingrid Lönnstedt", "Milica Ng", "Melanie Bahlo", "Luyi Tian", "Ingrid Lönnstedt", "Milica Ng", "Melanie Bahlo" ], "abstract": "Background: The commercially available 10x Genomics protocol to generate droplet-based single-cell RNA-seq (scRNA-seq) data is enjoying growing popularity among researchers. Fundamental to the analysis of such scRNA-seq data is the ability to cluster similar or same cells into non-overlapping groups. Many competing methods have been proposed for this task, but there is currently little guidance with regards to which method to use. Methods: Here we use one gold standard 10x Genomics dataset, generated from the mixture of three cell lines, as well as three silver standard 10x Genomics datasets generated from peripheral blood mononuclear cells to examine not only the accuracy but also robustness of a dozen methods. Results: We found that some methods, including Seurat and Cell Ranger, outperform other methods, although performance seems to be dependent on the complexity of the studied system. Furthermore, we found that solutions produced by different methods have little in common with each other. Conclusions: In light of this, we conclude that the choice of clustering tool crucially determines interpretation of scRNA-seq data generated by 10x Genomics. Hence practitioners and consumers should remain vigilant about the outcome of 10x Genomics scRNA-seq analysis.", "keywords": [ "Clustering", "Single-Cell RNA-seq", "Benchmarking", "10x Genomics" ], "content": "Introduction\n\nSingle-cell RNA-sequencing (scRNA-seq) studies have opened the way for new data-driven definitions of cell identity and function. No longer is a cell’s type determined by arbitrary hierarchies and their respective predefined markers. Instead, a cell’s transcriptional and epigenomic profile can now be used1 to accomplish this task. This is achieved using computational methods for scRNA-seq that characterize cells into novel and known cell types. Characterization consists of two steps: (i) unsupervised or semi-supervised clustering of same or similar cells into non-overlapping groups, and (ii) labeling clusters, i.e. determining the cell type, or related cell types, represented by the cluster. Here, we focus on the first step of this process.\n\nResearch into clustering has produced many algorithms for the task, including over 60 tools specifically designed for scRNA-seq2. Due to the relative youth of the field, there are currently no rules guiding the application of these clustering algorithms. If tools’ performances have been tested outside synthetic scenarios, testing seems to be confined to scenarios with limited biological variability. Furthermore, most tools were developed and consequently tested only on the Fluidigm C1 protocol, despite considerable differences in throughput capabilities and sensitivities3 in the different scRNA-seq platforms. Here we focus solely on clustering performance on medium-sized scRNA-seq data generated by 10x Genomics as it is currently the most widely used platform. Commercially available scRNA-seq platforms, like 10x Genomics’ Chromium, are being widely adopted due to their ease of use and relatively low cost per cell4. The 10x Genomics protocol uses a droplet-based system to isolate single cells. Each droplet contains all the necessary reagents for cell lysis, barcoding, reverse transcription and molecular tagging. This is followed by pooled PCR amplification and 3’ library preparation, after which standard Illumina short-read sequencing can be applied5. Unlike other commercially available scRNA-seq protocols, like Fluidigm C1, 10x Genomics allows for sequencing of thousands of cells albeit at much shallower read depths per cell, and without allowing the use of fluorescence markers to establish cell identity. As such the 10x Genomics platform is particularly suited to detailed characterization of heterogeneous tissues.\n\n\nMethods\n\nIn this study, we performed comprehensive evaluation of a dozen clustering methods (Table 1). We focused on analysis methods available in the R language, as this is one of the most commonly used programming languages for scRNA-seq data analysis. The exception to this is the 10x Genomics software Cell Ranger. Our evaluation comprised four core aspects: (i) accuracy of clustering solutions compared to a gold standard (near absolute truth, limited variability and complexity), (ii) performance of clustering methods using silver standard data (no absolute truth, realistic variability and complexity), (iii) stability of clustering solutions, and (iv) miscellaneous characteristics, such as time and practicality.\n\nGold standard. Three human lung adenocarcinoma cell lines, HCC827, H1975 and H2228, were cultured separately. The cell lines were obtained from ATCC and cultured in Roswell Park Memorial Institute 1640 medium with 10% fetal bovine serum (FBS, catalog number: 11875-176; Thermo Fisher Gibco) and 1% penicillin-streptomycin. The cells were grown independently at 37°C with 5% carbon dioxide until near 100% confluence. Before mixing cell lines, cells were dissociated into single-cell suspensions in FACS buffer (phosphate-buffered saline (PBS); catalog number: 14190-144; Thermo Fisher Gibco) with 5% FBS, Corning, catalog number: 35-076-CV), stained with propidium iodide (catalog number: P21493; Thermo Fisher FluoroPure) and 120,000 live cells were sorted for each cell line by FACS (BD FACSAria III flow cytometer, BD FACSDiva software version 7.0; BD Biology) to acquire an accurate equal mixture of live cells from the three cell lines. The resulting mixture was then processed by the Chromium Controller (10x Genomics) using single Cell 3’ Reagent Kit v2 (Chromium Single Cell 3’ Library & Gel Bead Kit v2, catalog number: 120237; Chromium Single Cell A Chip Kit, 48 runs, catalog number: 120236; 10x Genomics) (see Table 2). Afterwards the library was sequenced using Illumina NextSeq500 and V4 chemistry (NextSeq 500/550 High Output Kit v2.5, 150 Cycles, catalog number; 20024907; Illumina) with 100bp paired end reads. RTA (version 1.18.66.3; Illumina) was used for base calling.\n\n*https://support.10xgenomics.com/single-cell-gene-expression/datasets/1.1.0/pbmc3k\n\n+https://support.10xgenomics.com/single-cell-gene-expression/datasets/1.2.0/pbmc4k\n\nSilver standard. We consider three human peripheral blood mononuclear cells (PBMCs) scRNA-seq datasets to be the silver standard (Table 2). All datasets were generated using the 10X Genomics droplet system combined with Illumina sequencing. The Australian Genome Research Facility in partnership with CSL generated one dataset using the 10x Genomics Chromium system (Dataset 1). Two datasets were generated by 10x Genomics and are publicly available (Datasets 2 and 3). Of these, one dataset was generated with an earlier version of the microfluidics instrument, the 10x Genomics GemCode Controller (Dataset 2). The second dataset was generated with the latest instrument, the 10x Genomics Chromium Controller (Dataset 3).\n\nFor the first dataset, PBMCs were isolated from whole blood obtained through the Australian Red Cross Blood Service in the following manner. First, 50ml of blood was diluted using 50ml of PBS (catalog number: D8537-500ml; Sigma-Aldrich). We then added 30ml of Ficoll-Paque medium (catalog number: Catalog: 17-1440-03; GE Healthcare). We then centrifuged at room temperature for 20 minutes at 400 g and carefully removed the interface layer containing PBMCs, located between the top plasma layer and middle layer (Heraeus Multifuge 3 S-R Centrifuge; Thermo Fisher Scientific). To remove the supernatant, we further centrifuged at 400 g for 10 minutes at room temperature. This process was repeated to remove the contaminating Ficoll medium or platelets. Finally, cells were resuspended in 20ml of cell culture media with 5% FBS (RPMI-1640 Medium, catalog number: R0884-500ml, Sigma-Aldrich) and counted (Nikon Eclipse TS100 Microscope; Nikon). The resulting mixture was then processed by the Chromium Controller (10x Genomics) using single Cell 3’ Reagent Kit v2 (Chromium Single Cell 3’ Library & Gel Bead Kit v2, catalog number: 120237; Chromium Single Cell A Chip Kit, 48 runs, catalog number: 120236; 10x Genomics). Afterwards the library was sequenced using HiSeq2500 (Illumina) and V4 chemistry (HiSeq PE Cluster Kit v4 cBot, catalog number: PE-401-4001; HiSeq SBS Kit V4 50 cycles, catalog number: FC-401-4002; Illumina) with 101bp paired end reads. RTA (version 1.18.66.3; Illumina) was used for base calling.\n\nWe used the 10x Genomics software, Cell Ranger (version 2.0.0) to align, de-duplicate, filter barcodes and quantify genes for all datasets. Note that we aligned reads with Cell Ranger to the GRCh38 (version 90) genome annotation. Using the Bioconductor package scater17 (version 1.6.3), we then removed low quality data from cells with low library size or low number of expressed gene transcripts. We also removed cells with a high mitochondrial read proportion as this can indicate apoptosis, also known as programmed cell death. Stressed cells undergoing apoptosis have an aberrant transcriptome profile in comparison to a living cell and have previously been acknowledged to adversely influence transcriptome studies13.\n\nGold standard. By exploiting the genetic differences between the three different cell lines we were able to establish near absolute truth in the gold standard dataset. To this end we first called single nucleotide variants (SNVs) in publicly available bulk RNA-seq of the same cell lines (GSE86337)18. Drawing on these SNVs, we then apply demuxlet19 (version 0.0.1), which harnesses the natural genetic variation between the cell lines to determine the most likely identity of each cell. We observe almost complete concordance between the result from demuxlet and clustering of cells seen in dimension reduction visualizations of the data (compare Supplementary Figure 1).\n\nSilver standard. For the silver standard data, we compared clustering solutions to a cell labeling approach by 10x Genomics5. This approach finds the cell type in a reference dataset which most closely resembles the expression in the cell. The reference dataset contains 11 isolated cell types sequenced using the 10x Genomics system. While this labeling does not constitute truth, it has been found to be perform well in comparison with marker-based classification5. Furthermore, the proportions of cells assigned to the 11 cell types by the supervised labeling approach were consistent with the literature (see Supplementary Table 1)20,21.\n\nWe based our selection of method on the online list within www.scrna-tools.org2 in October 2017. We only considered methods with an R package that had sufficient documentation to enable easy installation and execution and had at least one preprint or publication associated with it. We also excluded any methods that required extensive prior information not provided in the package. We also excluded any methods that continually failed to run (e.g. Linnorm22 and Monocle23). This resulted in the evaluation of 12 methods (see Table 1). Note that for some of the R packages the primary focus is not clustering, but the package authors explicitly describe how their packages can be applied to achieve clustering of the scRNA-seq data.\n\nThe aim of this study is to provide guidance for the use of clustering methods to non-experts. Hence, we use all clustering methods with their default parameters as this represents the most common use case. In the case of countClust and SIMLR parameters included the number of clusters, which we set to 3 and 8 for the gold standard and silver standard datasets, respectively. Marker genes were required for the analysis with scran, which we obtained by performing differential expression analyses on GSE86337 and an in-house dataset of isolated cell types in PBMCs24 for the gold standard and silver standard datasets, respectively. Furthermore, we also followed upstream data handling, such as filtering of genes and normalization, as described in the documentation of the respective clustering method. We concede that it is possible that more care in the upstream data handling and selection of parameters could result in different results. However, confronted with the extremely large number of parameter choices, we believe that this evaluation suffices to identify strengths and weaknesses of each method.\n\nTo evaluate the performance and similarity of different clustering solutions, we rely on three different metrics. We use the adjusted Rand index (ARI)25 and the normalized mutual information (NMI)26, two metrics routinely applied in the field of clustering, to assess the similarity of clustering solutions or their similarity to a known truth. Both metrics can take values from 0 to 1, with 0 signifying no overlap between two groupings and 1 signifying complete overlap. These metrics are also applicable in the absence of known cluster labels. Finally, we also use a homogeneity score27. This score takes the value 1 when all of its clusters contain only data points that are members of a single known group. Values of this score closer to 0 indicate that clusters contain mixed known groups. Unlike NMI and ARI, this score requires knowledge of an underlying truth.\n\nLet X be a finite set of size n. A clustering solution C is a set C1, . . . , Ck of non-empty disjoint subsets of X such that their union equals X. Let C′=C1′,...,Cl′ be a second clustering solution or the supervised labeling solution with the same properties. The contingency table M = (mij) of the pair of sets C, C′ is a k × l matrix whose i, j-th entry equals the number of elements in the intersection of clusters Ci and Cj′:\n\n\n\nARI\n\n\n\nwhere t1 = ∑i=1k(\|Ci\|2),t2=∑j=1l(\|Cj′\|2) and t3=2t1t2n(n−1)· For ease of notation this is referred to as ARI in the text, dropping the reference to specific pairs of sets. Furthermore, we also distinguish between ARI_truth as a comparison of a clustering solution to an underlying known or suspected truth and ARI_comp, which refers to a comparison between two clustering solutions.\n\nNMI\n\n\n\nwhere H(C) = I(C, C) is the entropy of C. Note that\n\n\n\nwhere P(i, j) = mijn and P(i) = \|Ci\|n, is the mutual information of C and C‘.\n\nHomogeneity. Now let us assume C′ is the known and correct grouping of the cells. Then,\n\n\n\nTo test the robustness of different clustering methods we pursued a sampling strategy in terms of genes and cells. We used Dataset 3 for the cell robustness evaluation with regards to cells, since it had the most number of cells. Similarly, we used Dataset 1 for the robustness evaluation with regards to genes, since it had the most number of non-zero genes after filtering. The impact of different aligners and preprocessing was assessed using all appropriate combinations of programs.\n\nCells. We randomly sampled 3,000 cells in Dataset 3 (out of the total of 4,292 that were available after filtering), generating five (non-independent) datasets. For every combination of two datasets (10 combinations in total) we then investigated for each clustering method separately how often cells contained in all five sampled datasets were assigned to the same cluster using the ARI_comp.\n\nGenes. We randomly filtered half of all genes in Dataset 1 (out of the total of 58,302 genes), generating 10 datasets. For every combination of two datasets (45 combinations in total) we then investigated for each clustering method separately how often cells were assigned to the same cluster using the ARI_comp.\n\nAligners and preprocessing pipelines. In order to assess the affect of using different preprocessing pipelines on the data, we applied the Bioconductor package scPipe28 (version 1.0.6) to the raw data. Like Cell Ranger, scPipe can be used to align, deduplicate, filter barcodes and quantify genes. Since scPipe is modular, we tried it with both the STAR29 (version 020201) and Subread30 (version 1.5.2) aligners. In order to ensure comparability we aligned reads to the same GRCh38 genome annotation and repeated quality control with scater. We investigated the similarity of clustering solutions applied to the differently preprocessed and aligned versions of the same dataset by ARI_comp.\n\nEach execution of a method on a dataset was performed in a separate R session. Each task was allocated as many CPU cores of a 24 core Intel(R) Xeon(R) CPU E5-2690 v3 @ 2.60GHz as specified by the default parameters. The base::set.seed was overridden in order to prevent stochasticity and thus give reduced unwanted variation in the results. Timings for each method include any preprocessing steps.\n\nWe also investigated what properties of each cell’s data were driving the clustering solutions produced by the different methods as well as the inferred cell labels. Properties of a cell’s data refer to features such as the number of total reads that included the cell’s barcode, the total number of gene transcripts found for this cell, etc. To this end, we used linear mixed models where cell data properties were predicted using the indicators for cluster membership. We predicted cell data properties and not cluster membership for modeling ease. The adjusted R2 of these models was used to assess which properties influenced the clustering solutions. Properties investigated included: (i) the total number of detected gene transcripts, (ii) the total read count, and (iii) the percentages of reads aligning respectively to ribosomal proteins, mitochondrial genes and ribosomal RNA.\n\n\nResults\n\nGold standard data. For the gold standard data set consisting of three cell types, half of the tested clustering methods overestimated the true number of different cell types in the data. Methods with cluster number estimations close to the correct number of different cell types included methods with prior information, such as SIMLR, countClust and scran, as well as ascend, Cell Ranger, RaceID and CIDR (Figure 1). The clustering solutions produced by these methods, with the exception of countClust, largely reflect the cell types. This is indicated by ARI_truth >0.8. The remaining methods overestimated the number of clusters by 2 to 85 clusters, with SC3 and RaceID2 representing the extremes, both estimating more than 20 clusters (see t-SNE plots in Supplementary Figure 1 for the impact). As a consequence of the greater number of estimated clusters, the ARI_truth of the other clustering methods is lower than 0.8. To see whether these methods split cell types into several clusters or instead assign cells types randomly to clusters, we also investigate the homogeneity of the clustering solutions with respect to the known labeling. Apart from countClust and RCA, all methods have extremely high homogeneity, indicating that they split cell types into more subtypes, rather than randomly creating more cell types, which is reassuring.\n\n(a) ARI_truth of each method with regards to the truth versus the number of clusters. The dashed line indicates the true number of clusters.(b) Homogeneity of clusters of each method, given the truth.\n\nSilver standard data. We labeled the cells in each of the three silver standard datasets as one of 11 different PBMC cell populations. When using the ARI_truth to compare the likeness of the clustering solutions and the labels, no method produced solutions that were uniformly the most similar to the inferred labels (Figure 2). Two methods, ascend and countClust, tended to estimate smaller number of clusters and consequently did not agree with the labeling. Only Seurat, SC3 and Cell Ranger achieved an ARI_truth above 0.4 for at least two out of the three silver standard datasets. All methods considerably improved their ARI_truth when we subset to more confidently labeled cells (see Supplementary Figure 2). RCA was particularly affected, showing much greater similarity for more confidently labeled cells. We also calculated the homogeneity of each method in each dataset with respect to the inferred labeling. Generally, most methods exhibited significantly lower performance on Dataset 2, which was generated with an older version of the 10x Genomics technology than that used to generate Datasets 1 and 3. Apart from RCA all methods had much lower accuracy than for the gold standard data, indicating that most clusters represent mixtures of different inferred cell types. The exception is SC3’s clustering solution of Dataset 3, which achieved an homogeneity score above 0.7.\n\n(a) ARI_truth of each method in each dataset, as indicated by different shapes, with regards to the supervised cell labeling versus the number of clusters. The dashed line indicates the number of cell populations estimated by the supervised cell labeling approach. (b) Homogeneity of clusters with regards to the inferred labeling for each method and each dataset. Different datasets are indicated by transparency.\n\nInterestingly, similar performance when compared to the labeling did not imply that cluster solutions were similar (compare Figure 3). In fact, only a few methods resulted in clustering solutions that were similar. For all three datasets, a group of five methods (RCA, scran, Seurat, SIMLR and TSCAN) produced similar results, while the other seven methods appeared dissimilar between each other and to the set of five methods.\n\nSimilarity of all combinations of clustering methods as estimated by ARI_comp (lower triangle) and NMI (upper triangle) in (a) Dataset 1, (b) Dataset 2, and (c) Dataset 3. The similarity is indicated by the color; yellow indicating no similarity and purple indicating complete overlap. The diagonals give the number of clusters estimated by each respective method.\n\nStability. We evaluated the stability of the clustering methods by examining three different features: (i) filtering of cells, (ii) filtering of genes (Figure 4), and (iii) use of different aligners (Figure 5). When assessing the stability with regards to input, countClust, RaceID and RaceID2 did not appear very robust. The robustness of ascend, countClust and RaceID was variable. Due to its reliance on reference profiles RCA is extremely robust, achieving ARI_comp above 0.9 consistently. Seurat, SIMLR and Cell Ranger demonstrated robustness with regards to input, but also exhibited robustness when changing gene filtering procedures (compare Figure 4b). CIDR appeared to be very sensitive to changes in gene filtering, which may be due to its imputation feature.\n\nWe also investigated how the stability of the clustering method was affected by the use of different aligners (Figure 5). In particular, we used Cell Ranger and ScPipe28 with Subread30, or STAR29. We found that different aligners largely result in the same gene counts, but with some notable exceptions for processed pseudogenes (see Supplementary Figure 4, Supplementary Figure 5 and Supplementary Figure 6). Not all methods were able to be used in conjunction with scPipe. This included ascend and SIMLR, which failed to run, and Cell Ranger, which requires output from its own preprocessing pipeline. However we were able to evaluate eight methods. Apart from RaceID2 and RCA, all tested methods appeared robust.\n\nTukey boxplots of ARI_comp results from the comparison of clustering solutions of the same method when (a) cell input was varied in Dataset 3 and (b) gene input was varied in Dataset 1. Note that RaceID only estimated 1 cluster when genes were varied.\n\nOnly the subset of nine methods that worked in conjunction with all three aligners are shown.\n\nMiscellaneous properties. Running time varies substantially between different methods. Methods like RaceID and RaceID2 take prohibitively long and thus do not lend themselves to interactive analysis when applied to 10x Genomics data (Figure 6). The fastest methods were RCA and TSCAN, with both taking less than 25 seconds on average for the entire dataset analysis. Their fast running time is due to both of these methods offering little flexibility or intermediate results during their analysis (compare Table 1). By contrast Seurat’s relatively long running time is partially due to extensive quality control during the analysis. Also note that methods differed in the quality of their documentation. For example, tools like Cell Ranger and Seurat offer detailed documentation, with many different use cases as well as tutorials. Tools, which are not found on Bioconductor, such as RaceID, RaceID2, ascend and RCA have more limited documentation.\n\nThe variation in the percentage of reads aligning to ribosomal protein genes strongly predicted all clustering solutions as well as the inferred cell labels (see Figure 7). Expression of ribosomal protein genes has been successfully used to discriminate cell types belonging to different hematopoietic lineages31. Hence, it may be the case that overall mRNA amount of ribosomal protein genes can also serve as a discriminator. Furthermore, differences in abundance of ribosomal protein genes are likely to drive variation in PBMCs scRNA-seq datasets, as they typically account for a large proportion of reads (around 40% in all three datasets). In combination with ribosomal protein genes being less affected by dropout due to their relatively high expression, it is perhaps unsurprising that clustering solutions of all methods foremost reflect differences in the amount of ribosomal protein genes between cells.\n\nMost methods’ solutions were much more driven by the total number of features and total number of counts than the inferred solution. TSCAN was particularly affected (R2 = 0.52), but for both ascend and RaceID2 similar effects were observed. It can be speculated that this strong influence of total number of features and total number of count on their clustering solutions points to a failure to appropriately normalize the data.\n\nFor every method and every feature the adjusted R2 of the linear model fitting the feature by the clustering solution is presented.\n\n\nDiscussion\n\nMost biological conclusions obtained from droplet-based scRNA-seq data crucially rely on accurate clustering of cells into homogeneous groups. Indeed, one can argue that it is the very act of clustering that unlocks the technology’s potential for discovery. Therefore it is not surprising that according to several repositories, such as www.omicstools.org and www.scRNA-tools.org2, many of the tools developed for scRNA-seq specifically focus on clustering. With so many choices, it is thus important to evaluate their performance for droplet based protocols, such as 10x Genomics, specifically.\n\nIn this study, we presented our evaluation of a dozen clustering method on scRNA-seq 10x Genomics data. The results of our investigations will be useful for method users, as we provide clear and practical guidelines. Nonetheless, our evaluation has several limitations:\n\n• Inclusion of methods limited to R packages and methods published before October 2017\n\n• Parameter selection limited to defaults\n\n• No assessment of robustness to noise and parameter changes\n\n• No assessment of ability to discover rare cell populations\n\n• Evaluation of more silver standard datasets from systems other than PBMCs\n\n• No evaluation of quality of code and documentation\n\n• No assessment of scalability of methods\n\nOur evaluations suggest that Seurat and Cell Ranger provide the most stable and accurate clustering solutions for 10x Genomics scRNA-seq data. While Seurat performed slightly better, the choice between Seurat and Cell Ranger, in our opinion, should be informed by the user’s familiarity with statistical concepts, which enable them to make the informed parameter choices required in Seurat. More generally we find that different clustering methods resulted in very different solutions. The good performance of Seurat as well as the vast difference between clustering methods have also been observed by Duò et al.32 in a benchmarking study including multiple scRNA-seq protocols. Our investigations suggest that biological differences between cells, such as cell type or state, and technical variation between cells (as well as combinations of biological differences and technical variation) all drive clustering. However, which aspects are captured by which clustering method remain to be confirmed. Our study merely pinpoints some of the drivers of performance, but not their origin, and thus cannot anoint an overall best method.\n\nWe recommend that practitioners and consumers of results generated from 10x Genomics scRNA-seq data alike remain vigilant about the outcome of their analysis, and acknowledge the variability and likelihood of undesired influences. The choice of clustering tool for scRNA-seq data generated by the 10x Genomics platform crucially determines interpretation. Hence, at least two clustering tools should be routinely applied to 10x Genomics scRNA-seq data in order to offer more than one subjective interpretation and hence increase robustness and confidence in any results.\n\n\nData availability\n\nRepository: Gold Standard Dataset. Single cell profiling of 3 Human Lung Adenocarcinoma cell lines, GSE111108\n\nRepository: Silver Standard Dataset 1. Single cell profiling of peripheral blood mononuclear cells from healthy human donor, GSE115189\n\nRepository: Silver Standard Dataset 2. 3k PBMCs from a Healthy Donor, https://support.10xgenomics.com/single-cell-gene-expression/datasets/1.1.0/pbmc3k\n\nRepository: Silver Standard Dataset 3. 4k PBMCs from a Healthy Donor, https://support.10xgenomics.com/single-cell-gene-expression/datasets/1.2.0/pbmc4k\n\n\nSoftware availability\n\nAll code is available for download at: https://github.com/SaskiaFreytag/cluster_benchmarking_code.\n\nArchived code at time of publication: 10.5281/zenodo.1324576\n\nLicense: MIT License\n\n\nConsent\n\nWritten informed consent for publication of the participant’s transcriptomic information was obtained (Australian Red Cross Blood Service Supply Agreement 18-03VIC-07).", "appendix": "Author contributions\n\n\n\nFreytag S: Conceptualization, Data Curation, Funding Acquisition, Formal Analysis, Investigation, Methodology, Software, Visualization, Writing – Original Draft Preparation, Writing – Review & Editing\n\nTian L: Investigation, Writing – Review & Editing\n\nLönnstedt I: Conceptualization, Methodology, Writing – Review & Editing\n\nNG M: Conceptualization, Investigation, Funding Acquisition, Methodology, Writing – Review & Editing\n\nBahlo M: Supervision Conceptualization, Investigation, Funding Acquisition, Methodology, Writing – Review & Editing\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nWe would like to thank the Australian Genome Research Facility and the Genomics Innovation Hub for their generous support of this project, including funding. This work was also supported by the Victorian Government’s Operational Infrastructure Support Program and Australian Government NHMRC IRIIS. MB is funded by NHMRC Senior Research Fellowship 110297 and NHMRC Program Grant 1054618.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe gratefully acknowledge the constructive comments and experimental work of Azadeh Seidi, Mark Biondo and Nicolas J. Wilson. Additionally, we want to acknowledge Mark Robinson for his great advice.\n\n\nSupplementary material\n\nSupplementary Figure 1. T-SNE plot of the gold standard data after filtering and normalization with the package scater. Shapes indicate the cell identity as established by demuxlet. The colors indicate in (a) RaceID2 clustering, (b) SC3 clustering and in (c) Seurat clustering. It can be observed that these three programs present different degrees of complexity.\n\nClick here to access the data.\n\nSupplementary Figure 2. ARI_truth of each method on all three silver standard datasets given the subset of the cells that reached respective minimum correlation using the cell labeling approach by Zheng et al.5.\n\nClick here to access the data.\n\nSupplementary Figure 3. Median of ARI_comp of each method when cell input is changed versus median of ARI_comp of each method when gene input is changed. See Figure 4 for variability associated with each methods’ ARI_comp.\n\nClick here to access the data.\n\nSupplementary Figure 4. UpSeTR Venn diagram to compare of number genes detected in the gold standard dataset by different programs used for preprocessing. The vast majority of genes are detected by all programs.\n\nClick here to access the data.\n\nSupplementary Figure 5. Total counts for the same barcode as measured when processed with (a) Cell Ranger versus ScPipe Subread, (b) Cell Ranger versus ScPipe STAR, and (c) ScPipe STAR versus ScPipe STAR. Only barcodes are shown that appear in all three versions of the processed dataset.\n\nClick here to access the data.\n\nSupplementary Figure 6. Tukey boxplots showing the correlations between gene counts of particular gene category for all three comparisons between preprocessing programs used on gold standard dataset. Processed pseudogenes’ counts seem to differ depending on program used.\n\nClick here to access the data.\n\nSupplementary Table 1. Proportion of cell types in different silver standard datasets as estimated by supervised cell labeling.\n\nClick here to access the data.\n\n\nReferences\n\nTanay A, Regev A: Scaling single-cell genomics from phenomenology to mechanism. Nature. 2017; 541(7637): 331–338. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZappia L, Phipson B, Oshlack A: Exploring the single-cell RNA-seq analysis landscape with the scRNA-tools database. PLoS Comput Biol. 2018; 14(6): e1006245. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZiegenhain C, Vieth B, Parekh S, et al.: Comparative Analysis of Single-Cell RNA Sequencing Methods. Mol Cell. 2017; 65(4): 631–643.e4. PubMed Abstract \| Publisher Full Text\n\nHaque A, Engel J, Teichmann SA, et al.: A practical guide to single-cell RNA-sequencing for biomedical research and clinical applications. Genome Med. 2017; 9(1): 75. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZheng GX, Terry JM, Belgrader P, et al.: Massively parallel digital transcriptional profiling of single cells. Nat Commun. 2017; 8: 14049. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSenabouth A, Lukowski S, Alquicira J, et al.: ascend: R package for analysis of single cell RNA-seq data. bioRxiv. 2017; 207704. Publisher Full Text\n\nLin P, Troup M, Ho JW: CIDR: Ultrafast and accurate clustering through imputation for single-cell RNA-seq data. Genome Biol. 2017; 18(1): 59. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDey KK, Hsiao CJ, Stephens M: Visualizing the structure of RNA-seq expression data using grade of membership models. PLoS Genet. 2017; 13(3): e1006599. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGrün D, Lyubimova A, Kester L, et al.: Single-cell messenger RNA sequencing reveals rare intestinal cell types. Nature. 2015; 525(7568): 251–5. PubMed Abstract \| Publisher Full Text\n\nGrün D, Muraro MJ, Boisset JC, et al.: De Novo Prediction of Stem Cell Identity using Single-Cell Transcriptome Data. Cell Stem Cell. 2016; 19(2): 266–277. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLi H, Courtois ET, Sengupta D, et al.: Reference component analysis of single-cell transcriptomes elucidates cellular heterogeneity in human colorectal tumors. Nat Genet. 2017; 49(5): 708–718. PubMed Abstract \| Publisher Full Text\n\nKiselev VY, Kirschner K, Schaub MT, et al.: SC3: consensus clustering of single-cell RNA-seq data. Nat Methods. 2017; 14(5): 483–486. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLun AT, Bach K, Marioni JC: Pooling across cells to normalize single-cell RNA sequencing data with many zero counts. Genome Biol. 2016; 17(1): 75. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nButler A, Hoffman P, Smibert P, et al.: Integrating single-cell transcriptomic data across different conditions, technologies, and species. Nat Biotechnol. 2018; 36(5): 411–420. PubMed Abstract \| Publisher Full Text\n\nWang B, Ramazzotti D, De Sano L, et al.: SIMLR: A Tool for Large-Scale Genomic Analyses by Multi-Kernel Learning. Proteomics. 2018; 18(2): 1700232. PubMed Abstract \| Publisher Full Text\n\nJi Z, Ji H: TSCAN: Pseudo-time reconstruction and evaluation in single-cell RNA-seq analysis. Nucleic Acids Res. 2016; 44(13): e117. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMcCarthy DJ, Campbell KR, Lun AT, et al.: Scater: pre-processing, quality control, normalization and visualization of single-cell RNA-seq data in R. Bioinformatics. 2017; 33(8): 1179–1186. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHolik AZ, Law CW, Liu R, et al.: RNA-seq mixology: designing realistic control experiments to compare protocols and analysis methods. Nucleic Acids Res. 2017; 45(5): e30. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKang HM, Subramaniam M, Targ S, et al.: Multiplexed droplet single-cell RNA-sequencing using natural genetic variation. Nat Biotechnol. 2018; 36(1): 89–94. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSasaki Y, Darmochwal-Kolarz D, Suzuki D, et al.: Proportion of peripheral blood and decidual CD4+ CD25bright regulatory T cells in pre-eclampsia. Clin Exp Immunol. 2007; 149(1): 139–145. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJing Y, Gravenstein S, Chaganty NR, et al.: Aging is associated with a rapid decline in frequency, alterations in subset composition, and enhanced Th2 response in CD1d-restricted NKT cells from human peripheral blood. Exp Gerontol. 2007; 42(8): 719–732. PubMed Abstract \| Publisher Full Text\n\nYip SH, Wang P, Kocher JA, et al.: Linnorm: improved statistical analysis for single cell RNA-seq expression data. Nucleic Acids Res. 2017; 45(22): e179. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTrapnell C, Cacchiarelli D, Grimsby J, et al.: The dynamics and regulators of cell fate decisions are revealed by pseudotemporal ordering of single cells. Nat Biotechnol. 2014; 32(4): 381–386. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nde Graaf CA, Choi J, Baldwin TM, et al.: Haemopedia: An Expression Atlas of Murine Hematopoietic Cells. Stem cell reports. 2016; 7(3): 571–582. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHubert L, Arabie P: Comparing partitions. J Classif. 1985; 2(1): 193–218. Publisher Full Text\n\nStudholme C, Hill DLG, Hawkes DJ: An overlap invariant entropy measure of 3D medical image alignment. Pattern Recogn. 1999; 32(1): 71–86. Publisher Full Text\n\nRosenberg A, Hirschberg J: V-measure: A conditional entropy-based external cluster evaluation measure. In Proceedings of the 2007 joint conference on empirical methods in natural language processing and computational natural language learning (EMNLP-CoNLL). 2007. Reference Source\n\nTian L, Su S, Amann-Zalcenstein D, et al.: scPipe: a flexible R/Bioconductor preprocessing pipeline for single-cell RNA-sequencing data. bioRxiv. 2017; 175927. Publisher Full Text\n\nDobin A, Davis CA, Schlesinger F, et al.: STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013; 29(1): 15–21. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiao Y, Smyth GK, Shi W: The Subread aligner: fast, accurate and scalable read mapping by seed-and-vote. Nucleic Acids Res. 2013; 41(10): e108. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGuimaraes JC, Zavolan M: Patterns of ribosomal protein expression specify normal and malignant human cells. Genome Biol. 2016; 17(1): 236. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDuò A, Robinson MD, Soneson C: A systematic performance evaluation of clustering methods for single-cell RNA-seq data [version 1; referees: 2 approved with reservations]. F1000Res. 2018; 7: 1141. Publisher Full Text" }	[ { "id": "37232", "date": "28 Aug 2018", "name": "Joshua W. K. Ho", "expertise": [ "Reviewer Expertise Bioinformatics", "single-cell transcriptomics" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper presents a well-designed and comprehensive evaluation of widely used clustering algorithms for medium-sized 10x Genomics scRNA-seq data. Clustering is a highly active area of research in scRNA-seq data analysis. With so many published clustering tools available, it is often difficult to choose the most appropriate tool. This paper attempts to address this problem by systematically comparing the performance of 12 commonly used clustering tools. The evaluation results should serve as an important guide to bioinformatics practitioners. This paper is a very useful contribution to the field.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "37231", "date": "29 Aug 2018", "name": "Shila Ghazanfar", "expertise": [ "Reviewer Expertise Statistics", "statistical bioinformatics" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nFreytag and colleagues provide a comprehensive comparison of clustering methods - specifically designed for scRNA-Seq data - on data collected using the popular droplet-based 10x Genomics platform. A total of four datasets, comprising a Gold standard mixture of cell lines as well as three Silver standard PBMC datasets, were compared in terms of accuracy, stability as well as other metrics like runtime and ease of use. Freytag et al also perform an analysis to try to determine the factors influencing the resulting clusterings for the Silver standard datasets.\nIt is a very challenging task to perform a comprehensive characterization and comparison of clustering methods on such types of high-dimensional data, due to the sheer number of choices that need to be made, the difficulty in establishing ideal performance, and the relative lack of ground truth. Freytag et al do a great job of addressing these challenges and working towards providing an overall recommendation of clustering methods for non-expert practitioners, while stressing the need for careful interpretation of such results.\n\nWith this in mind, I have some comments/suggestions, as well as a number of minor comments/suggestions, as follows:\nComments to authors\nLinnorm and Monocle failed - expand on why? I understand that this is indeed a limitation especially for a non-expert practitioner, but it would be good to have an understanding towards what the issue might have been.\nCould use a flowchart to summarise the study and various comparisons, as well which methods could no longer be compared (e.g. methods that could not work within the scPipe framework).\nDifferent upstream data handling was performed for each clustering method. How much of a difference was observed just due to this preprocessing, as opposed to the actual clustering step? I understand that each method provides their own preprocessing as part of the method, but at least some of these methods would have been developed with plate-based and/or non-UMI-based scRNA-Seq in mind, so may not be intended for the context of 10x Genomics data. Again I understand that you're comparing methods 'out of the box' but it would be insightful to see what differences there are. I suggest a figure like an upsetR plot for the genes/cells filtered and a correlation heatmap of the expression values themselves.\nCould you summarise the distance metrics used in the clustering and if there is a general flavour to the clustering algorithm? e.g. hierarchical, k-means, density-based etc. How do these relate in terms of overall accuracy, stability and other metrics?\nStability assessment - mentions that half of the 58,302 genes were randomly selected, but Table 1 says 24,654 total genes detected. There's a big discrepancy between these two so please clarify; if half of the 58,302 genes were selected then a large proportion of genes would have identically zero rows. Also Table 1 shows Dataset 3 had the highest number of 'total genes detected', so how was Dataset 1 the one with \"most number of non-zero genes after filtering\"?\nRun time section - What do you mean by 'overridden'? And for which aspects of the analysis steps was this done?\nFigure 4 - These boxplots show ARI among multiple clustering solutions, so a method that gives a consistently bad result is still high (e.g. in this case the RCA method). Suggest an analogous set of boxplots but with ARI_truth, is there a similar variability observed, as seen in these boxplots?\nGene-wise stability analysis - I'm actually unsure how realistic this particular comparison is. It would be insightful to assess clusterings depending on different levels of gene filtering stringency (in the initial Cell Ranger read processing), or stringency on selection of features based on various criteria like highly variable genes.\nFigure 7 - Please clarify how 'total number of features' is a cell-specific quantity. Do you mean total number of non-zero features? Was this analysis also performed on the Gold Dataset and what overall similarities could be observed?\nFactors influencing clustering solutions - It would be interesting to consider the factors associated with 'correct' cluster assignment for cells. Optionally suggest to perform this for either the Gold Dataset or the Silver datasets and perform a logistic regression with the response being success/failure of a cell to belong to the cluster most associated with the 'true' cell type group. There is an added subtlety as far as matching clusters with cell type groups goes, but I think there are a few reasonable ways to perform this (e.g. assign candidate clusters to the 'true' groups by taking the higher proportion of cell overlap, and allow multiple candidate clusters to match to a single true group). Performing this kind of analysis could shed light on properties of cells that don't tend to cluster correctly, and if there is consistency in this across multiple disparate datasets.\nMinor comments\nTable 1 - countClust 'version' formatted with verbatim.\nTable 1 - I would suggest the 'properties' column could be better presented in a checklist format, with ticks/crosses for fulfilling various criteria listed.\nSection beginning \"silver standard\" - 10x is capitalised.\nSupplementary Figure 1 - legend fallen off panel a), needs a higher resolution or larger points\nNMI definition - trailing parenthesis in denominator\ntypo - assess the effect*\nFigure 2a - I found this quite busy, hard to interpret. Suggest to add shading that covers the points for same method or to facet by dataset. I don't believe the ARI values are particularly comparable between datasets so I would prefer facetting by dataset.\nFigure 3 - rows/columns are ordered differently between panels, what's driving this difference?\nSupplementary Figure 3 was not mentioned in the main text\nSupplementary Figure 4 is a two page pdf, with the first page blank\nFigure 6 - Figure caption says Dataset 1 but reports 29,151 genes. Do you mean the Gold Dataset and 29,451 genes? If not, please clarify which data and how many genes.\nDiscussion - One instance of \"Seurat\" is missing verbatim format\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4301", "date": "19 Dec 2018", "name": "Saskia Freytag", "role": "Author Response", "response": "We would like to thank the reviewer for reviewing our manuscript and for their constructive comments. Below are point-by-point responses to the individual comments. Linnorm and Monocle failed - expand on why? I understand that this is indeed a limitation especially for a non-expert practitioner, but it would be good to have an understanding towards what the issue might have been. Linnorm failed because its calculations would time out. Monocle failed because the dispersion could not be calculated. However, neither of the programs was tried using their newer package versions corresponding to R version 3.5.0. We have included a statement in the manuscript to this regard. Could use a flowchart to summarize the study and various comparisons, as well which methods could no longer be compared (e.g. methods that could not work within the scPipe framework). While we were unable to summarize our study design effectively in a flowchart, we have summarized it in a table (see Supplementary Table 2). We hope that this will clarify the various assessments performed in this paper. Different upstream data handling was performed for each clustering method. How much of a difference was observed just due to this preprocessing, as opposed to the actual clustering step? I understand that each method provides their own preprocessing as part* of the method, but at least some of these methods would have been developed with plate-based and/or non-UMI-based scRNA-Seq in mind, so may not be intended for the context of 10x Genomics data. Again I understand that you're comparing methods 'out of the box' but it would be insightful to see what differences there are. I suggest a figure like an upsetR plot for the genes/cells filtered and a correlation heatmap of the expression values themselves. We agree that different data handling influences the performance of each clustering method, which were indeed designed with different single cell technologies in mind, and that the effect of this would be interesting to further investigate. However, the `black-box` nature of some of the investigated methods means that even recording these differences is challenging. Take Seurat as an example it is unclear whether to report the number of genes passing the filtering step or the number of genes that are used in the clustering. Instead, we would like to refer you to the recent benchmarking study of clustering methods for scRNA-seq by Duó et al, where the authors investigated the effects of different gene filtering on clustering solutions. Could you summarise the distance metrics used in the clustering and if there is a general flavour to the clustering algorithm? e.g. hierarchical, k-means, density-based etc. How do these relate in terms of overall accuracy, stability and other metrics? Thank you for the suggestion. We have updated the table summarizing the properties of the different clustering methods and added a discussion regarding how different flavors of clustering methods relate to overall performance (see Table 1 and Discussion). Stability assessment - mentions that half of the 58,302 genes were randomly selected, but Table 1 says 24,654 total genes detected. There's a big discrepancy between these two so please clarify; if half of the 58,302 genes were selected then a large proportion of genes would have identically zero rows. Also Table 1 shows Dataset 3 had the highest number of 'total genes detected', so how was Dataset 1 the one with \"most number of non-zero genes after filtering\"? You are correct. We randomly selected half of 58,302 genes of which many were zero. We have since replaced this analysis, as per your suggestion, with an analysis that assesses stability when keeping only the top 10th, 20th, 30th, 40th, and 50th percentile of all genes including the ones not detected. With regards to the number of detected genes in dataset 1 and dataset 3, indeed dataset 3 had more detected genes. Thank your for correcting this. Run time section - What do you mean by 'overridden'? And for which aspects of the analysis steps was this done? We meant to say that a seed had been set to provide reproducibility of all parts of the analysis that involve randomness. This has been corrected in the manuscript. Figure 4 - These boxplots show ARI among multiple clustering solutions, so a method that gives a consistently bad result is still high (e.g. in this case the RCA method). Suggest an analogous set of boxplots but with ARI_truth, is there a similar variability observed, as seen in these boxplots? Thank you for the suggestion, we have included a boxplot with ARI_truth. Gene-wise stability analysis - I'm actually unsure how realistic this particular comparison is. It would be insightful to assess clusterings depending on different levels of gene filtering stringency (in the initial Cell Ranger read processing), or stringency on selection of features based on various criteria like highly variable genes. We have replaced the gene-wise stability analysis with an assessment of the performance when keeping only the top 10th, 20th, 30th, 40th, and 50th percentile of all genes (compare Figure 6). We think that this is more insightful as it is closer to filtering performed during analysis. Figure 7 - Please clarify how 'total number of features' is a cell-specific quantity. Do you mean total number of non-zero features? Was this analysis also performed on the Gold Dataset and what overall similarities could be observed? Indeed we do mean the number of non-zero genes and we have replaced this in the figure with “number of detected genes”. We also include the same analysis on the gold standard dataset in the Supplementary (Supplementary Figure 11). Factors influencing clustering solutions - It would be interesting to consider the factors associated with 'correct' cluster assignment for cells. Optionally suggest to perform this for either the Gold Dataset or the Silver datasets and perform a logistic regression with the response being success/failure of a cell to belong to the cluster most associated with the 'true' cell type group. There is an added subtlety as far as matching clusters with cell type groups goes, but I think there are a few reasonable ways to perform this (e.g. assign candidate clusters to the 'true' groups by taking the higher proportion of cell overlap, and allow multiple candidate clusters to match to a single true group). Performing this kind of analysis could shed light on properties of cells that don't tend to cluster correctly, and if there is consistency in this across multiple disparate datasets. We did perform the suggested analysis. However, results from this analysis did not give any insights beyond the already conducted analysis (see https://github.com/SaskiaFreytag/cluster_benchmarking_code/tree/master/revision_figure). Hence, we chose not to include this in the manuscript. Table 1 - countClust 'version' formatted with verbatim. Thank you for noticing, this has been corrected. Table 1 - I would suggest the 'properties' column could be better presented in a checklist format, with ticks/crosses for fulfilling various criteria listed. Table 1 is now Supplementary Table 1. Unfortunately, properties differ too much to adequately represent these in a checklist. Section beginning \"silver standard\" - 10x is capitalised. Thank you for noticing, this has been corrected. Supplementary Figure 1 - legend fallen off panel a), needs a higher resolution or larger points We have increased the resolution. NMI definition - trailing parenthesis in denominator Thank you for noticing, this has been corrected. typo - assess the effect** Thank you for noticing, this has been corrected. Figure 2a - I found this quite busy, hard to interpret. Suggest to add shading that covers the points for same method or to facet by dataset. I don't believe the ARI values are particularly comparable between datasets so I would prefer facetting by dataset. We agree with the reviewer and now use faceting. Figure 3 - rows/columns are ordered differently between panels, what's driving this difference? The difference by clustering on the similarity across methods, i.e. more similar methods are closer to each other. We have included a statement explaining this in the figure description. Supplementary Figure 3 was not mentioned in the main text We now mention this Supplementary Figure. Supplementary Figure 4 is a two page pdf, with the first page blank We have corrected this error. Figure 6 - Figure caption says Dataset 1 but reports 29,151 genes. Do you mean the Gold Dataset and 29,451 genes? If not, please clarify which data and how many genes. Note that this figure has been replaced. We indeed meant Dataset 1, but with only half the genes. Discussion - One instance of \"Seurat\" is missing verbatim format Thank you for noticing, this has been corrected." } ] }, { "id": "37228", "date": "31 Aug 2018", "name": "Stephanie C Hicks", "expertise": [ "Reviewer Expertise Statistics", "genomics", "analysis of single-cell data" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nFreytag et al. have produced a nice research article on assessing methods for clustering scRNA-seq data from the 10x Genomics platform. I was excited to read the article to learn about what they recommend using. I have made some suggestions below for improvements that are mostly related to providing more intuition and higher-level summaries. This is mostly because as a user of these methods, at the end of the paper, I still felt a little confused about which method the authors would recommend using. I hope the authors can update the article with some of the suggestions:\nWhile the authors have provided detailed comparisons (running time, cluster stability, use of different aligners, different genes, etc), the biggest suggestion would be that the authors provide a higher-level summary of what the authors would suggest a user use to cluster his/her data. At the end of reading this paper, I felt a little overwhelmed at the amount of comparisons across various datasets. It's hard to look at Figs 1-7 and get an overall summary of which method to use. The authors do state in the abstract \"We found that some methods, including Seurat and Cell Ranger, outperform other methods, although performance seems to be dependent on the complexity of the studied system\", but it would be great if the authors could somehow provide a visual high-level summary of how they came to that conclusion, or elaborate in the discussion on that.\n\nFor the \"gold standard\" data, what was the percent of each human lung cell lines (HCC827, H1975, H2228) that were mixed together? Equal proportions? Was the reason you needed to use demuxlet was because the cell lines were mixed up for sequencing? It would be great if the authors could elaborate on the experimental design.\n\nIs the \"gold standard\" data available with the SNVs called for each cell. It would be useful to have this count matrix and corresponding phenotypic information about each cell in a SingleCellExperiment object for others to have access to.\n\nIt would be great if the authors could include another example dataset with a batch effect in it or something with a slightly less clean design, given most datasets are not quite this \"clean\". Also, maybe different clustering methods would perform better / worse depending on they data contained rare vs common cell types or included more or less diversity.\n\nThere is a TENxPBMCsData package (https://github.com/kasperdanielhansen/TENxPBMCData) that has been submitted to Bioconductor (similar to the TENxBrainData). This includes all PBMC 10X datasets currently listed on their site and loads in a SingleCellExperiment object into R. For the Silver Standard Datasets, you might incorporate this into your workflow.\n\nHow did you (or Cell Ranger) deal with empty droplets or swapped barcodes on the 10x platform? This seems relevant for discovering cell types using some form of clustering.\nSupplemental Table 1 could use a caption and a label at the top saying \"Supplemental Table 1\". I had many tabs open with different supplemental figures and tables, and was getting confused about which was which one.\n\nWhy did Linnorm and Monocle \"continually failed to run\"? Did the authors contact the original authors of Linnorm and Monocle to determine if there was a problem with the actual software or if it was a problem with the implementation of the software? It would be great if the authors could elaborate.\n\nI agree with this statement: \" We concede that it is possible that more care in the upstream data handling and selection of parameters could result in different results.\" This is true for almost all benchmarking papers. Given the authors are working within the R/Bioconductor framework, it would be great if the authors could use something like SummarizedBenchmark (http://bioconductor.org/packages/release/bioc/vignettes/SummarizedBenchmark/inst/doc/SummarizedBenchmark.html) to keep track of these parameters.\n\nCould the authors elaborate on how they decided which performance metrics to use?\n\nWhat does this mean: \"The impact of different aligners and preprocessing was assessed using all appropriate combinations of programs\"? Could the authors be more specific?\nI'm a little concerned about how much the solutions differ between methods and parameter choices. I understand the point of this paper is to make comparisons between already published methods, but as the authors are now very familiar with these methods, it would be great if they could provide some more practical guidance. What would the authors suggest using?\n\nFig 1 -- Could the authors hypothesize on why Seurat, TSCAN, RCA, SC3, RaceID, RaceID2 are estimating so many clusters? Also, why does countClust tend to underestimate the number of clusters? It would be great if the authors could provide some intuition.\n\nFig 3 -- If I'm understanding, ascend and countClust produce clusters that are very different than the rest?\n\nThank you to the authors for making their code publicly available!\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4300", "date": "19 Dec 2018", "name": "Saskia Freytag", "role": "Author Response", "response": "We would like to thank the reviewer for reviewing our manuscript and for their constructive comments. Below are point-by-point responses to the individual comments. While the authors have provided detailed comparisons (running time, cluster stability, use of different aligners, different genes, etc), the biggest suggestion would be that the authors provide a higher-level summary of what the authors would suggest a user use to cluster his/her data. At the end of reading this paper, I felt a little overwhelmed at the amount of comparisons across various datasets. It's hard to look at Figs 1-7 and get an overall summary of which method to use. The authors do state in the abstract \"We found that some methods, including Seurat and Cell Ranger, outperform other methods, although performance seems to be dependent on the complexity of the studied system\", but it would be great if the authors could somehow provide a visual high-level summary of how they came to that conclusion, or elaborate in the discussion on that. We have added a discussion section in which we summarize the results across all evaluations. This discussion section includes a visual high-level summary (Figure 9). For the \"gold standard\" data, what was the percent of each human lung cell lines (HCC827, H1975, H2228) that were mixed together? Equal proportions? Was the reason you needed to use demuxlet was because the cell lines were mixed up for sequencing? It would be great if the authors could elaborate on the experimental design. We mixed the cell lines in equal proportions. Due to using 10x Genomics technology, the cell lines were mixed up in the process but could be deconvoluted using demuxlet (ref?). We have elaborated on this further in the manuscript to clarify the experimental design. Is the \"gold standard\" data available with the SNVs called for each cell. It would be useful to have this count matrix and corresponding phenotypic information about each cell in a SingleCellExperiment object for others to have access to. We have made all datasets as SingleCellExperiment objects, including their phenotypic information, available on Github at https://github.com/bahlolab/cluster_benchmark_data . We have added information regarding the availability of all processed datasets to the manuscript. It would be great if the authors could include another example dataset with a batch effect in it or something with a slightly less clean design, given most datasets are not quite this \"clean\". Also, maybe different clustering methods would perform better / worse depending on they data contained rare vs common cell types or included more or less diversity. We agree that investigating the performance of clustering approaches on ”messy” scRNA-seq designs would be very interesting. However, this is beyond the scope of this paper, as it requires the application of sophisticated batch correction methods. Such methods should generally be performed by experts rather than beginners, who were the target audience of this paper. We have added a discussion to this effect. Finally, in order to investigate whether methods perform better or worse in more or less diverse situations, one requires either simulations or mixture experiments. These were beyond the scope of this paper. However, we now refer the readers of our manuscript to the recent benchmarking study of scRNA-seq clustering methods by Duó et al, which investigates just such a scenario. For most methods they did not observe overt differences. There is a TENxPBMCsData package (https://github.com/kasperdanielhansen/TENxPBMCData) that has been submitted to Bioconductor (similar to the TENxBrainData). This includes all PBMC 10X datasets currently listed on their site and loads in a SingleCellExperiment object into R. For the Silver Standard Datasets, you might incorporate this into your workflow. We decided to incorporate all moderately large fresh PBMC samples included in the TENxPBMCs into our workflow. This also provided us with an opportunity to update the package versions for the individual clustering tools for our silver standard benchmarking and stability analyses. How did you (or Cell Ranger) deal with empty droplets or swapped barcodes on the 10x platform? This seems relevant for discovering cell types using some form of clustering. We added the following explanation: “Cell Ranger filters any barcode that contains less than 10% of the 99th percentile of total UMI counts per barcode, as these are considered to be barcodes associated with empty droplets. The barcode by design can take one of 737,000 different sequences that comprise a whitelist. This feature allows the performance of error correction when the observed barcode does not match any barcode on the whitelist due to sequencing error.” Supplemental Table 1 could use a caption and a label at the top saying \"Supplemental Table 1\". I had many tabs open with different supplemental figures and tables, and was getting confused about which was which one. We added a caption on the top of all Supplemental Tables. Why did Linnorm and Monocle \"continually failed to run\"? Did the authors contact the original authors of Linnorm and Monocle to determine if there was a problem with the actual software or if it was a problem with the implementation of the software? It would be great if the authors could elaborate. Linnorm failed because its calculations would time out. Monocle failed because the dispersion could not be calculated. However, neither of the programs was tried using their newer package versions corresponding to R version 3.5.0 nor were any of the packages’ authors contacted. We have included a statement in the manuscript to this regard. I agree with this statement: \" We concede that it is possible that more care in the upstream data handling and selection of parameters could result in different results.\" This is true for almost all benchmarking papers. Given the authors are working within the R/Bioconductor framework, it would be great if the authors could use something like SummarizedBenchmark (http://bioconductor.org/packages/release/bioc/vignettes/SummarizedBenchmark/inst/doc/SummarizedBenchmark.html) to keep track of these parameters. We did take a look at the SummarizedBenchmark package, but did not find it suitable for our needs. However, we understand the need to provide all parameters (including defaults) used in the individual analyses and thus have added additional files providing this information to the GitHub respository. Could the authors elaborate on how they decided which performance metrics to use? We used performance metrics commonly used in the clustering literature. We also made sure that the selected metrics were applicable in the absence of known cluster labels. Furthermore, they share the advantages of bounded ranges and no assumptions regarding cluster structures. Additionally they offer complementary insights. We have added this explanation to the manuscript. What does this mean: \"The impact of different aligners and preprocessing was assessed using all appropriate combinations of programs\"? Could the authors be more specific? We meant to say that we assessed the impact of combinations of different aligners and preprocessing (i.e. CellRanger or scPipe) for all possible clustering methods. Some clustering methods, like ascend, failed to run for scPipe generated output and it was too challenging to run the CellRanger clustering approach on scPipe generated output. I'm a little concerned about how much the solutions differ between methods and parameter choices. I understand the point of this paper is to make comparisons between already published methods, but as the authors are now very familiar with these methods, it would be great if they could provide some more practical guidance. What would the authors suggest using? We suggest using several clustering methods ideally with multiple parameter choices in order to ensure that biological results are not artifacts of method or parameter choice. Unfortunately, we do not feel in a position to give specific practical advice for the specific use of individual methods, as optimal parameter choices depend on many different factors including the type of biological system studied. Fig 1 -- Could the authors hypothesize on why Seurat, TSCAN, RCA, SC3, RaceID, RaceID2 are estimating so many clusters? Also, why does countClust tend to underestimate the number of clusters? It would be great if the authors could provide some intuition. We believe that many methods tended to overestimate the number of clusters in the gold standard dataset, because the cell lines may be heterogeneous with regards to other biological factors, such as cell state. Consequently, in such a scenario methods may split cells of the same population but in different cell states into multiple clusters. We have no intuition as to why countClust underestimates the number of clusters. Fig 3 -- If I'm understanding, ascend and countClust produce clusters that are very different than the rest? Yes that is correct." } ] } ]	1	https://f1000research.com/articles/7-1297
https://f1000research.com/articles/7-1950/v1	18 Dec 18	{ "type": "Case Report", "title": "Case Report: Synchronous adenoid cystic variant of basal cell carcinoma in the right lumbar region: case report of an incidental finding in a patient with breast carcinoma", "authors": [ "Sucheta Gandhe", "Rahul Patil", "Rajnish Nagarkar", "Sucheta Gandhe", "Rahul Patil" ], "abstract": "Skin cancer has emerged as a major problem for light-skinned people globally. Basal cell carcinoma (BCC) is one of the most common forms of skin cancer. BCC is associated with significant morbidity. There are multiple histological patterns of BCC, wherein the adenoid cystic variant is a rare form. A 75-year-old female with a history of breast carcinoma visited our centre for a routine follow-up. The patient was diagnosed with invasive breast carcinoma and underwent left breast conservation surgery in August 2018. At follow-up, the patient complained of itchiness, redness, and ulceration over a long-standing mole located at the right lumbar region. The lesion was excised and histopathologically diagnosed as the adenoid cystic variant of BCC. Adenoid cystic variant of BCC is an uncommon presentation. Identifying the mole in the lumbar region with clinical signs and symptoms was an incidental finding. In most cases, skin moles are benign. However, this case is of considerable interest as the patient presented with two primary cancers of different pathological characteristics within 3 months. The patient is currently doing well and is due for follow-up.", "keywords": [ "Basal cell carcinoma", "adenoid cystic", "non-melanoma skin cancer" ], "content": "Introduction\n\nSkin cancer has emerged as a major problem for light-skinned people worldwide. Skin tumours are classified based on the melanin concentration. There are two types of skin tumours, non-melanoma skin cancer (NMSC) and malignant melanomas (MM). As compared to NMSC, MM has a more aggressive modality with an estimated mortality of 85% for all fatal skin cancers1. The most common NMSCs include basal cell carcinoma (BCC), which constitutes approximately 80% of new cases, and squamous cell carcinoma (SCC), which constitutes approximately 20% of new cases1. BCC is derived from the basal cells while SCC is derived from the squamous cells of the epidermis. SCC is an aggressive form of skin cancer with a high risk of metastasis2. BCC is attributed with a slow growth tendency, high risk of recurrence, and a low mortality rate3. We present a case report synchronous adenoid cystic, a rare variant of BCC of right lumbar region in a patient with breast carcinoma.\n\n\nCase report\n\nA 75-year-old woman with previously diagnosed breast cancer came for a routine follow-up in November 2018 at HCG Manavata Cancer Centre, Nashik, India. The patient was diagnosed with Stage IIA (T2N0) breast cancer in August 2018. The patient underwent left breast conservation surgery in August 2018. She was referred for radiotherapy. She completed radiotherapy (52.7 Gy in 20 fractions) for 20 days in September 2018. At follow-up, the patient complained of itchiness, redness, and ulceration on the long-standing mole. The lesion measured 1.4×0.6 cm. The patient underwent wide local excision at our centre. The lesion was sent for haematoxylin and eosin (H&E) histopathological examination.\n\n\nDiagnostic assessment\n\nThe multiple sections that were examined showed ulcerated epidermis. The dermis was observed to have a tumour with basaloid pattern and peripheral pallisidation, arising from the basal layer of the epidermis. Multiple punched out cystic spaces were observed. Multiple pigment laden macrophages were also observed. The mitotic rate was 9–10 per HPF). Necrosis was not observed. Perineural invasion or lympho-vascular emboli were not observed. Morphological examination of the cells resulted in a diagnosis of the adenoid cystic variant of BCC (Figure 1 and Figure 2). The patient is under close observation.\n\n\nDiscussion\n\nAdenoid cystic variant of BCC is a rare histological entity. Furthermore, ADBCC in the lumbar region is an unusual site of presentation. The diagnosis of ADBCC is important considering its poor prognosis as compared to other variants of BCC. Skin moles are usually benign in nature. However, skin mole presenting with clinical symptoms should not be ignored in patients with already diagnosed malignancy. Our multidisciplinary team of experts ensure robust and comprehensive patient care services. Due to our strengths, the second primary malignancy (ADBCC) was diagnosed early.\n\nClinicians should not ignore patients with different clinical manifestations other than those reported initially. Assessment of new lesions reported by patients at follow-up should be a norm in clinical practice.\n\nIt is essential to report such cases in the literature as it helps disseminate knowledge on synchronous malignancies.\n\nBCC is the most commonly reported skin cancer4. It is neither a lethal nor a metastatic disease. However, BCC is characterized by significant morbidity which is secondary to local invasion or destruction4. Based on current evidence, the incidence of BCC is 226 per 100,000 people while the age-adjusted prevalence is 343 per 100,000 people4. The major risk factor for BCC is ultraviolet light exposure. Other risk factors for BCC include ultraviolet A or psoralen therapy, immunosuppressive medications, radiation therapy, and chronic arsenic toxicity4.\n\nThe increase in incidence rates of BCC is attributed with lifestyle or environmental changes along with behavioural risk factors. Exposure to UV light, specifically UVB induces mutations in tumour suppressor genes which play a key role in the overall pathogenesis of BCC5. Younger age and history of blistering sunburn has been associated with BCC. However, intermittent or continuous exposure throughout life as a risk factor remains unclear to cause BCC5. Exposure to UV radiation early in life should be considered as a key risk factor as compared to overall cumulative exposure5.\n\nBCC is a common, locally invasive epithelial malignancy of the skin and appendages. An estimated 10 million people are diagnosed with BCC every year worldwide6. The histology of BCC is predictable while certain variants are rare to observe. Some of the rare histological variants include adenoid, cystic, morpheaform, pigmented, infundibulocystic, and miscellaneous variants. These variants account for nearly less than 10% of all BCCs6. Adenoid BCC (ADBCC) is a rare histopathological variant with an incidence of approximately 1.3%6. The clinical appearance of the adenoid lesion can be pigmented or non-pigmented nodule or an ulcer without predilection for any particular site6.\n\nADBCC is an extremely rare, low-grade, and differentiated malignancy. It is reported at various sites such as the inner canthus of the eye, leg, axillae, back, chin, and forehead6. It has also been reported in the cervix and prostate6. The rarity of this lesion is evidenced by the paucity of data in the literature. It is important for pathologists to differentiate and make an accurate diagnosis of ADBCC as it often mimics primary cribriform apocrine carcinoma or cutaneous adenoid cystic carcinoma6. A similar case of BCC and breast cancer in a single patient was reported in 20147. It is essential to make an accurate diagnosis of ADBCC due to the prognostic differences of these conditions. By writing up this case we hope to contribute to the body of literature about ADBCC, in particular its unique histological characteristics. The treatment of BCC includes surgical and non-surgical options such as excision with primary closure, grafts, and radiotherapy8. In our case, the patient had undergone wide local excision8.\n\n\nConclusion\n\nAdenoid cystic variant of BCC is a rare histological entity. We report synchronous adenoid cystic variant of BCC of the right lumbar region in a patient with breast cancer. Clinicians should be vigilant about patients with different clinical manifestations. Multidisciplinary support was our core strength in assessing the patient with a new lesion which in turn led to an incidental finding of a rare skin cancer.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\n\nConsent\n\nWritten informed consent for publication of their clinical details and images was obtained from the patient.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgments\n\nWe would like to thank Mr. Lyndon Fernandes for his editorial assistance.\n\n\nReferences\n\nLihachev A, Lihacova I, Plorina EV, et al.: Differentiation of seborrheic keratosis from basal cell carcinoma, nevi and melanoma by RGB autofluorescence imaging. Biomed Opt Express. 2018; 9(4): 1852–1858. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKang KW, Lee DL, Shin HK, et al.: A Retrospective Clinical View of Basal Cell Carcinoma and Squamous Cell Carcinoma in the Head and Neck Region: A Single Institution's Experience of 247 Cases over 19 Years. Arch Craniofac Surg. 2016; 17(2): 56–62. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMackiewicz-Wysocka M, Bowszyc-Dmochowska M, Strzelecka-Węklar D, et al.: Basal cell carcinoma - diagnosis. Contemp Oncol (Pozn). 2013; 17(4): 337–42. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAwad R, Andrade JCB, Mousa H, et al.: Invasive Basal Cell Carcinoma of the Skin Treated Successfully with Vismodegib: A Case Report. Perm J. 2018; 22: 17–181. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTotonchy M, Leffell D: Emerging concepts and recent advances in basal cell carcinoma [version 1; referees: 2 approved]. F1000Res. 2017; 6: 2085. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSaxena K, Manohar V, Bhakhar V, et al.: Adenoid basal cell carcinoma: a rare facet of basal cell carcinoma. BMJ Case Rep. 2016; 2016: bcr2015214166. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMorelle A, Cericatto R, Krepischi AC, et al.: Clinical and genetic characterization of basal cell carcinoma and breast cancer in a single patient. SpringerPlus. 2014; 3: 454. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWong CS, Strange RC, Lear JT: Basal cell carcinoma. BMJ. 2003; 327(7418): 794–798. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "50650", "date": "04 Jul 2019", "name": "Semir Vranic", "expertise": [ "Reviewer Expertise Pathology" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors should provide more info on the primary breast cancer; what was the status of ER, PR and Her2 in primary breast cancer? It is unusual that primary breast cancer (pT2No) is only treated by radiotherapy.\nWhat was the immunohistochemical profile of ACC variant of BCC? This must be clearly provided. There are also rare variants of primary ACC of the skin as well as ACC of the breast (including basaloid variants that are usually high grade cancers). It is not easy to appreciate mitotic figures in skin cancer (I am a pathologist).\n\nIs the background of the case’s history and progression described in sufficient detail? No\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? No\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? No\n\nIs the case presented with sufficient detail to be useful for other practitioners? No", "responses": [] }, { "id": "53428", "date": "09 Sep 2019", "name": "Mariam Totonchy", "expertise": [ "Reviewer Expertise cutaneous oncology", "non-melanoma skin cancer", "melanoma", "cutaneous/dermatologic surgery and reconstruction" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nOverall, I think this is an interesting case presentation of a rare subtype of basal cell carcinoma, however, I think there are several points that need to be improved upon.\n\nThere are several statements that should be amended. For example, BCC is the most common form of skin cancer, and it would be best to state this directly in the introduction. It is important to correct that skin tumors are not classified based on melanin concentration but on the identified cell of origin. Compared to other tumors, cutaneous SCC does not have a high risk of metastasis (~1.9-4%) so I would be careful about the wording there. Also, compared to other cutaneous malignancies, BCC does not have a high risk of recurrence.\nI would include additional important details regarding the patient, including her Fitzpatrick skin type, type of breast cancer, and if the location of the BCC was within the radiation field (unlikely given the description, but would be important to know). Follow-up information would also be helpful to report on the patient. I would also include in the discussion any demographic information that is available on patients with adenoid cystic BCCs as reported in the literature.\nYou state that the adenoid cystic variant of BCC emerged from a pre-existing mole, however, no collision lesion was identified in the description of the histopathology. It would be interesting to the overall discussion if a nevus was identified adjacent to the BCC as this may have important implications. Regardless of whether or not a nevus was identified on pathology, this should be commented upon and is important for the overall description of the case. BCCs are often pigmented in patients with darker skin types, so this would be an important point to clarify.\nI do not think that there is enough evidence presented in this report or in the literature to suggest that adenoid cystic BCC has a poor prognosis as compared to BCC. I would remove this statement or provide additional evidence to back the claim.\n\nIs the background of the case’s history and progression described in sufficient detail? Partly\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Partly\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Partly\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1950
https://f1000research.com/articles/7-1949/v1	18 Dec 18	{ "type": "Research Article", "title": "Intra-canal medication containing silver nanoparticle versus calcium hydroxide in reducing postoperative pain: A randomized clinical trial", "authors": [ "Fatma El Zahraa El Abbasy", "Salsabyl Ibrahim", "Olfat Shaker", "Geraldine Ahmed", "Salsabyl Ibrahim", "Olfat Shaker", "Geraldine Ahmed" ], "abstract": "Background: Pain of endodontic origin can be annoying for patients and endodontist. Pain relief is more important to the patient than treatment success. Numerous factors such as over instrumentation, over filling, debris extrusion can cause postoperative pain. However, bacteria found in the root canal space is the most important factor. Therefore mechanical preparation is an important step in elimination of micro-organisms from the root canal. It has been reported that micro-organisms can still survive inside the root canal even after mechanical preparation. Hence, the use of intra-canal medicaments in between visits for reduction of bacteria inside the root canal space has been recommended. The aim of this study was to assess the ability of silver nanoparticles versus calcium hydroxide used as intra-canal medication in reducing pain in necrotic teeth with apical periodontitis. Methods: Thirty-four participants were randomly divided into 2 groups, 17 in each group according to intra-canal medication used silver nanoparticles and calcium hydroxide (AgNPs and Ca(OH)2). Each patient was given pain scale chart numerical rating scale (NRS) in order to record his/her pain level before any intervention followed by placement of intra-canal medicament for 1 week. Postoperative pain was recorded at 4, 12, 24, 48 hours. Results: Pre-operatively; there was no statistically significant difference between mean pain scores in the two groups. After 4, 12 as well as 24 hours, Ca(OH)2 group showed statistically significantly higher mean pain score than AgNPs group. After 48 hours; there was no statistically significant difference between mean pain scores in the two groups. Conclusions: There was a statistically significant difference in postoperative pain following 4, 12, and 24 hours where AgNPs group resulted in reduction of pain more than Ca(OH)2 group. At 48 hours, there was no statistically significant difference. Trial registration: PACTR PACTR201602001444180 26/01/2016", "keywords": [ "calcium hydroxide", "silver nanoparticles", "intra-canal medication", "post-operative pain" ], "content": "Introduction\n\nPain in endodontic treatment is a major concern for patients and clinicians. Many factors can cause pain like bacteria, chemical mediators, change in cyclic mediators, change in periapical tissue pressure, and psychological factors. However, the presence of microorganisms as a result of failure to properly disinfect the canal is the most common cause of pain. Infected root canals contain a wide range of microorganisms, which may result in the production of enzymes and endotoxins resulting in persistence of pain1.\n\nCalcium hydroxide is the most commonly used intra-canal medicament and is considered the gold standard. However, the ability of calcium hydroxide medication to completely eradicate bacterial species from the root canal has been questioned2. Also, Al-Zaka (2007)3 found that chlorhexidine intra-canal medicament significantly reduced postoperative pain more than calcium hydroxide medication ,where at 4 hours, 20 of 100 patients in chlorhexidine group had no pain compared to only 4 of 100 patients in calcium hydroxide group, and at 24 hours, 44 of 100 patients in chlorhexidine group had no pain compared to 24 in calcium hydroxide group.\n\nRecently, nano-technology has attracted attention, especially those containing silver4. Silver nanoparticles are advantageous as they are biocompatible and have antimicrobial activity5, silver ions can cause damage to bacterial cell wall. The antibacterial properties of silver nanoparticles has been demonstrated4 against both gram- negative as well as gram- positive bacteria, fungi and viruses. Therefore, the aim of the present study was to compare calcium hydroxide and silver nanoparticles in terms of postoperative pain.\n\n\nMethods\n\nEach of the protocol, informed consent form, and recruitment materials was reviewed and approved by the Ethical review Committee, Faculty of Dentistry, Cairo University. The approval number is 1637.\n\nPatients were able to discuss anything with the researcher and the whole trial was explained to the patient. The researcher received a written consent from the patients that accepted to participate in the trial prior to starting any dental treatment.\n\nThe trial was registered with the Pan African Clinical Trials Registry (PACTR) on 26 January 2016 - PACTR201602001444180, trial name: comparison between two intra-canal medicament, calcium hydroxide and silver nanoparticle.\n\n34 participants, age range from 18–55 years, males or females, single rooted maxillary or mandibular teeth with: non-vital response of pulp tissue, tenderness to percussion, slight widening of lamina dura, were included in this study, (17 participants in the experimental group (silver nanoparticles: Nanotech, Dreamland Egypt, for its preparation, polyethylene glycol was used to formulate the preparation containing 75% (W/W) silver nanoparticles suspension6) and 17 participants in the control group (calcium hydroxide: Metapaste, Meta Biomed, CO, LTD, Korea #107677) representing Egyptian population which were recruited from Endodontic clinic, Faculty of Dentistry, Cairo University, between January 2017–January 2018. The aim of the study, treatment procedures, possible side effects and treatment alternatives were explained to the patients. Patients were asked to follow the general instructions and sign a printed informed consent that explained the aim of the study, and asked the patient to fill the outcomes data charts after the procedures, honestly and accurately and return them to us.\n\nPatients aging more than 55 years old or less than 18 years old, medically compromised patients, patients who had received antibiotic treatment during the last 3 months, pregnant females, teeth that couldn't be isolated with rubber dam, teeth with abscess, fistula or sinus tract, previously root canal treated teeth, and teeth with periodontal probing depth more than 4 mm were excluded from the study.\n\nBased on the previous paper by Singh et al. 2013, the sample size was calculated considering that a minimal clinical difference of 10 in numerical rating scale (NRS) score between two study group was clinically relevant. Using a power of 80%, a level of significance of 5% and considering a standard deviation of 9.0, 11 patients per group would be necessary. This number was increased to 13 to adjust for non-parametric usage and increased again to 17 in each group to compensate for losses during follow up. The sample size was calculated using the PS program (3.1.2 / August 2014).\n\nParticipants. Clinical examination was done tentatively using a diagnostic mirror and probe, presence of extensive caries or large restoration was detected, percussion with the back of a diagnostic mirror and palpation with the index finger to indicate the presence of any swelling or tenderness were done. An electric pulp tester (Denjoy DY310, Henan) was used to record the response of the affected tooth, with the adjacent and contra-lateral teeth used as a control. Preoperative pain was recorded, with each patient given pain scale chart (NRS) in order to record his/her pain level before any intervention (Extended data7). The pain scale (0–10 scale) consists of a line anchored by two extremes, \"No pain\" and \"the worst pain\", patients were asked to choose the mark that represented their level of pain from 0 to 10. Pain level was assigned as follows:\n\n0 reading represents \"no pain\"\n\n1–3 readings represent \"mild pain\"\n\n4–6 readings represent \"moderate pain\"\n\n7–10 readings represent \"severe pain\".\n\nRadiographic examination was done with a Phosphor Storage Plate (PSP) (Sordex (DIGORATM Optime DXR-60, Finland) wireless sensor taken using the bisecting angle technique to detect the presence of any widening of the periodontal membrane space. Final diagnosis revealed necrotic mandibular or maxillary single rooted teeth with apical periodontitis.\n\nSequence of endodontic clinical procedures. Local anesthesia: tooth was anaesthetized using nerve block or infiltration technique by selected local anesthesia (Articaine HCI 4% and Adrenaline 1:100,000).\n\nIsolation: teeth were properly isolated with rubber dam.\n\nAccess cavity preparation: a two stage access cavity preparation was performed. The first stage involved the removal of caries without exposing the pulp chamber using a high speed round diamond (Dentsply, Tulsa Dental, Dentsply Maillefer, USA) bur size 2. In the second stage, gaining access with complete de-roofing using a new round diamond bur size 2 was carried out.\n\nRoot canal instrumentation: the working length was checked with apex locator (Dent Port Zx J. Morita, Irvine, Japan) and confirmed by digital radiography.\n\nCoronal pre-flaring was performed using Gates-Glidden drills (Mani, Tochigi, Japan) sizes 2, 3 and 4.\n\nRoot canal preparation was then done using ProTaper Next rotary (Dentsply, Tulsa Dental, Dentsply Maillefer, TN, USA) system in the following sequence (X1, X2, X3, X4, X5) using an endodontic motor (Dentsply Maillefer, USA) with adjusted torque and speed according to the manufacturer's instructions using 2.5% sodium hypochlorite (Household cleaning products of Egypt, 10th of Ramadan) and EDTA gel (MD-Chelcream, Meta Biomed Co Ltd, Korea) between each file.\n\nIntra-canal medication and primary coronal seal:\n\n❖ The canals were dried with paper points (Dentsply Maillefer, Ballaigues, Switzerland) and plugged with selected intra-canal medication.\n\n❖ The access cavities were properly filled with glass ionomer filling to ensure proper sealing with no leakage of any oral fluids inside the root canal.\n\nThe patients were given the pain scale chart (NRS) to record their postoperative pain at 4, 12, 24, & 48 hours and return it back on the 2nd visit. Obturation was then carried out in the 2nd visit which was 7 days following the 1st visit.\n\n\nMethods: assignment of interventions\n\na- Sequence generation\n\nA random sequence was generated by computer software (http://www.random.org/), in the center of evidence-based dentistry (EBD), Cairo University.\n\nThe table was kept with the assistant supervisor (G.A.).\n\nb- Allocation concealment mechanism\n\nNumbered papers indicating the intra-canal medicament to be used were packed in opaque closed envelopes by G.A., where the patient picked up an envelope. After mechanical preparation, operator F.O. opened the envelope and used the intra-canal medicament assigned to that patient according to the number present inside the envelope.\n\nc- Implementation\n\nThe assistant supervisor G.A. was the one responsible for the generation of random sequence, assigned the patients to the intervention or control group and the only one who knew which medicament was used for each patient.\n\nd- Blinding\n\n-The study was double-blinded (the participants and the assessor).\n\n-Participants did not know which group they were treated with after they chose the opaque envelope which contained the number indicating which intra-canal medicament to be used.\n\n-Assessor, who assessed all results data, did not know which group the participants were related to.\n\nIf any harm was seen in the participants either in intervention or control groups they were recorded and reported at the end of the trial. The treatment according to the harm:\n\n- Pain: administration of Analgesics. (cataflam 60 mg, 1 tablet when needed).\n\n- Swelling: hot fomentation, mouth rinse with salty warm water, antibiotic administration in-case of presence of fever or lymphadenopathy.\n\n- Allergic reaction: referral for a physician for corticosteroid therapy. (Prednisolone 40 mg, 1 tablet/day for 3-10 days)\n\nNumerical data were explored for normality by checking the distribution of data and using tests of normality (Kolmogorov-Smirnov and Shapiro-Wilk tests). Age data showed normal (parametric) distribution, while pain data showed (non-parametric) distribution. Data were presented as mean, median, standard deviation (SD), minimum, maximum and 95% confidence interval (95% CI) for the mean values.\n\nFor parametric data; Student’s t-test was used to compare between mean age values in the two groups. For non-parametric data; Mann-Whitney U test was used to compare between the two groups. Friedman's test was used to study the changes by time in each group. Dunn’s test was used for pair-wise comparisons when Friedman's test is significant.\n\nQualitative data were presented as frequencies and percentages. Chi-square test (or Fisher’s Exact test when applicable) were used for comparisons regarding qualitative data. The significance level was set at P ≤ 0.05. Statistical analysis was performed with IBM® SPSS® Statistics Version 20 for Windows.\n\n\nResults\n\nThere was no significant difference between the two groups in terms of age (p=0.539) and no significant difference between the two groups in terms of gender distribution (p=1.000), Table 1.\n\nThe intensity and incidence of preoperative pain and postoperative pain at the predetermined periods were formulated as follows.\n\nComparison between the two groups regarding preoperative pain and postoperative pain at each time point: Table 2, Figure 1. Pre-operatively; there was no statistically significant difference between mean pain scores in the two groups. After 4, 12 as well as 24 hours, Ca(OH)2 group showed statistically significant higher mean pain score (P<0.001) than AgNPs group. After 48 hours; there was no statistically significant difference between mean pain scores in the two groups.\n\n: Significant at P ≤ 0.05, Different superscripts in the same column are statistically significantly different\n\nComparison between the two groups regarding intensity and incidence of preoperative and postoperative pain at each time point: Table 3, Figure 2. Pre-operatively; there was no statistically significant difference between intensity of pain in the two groups. After 4 hours, there was a statistically significant difference between the two groups. Ca(OH)2 group showed lower prevalence of mild pain and higher prevalence of moderate pain than AgNPs group. After 12 hours, there was a statistically significant difference between the two groups. All cases in Ca(OH)2 group showed severe pain while AgNPs group showed higher prevalence of mild and moderate pain and no cases with severe pain. After 24 as well as 48 hours; there was no statistically significant difference between incidences of postoperative pain in the two groups.\n\n: Significant at P ≤ 0.05\n\n\nDiscussion\n\nEffective control of intra-canal microbial load before obturation is the key element to the high success of root canal treatment8. Bacteria can be eliminated from the root canal space by mechanical instrumentation of the root canal which resulted in 50% reduction9,10 in endotoxins level in infected root canals. Endotoxins are recognized as foreign by the immune system which will result in a potent immune response resulting in 11 release of pro-inflammatory cytokines which lead to pain and the development of apical periodontitis12. Numerical rating scale (NRS) is a subjective method for scoring preoperative and postoperative pain. This is a valid and reliable method which has been widely used in endodontic literature,1,13 owing to its standardized assessment measure and reported better compliance when compared to other scales. In addition, it was preferred by the majority of patients amongst different demographic backgrounds14,15. Preoperative pain was recorded for each patient before starting treatment as it was considered a risk factor that can affect the postoperative pain16,17. Although calcium hydroxide is considered the gold standard for disinfecting the root canals, and has been widely used as an intra-canal medicament since 1920's18, its role in eliminating bacteria associated with apical infections is controversial. Microorganism have the ability to invade the dentinal tubules and buffer the high pH produced by calcium hydroxide compromising its antimicrobial activity19. Nanoparticles such as metallic, polymeric and bioactive were recently used as antimicrobial agents in medical and dental fields20,21. Silver nanoparticles bind to the negatively charged bacterial cell membrane, which affect the permeability and respiration of the bacteria causing its rupture and death19. Silver nanoparticles release silver ions which lead to bacterial death when it comes in contact with an aqueous media22,23. Silver nanoparticles intra-canal medicament was used in the present study as its antibacterial effect was the most commonly considered in the recent literature than the other types19. The incidence and intensity of postoperative pain, was recorded at 4, 12, 24, 48 hours after placement of intra-canal medication. Control group calcium hydroxide showed statistically significant higher mean pain score than silver nanoparticles group. After 48 hours, there was no statistically significant difference between mean pain scores in the two groups. There was a statistically significant increase in mean pain score from 12 to 24 hours, as from 24 to 48 hours there was a statistically significant decrease in mean pain score. Regarding the intensity of postoperative pain at the predetermined time periods, the two groups showed a statistically significant difference at 4, 12, 24 hours, whereas at 48 hours, the results were not statistically significant. Comparing preoperative and postoperative pain between both groups showed no statistically significant difference, neither in intensity nor incidence of pain at different follow up periods. In the present study, the assumption that patients who have acute preoperative pain are likely to experience more severe postoperative pain24–26 pre-operatively; as pain was present in all patients in both groups. However, there was no statistically significant difference between mean pain scores between the two groups. Silver in the form of nanoparticles has been used in various forms for treating burns and severe bacterial infected wounds and injuries and as an antimicrobial agent27. In this scenario the materials in nano-scale have been used as an anti-microorganism agent due to their physical and chemical properties26. Silver nanoparticles when used in small doses and with particles at 10 nm, appeared to have anti-inflammatory characteristics, and speed up the healing process of wounds, as well as modulate cytokines, and the induction and production of peripheral blood cells28. Calcium hydroxide showed higher mean pain score at 4, 12 and 24 hours compared to silver nanoparticles group. This may be attributed to the combined effect of silver nanoparticles as being anti-inflammatory and anti-microbial27. Calcium hydroxide release hydroxyl ions which increase alkalinity leading to death of bacteria and microorganisms inside the root canal19. This could explain the reduction of postoperative pain at 48 hour time period, however, Anjaneyulu and Nivedhitha29, concluded in their systematic review that the effect of calcium hydroxide in reducing postoperative pain can be increased when it is combined with other medications such as camphorated mono-chlorophenol or chlorhexidine and the use of calcium hydroxide alone isn’t effective to reduce postoperative pain. Orstavik and Hapassalo30, Yesiloy et al.31 found that certain bacteria present in the root canal system were resistant to the high pH environment produced by calcium hydroxide. Future studies can be directed to identify the the type of bacteria present in root canals, especially the harbored bacteria along full length of dentinal tubules to define the proper effect of the intra-canal medication used.\n\n\nConclusions\n\nThere was a statistically significant difference in postoperative pain following 4, 12, and 24 hours where AgNPs group resulted in reduction of pain more than Ca(OH)2 group. At 48 hours, there was no statistically significant difference.\n\n\nData availability\n\nUnderlying data is available from figshare\n\nFigshare: Dataset 1. Silver nanoparticle versus calcium hydroxide intra-canal medication https://doi.org/10.6084/m9.figshare.74413317\n\nLicence: CC0 1.0 Universal (CC0 1.0) Public Domain Dedication\n\nFigshare: Extended data. pain scale chart https://doi.org/10.6084/m9.figshare.74413317\n\nLicence: CC0 1.0 Universal (CC0 1.0) Public Domain Dedication\n\n\nReporting guidelines\n\nFigshare: CONSORT checklist and flowchart for article ‘Intra-canal medication containing silver nanoparticle versus calcium hydroxide in reducing postoperative pain: A randomized clinical trial’ https://doi.org/10.6084/m9.figshare.74413317\n\nLicence: CC0 1.0 Universal (CC0 1.0) Public Domain Dedication", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nSingh RD, Khatter R, Bal RK, et al.: Intracanal medications versus placebo in reducing postoperative endodontic pain--a double-blind randomized clinical trial. Braz Dent J. 2013; 24(1): 25–29. PubMed Abstract \| Publisher Full Text\n\nSathorn C, Parashos P, Messer H: Antibacterial efficacy of calcium hydroxide intracanal dressing: a systematic review and meta-analysis. Int Endod J. 2007; 40(1): 2–10. PubMed Abstract \| Publisher Full Text\n\nAl-Zaka IM: The incidence of posttreatment pain using two different intracanal medicaments. MDJ. 2007; 4(2): 110–116. Reference Source\n\nCheng L, Zhang K, Melo MA, et al.: Anti-biofilm dentin primer with quaternary ammonium and silver nanoparticles. J Dent Res. 2012; 91(6): 598–604. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMotshekga SC, Ray SS, Onyango MS, et al.: Microwave-assisted synthesis, characterization and antibacterial activity of Ag/ZnO nanoparticles supported bentonite clay. J Hazard Mater. 2013; 262: 439–446. PubMed Abstract \| Publisher Full Text\n\nBruniera JF, Silva-Sousa YT, Lara MG, et al.: Development of intracanal formulation containing silver nanoparticles. Braz Dent J. 2014; 25(4): 302–306. PubMed Abstract \| Publisher Full Text\n\nElAbbasy FE, Ibrahim S, Shaker O, et al.: Silver nanoparticle versus calcium hydroxide intra-canal medication. 2018. http://www.doi.org/10.6084/m9.figshare.7441331.v1\n\nSequeira JF Jr, Rôças IN: Clinical implications and microbiology of bacterial persistence after treatment procedures. J Endod. 2008; 34(11): 1291–1301.e3. PubMed Abstract \| Publisher Full Text\n\nGomes BP, Martinho FC, Vianna ME: Comparison of 2.5% sodium hypochlorite and 2% chlorhexidine gel on oral bacterial lipopolysaccharide reduction from primarily infected root canals. J Endod. 2009; 35(10): 1350–1353. PubMed Abstract \| Publisher Full Text\n\nVianna ME, Horz HP, Conrads G, et al.: Effect of root canal procedures on endotoxins and endodontic pathogens. Oral Microbiol Immunol. 2007; 22(6): 411–418. PubMed Abstract \| Publisher Full Text\n\nBaik JE, Kum KY, Yun CH, et al.: Calcium hydroxide inactivates lipoteichoic acid from Enterococcus faecalis. J Endod. 2008; 34(11): 1355–1359. PubMed Abstract \| Publisher Full Text\n\nJacinto RC, Gomes BP, Shah HN, et al.: Quantification of endotoxins in necrotic root canals from symptomatic and asymptomatic teeth. J Med Microbiol. 2005; 54(Pt 8): 777–783. PubMed Abstract \| Publisher Full Text\n\nKhattak YK, Shah SA, Alam F: Comparison of inter-appointment pain between calcium hydroxide mixed with 2% chlorhexidine and calcium hydroxide mixed with normal saline; a randomized controlled trial. JKCD. 2014; 5(1): 33–37. Reference Source\n\nPaice JA, Cohen FL: Validity of a verbally administered numeric rating scale to measure cancer pain intensity. Cancer Nurs. 1997; 20(2): 88–93. PubMed Abstract \| Publisher Full Text\n\nCaraceni A, Cherry N, Fainsinger R, et al.: Pain measurement tools and methods in clinical research in palliative care: recommendations of an Expert Working Group of the European Association of Palliative Care. J Pain Symptom Manage. 2002; 23(3): 239–255. PubMed Abstract \| Publisher Full Text\n\nSadaf D, Ahmad MZ: Factors associated with postoperative pain in endodontic therapy. Int J Biomed Sci. 2014; 10(4): 243–247. PubMed Abstract \| Free Full Text\n\nAlRahabi MK: Predictors, prevention, and management of postoperative pain associated with nonsurgical root canal treatment: A systematic review. J Taibah Univ Med Sci. 2017; 12(5): 376–384. Publisher Full Text\n\nSiqueira JF Jr, Lopes HP: Mechanisms of antimicrobial activity of calcium hydroxide: a critical review. Int Endod J. 1999; 32(5): 361–369. PubMed Abstract \| Publisher Full Text\n\nHaapasalo HK, Sirén EK, Waltimo TM, et al.: Inactivation of local root canal medicaments by dentine: an in vitro study. Int Endod J. 2000; 33(2): 126–131. PubMed Abstract \| Publisher Full Text\n\nNam W, Park SH, Choi GW: The effect of calcium hydroxide on post-treatment pain. J Korean Acad Conserv Dent. 2006; 31(2): 86–95. Publisher Full Text\n\nGuerreiro-Tanomaru JM, Pereira KF, Nascimento CA, et al.: Use of nanoparticulate zinc oxide as intracanal medication in endodontics: pH and antimicrobial activity. Acta Odontol Latinoam. 2013; 26(3): 144–8. PubMed Abstract\n\nMorones JR, Elechiguerra JL, Camacho A, et al.: The bactericidal effect of silver nanoparticles. Nanotechnology. 2005; 16(10): 2346–53. PubMed Abstract \| Publisher Full Text\n\nSotiriou GA, Pratsinis SE: Antibacterial activity of nanosilver ions and particles. Environ Sci Technol. 2010; 44(14): 5649–5654. PubMed Abstract \| Publisher Full Text\n\nSadaf D, Ahmad MZ: Factors associated with postoperative pain in endodontic therapy. Int J Biomed Sci. 2014; 10(4): 243–247. PubMed Abstract \| Free Full Text\n\nAli SG, Mulay S, Palekar A, et al.: Prevalence of and factors affecting post-obturation pain following single visit root canal treatment in Indian population: A prospective, randomized clinical trial. Contemp Clin Dent. 2012; 3(4): 459–463. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAli A, Olivieri JG, Duran-Sindreu F, et al.: Influence of preoperative pain intensity on postoperative pain after root canal treatment: A prospective clinical study. J Dent. 2016; 45: 39–42. PubMed Abstract \| Publisher Full Text\n\nNoronha VT, Paula AJ, Durán G, et al.: Silver nanoparticles in dentistry. Dent Mater. 2017; 33(10): 1110–1126. PubMed Abstract \| Publisher Full Text\n\nBhol KC, Schechter PJ: Topical nanocrystalline silver cream suppresses inflammatory cytokines and induces apoptosis of inflammatory cells in a murine model of allergic contact dermatitis. Br J Dermatol. 2005; 152(6): 1235–1242. PubMed Abstract \| Publisher Full Text\n\nAnjaneyulu K, Nivedhitha MS: Influence of calcium hydroxide on the post-treatment pain in Endodontics: A systematic review. J Conserv Dent. 2014; 17(3): 200–207. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOrstavik D, Hapassalo M: Disinfection by endodontic irrigants and dressings of experimentally infected dentinal tubules. Endod Dent Traumatol. 1990; 6(4): 142–149. PubMed Abstract \| Publisher Full Text\n\nYesiloy C, Whitaker E, Cleveland D, et al.: Antimicrobial and toxic effects of established and potential root canal irrigants. J Endod. 1995; 21(10): 513–515. PubMed Abstract \| Publisher Full Text" }	[ { "id": "42096", "date": "04 Jan 2019", "name": "Reham Hassan", "expertise": [ "Reviewer Expertise Endodontics" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe present article compares the postoperative pain after using either calcium hydroxide or silver nanoparticles as intra-canal medication.\n\nThe success of root canal therapy depends on the reduction or eradication of the microbial load inside root canals. However, total elimination of the bacterial population inside the root canal is difficult to accomplish.\n\nIt's well known that Ca(OH)2 does not show complete effectiveness in cases of persistent root canal infections, therefore different nano scale materials have been recently used as an antimicrobial agent, the post operative pain after using new materials is considerably an important factor\n\nThe authors definitely spent a lot of time carrying out the study and writing down their findings.\nThe Introduction section contains an adequate background regarding the use of silver nanoparticles for intra-canal medication.\n\nRegarding the methodology, several aspects should be pointed out:\n\nPatients inclusion and exclusion criteria were clearly stated Ethics committee approval and randomization procedure were provided The method of introduction of the silver nanoparticle to fill of canal space was not clearly mentioned\nResults section shows the main results clearly. Generally discussion section is good and deliberate. The provided CONSORT checklist and flowchart for this article is quit beneficial.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "52081", "date": "12 Aug 2019", "name": "Mothanna K. Al Rahabi", "expertise": [ "Reviewer Expertise Endodontic therapy", "Endodontic pain", "teaching endodontic." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis research has major defects:\nThe authors measure the pain during root canal therapy, not postoperative pain. Postoperative pain measured after root canal obturation, not after intracanal medication. So they need to modify the title to be as the following: Intra-canal medication containing silver nanoparticle versus calcium hydroxide in reducing pain during endodontic treatment .\n\nThere is a need to expand introduction to contain more information regarding endodontic pain and there is need to clarify the following statement: \"However, the ability of calcium hydroxide medication to completely eradicate bacterial species from the root canal has been questioned\". The authors compared silver nanoparticle and calcium hydroxide so there is no need to mention the following statement: \"chlorhexidine intra-canal medicament significantly reduced postoperative pain more than calcium hydroxide medication, where at 4 hours, 20 of 100 patients in chlorhexidine group had no pain compared to only 4 of 100 patients in calcium hydroxide group, and at 24 hours, 44 of 100 patients in chlorhexidine group had no pain compared to 24 in calcium hydroxide group\".\n\nThe authors wrote in sample size it represents the Egyptian population, do you think 34 patients represent the Egyptian population?\n\nIs silver nanoparticles intracanal medicament formula which is used in this study eligible to use with patients and approved by the medical authority in Egypt?\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1949
https://f1000research.com/articles/7-1332/v1	23 Aug 18	{ "type": "Research Article", "title": "Nucleocytoplasmic distribution of S6K1 depends on the density and motility of MCF-7 cells in vitro", "authors": [ "Viktoriia Kosach", "Kateryna Shkarina", "Anastasiia Kravchenko", "Yuliia Tereshchenko", "Evelina Kovalchuk", "Larysa Skoroda", "Mykhailo Krotevych", "Antonina Khoruzhenko", "Viktoriia Kosach", "Kateryna Shkarina", "Anastasiia Kravchenko", "Yuliia Tereshchenko", "Evelina Kovalchuk", "Larysa Skoroda", "Mykhailo Krotevych" ], "abstract": "Background: The ribosomal protein S6 kinase 1 (S6K1) is one of the main components of the mTOR/S6K signal transduction pathway, which controls cellular metabolism, autophagy, growth, and proliferation. Overexpression of S6K1 was detected in tumors of different origin including breast cancer, which was associated with a worse disease outcome. In addition, significant accumulation of S6K1 was found in the nuclei of breast carcinoma cells suggesting the implication of kinase nuclear substrates in tumor progression. However, this aspect of S6K1 functioning is poorly understood. The main aim of the present work was to study the subcellular localization of S6K1 in breast cancer cells with focus on cell migration. Methods: Multicellular spheroids of MCF-7 cells were generated using agarose-coated Petri dishes. Cell migration was initiated by spheroids seeding onto growth surface and subsequent cultivation for 24 and 72 hours. S6K1 subcellular localization was studied in human breast cancer and normal tissue, 2D and 3D MCF-7 cell culture using immunofluorescence analysis and confocal microscopy. Results: Analysis of histological sections of human breast cancer and normal tissue revealed predominantly nuclear localization of S6K1 in breast malignant cells and mainly cytoplasmic one in conditionally normal cells. In vitro studies of MCF-7 cells showed that the subcellular localization of S6K1 depends on the cell density in the monolayer culture. S6K1 relocalization from the cytoplasm into the nucleus was detected in MCF-7 cells migrating from multicellular spheroids onto growth surface. Immunofluorescence analysis of S6K1 and immunocoprecipitation assay revealed the colocalization and interaction between S6K1 and transcription factor TBR2 (T-box brain protein 2) in MCF-7 cells. Bioinformatical analysis revealed existence of several phosphorylation sites in TBR2 for S6K1 suggesting that TBR2 can be a target for phosphorylation and regulation by S6K1. Conclusions: Subcellular localization of S6K1 depends on the density and locomotor activity of the MCF-7 cells.", "keywords": [ "S6K1", "subcellular localization", "mTOR/S6K1 signaling pathway", "breast cancer", "TBR2", "Eomesodermin", "MCF-7 cell line" ], "content": "Introduction\n\nRibosomal protein S6 kinase 1 (S6K1) belongs to the AGC family of serine/threonine protein kinases (Ruvinsky & Meyuhas, 2006). It is involved in the regulation of crucial physiological processes, such as protein synthesis, ribosomal biogenesis, the G1/S-phase transition of the cell cycle, mRNA splicing, differentiation of specific cell types, and apoptosis. A large number of cellular targets makes S6K1 a key regulator of cell size, growth, and proliferation (Magnuson et al., 2012). S6K1 activity is under the control of the PI3K/Akt/mTOR signaling pathway, which is dysregulated in diverse human pathologies, including diabetes, obesity, neurodegenerative disorders, and cancer (Tavares et al., 2015). Overexpression of S6K1 was found in several tumor types, including breast cancer, and this was associated with a worse disease outcome (Bostner et al., 2015).\n\nIn mammalian cells, S6K1 is encoded by RPS6KB1 gene located at chromosome 17. Several isoforms of the S6K1 protein are known: 85kDa S6K1 and 70kDa S6K1 (p85S6K1 and p70S6K1 respectively), which originate from alternative translation initiation sites, and hypothetical p60S6K1, suggested to be a product of alternate mRNA translation as well (Kim et al., 2009). Recently, new 31kDa isoform of S6K1 (p31S6K1) encoded by mRNA splice variant was identified. It was demonstrated that translated p31S6K1 isoform does not have catalytic activity, but possesses oncogenic properties (Ben-Hur et al., 2013; Song & Richard, 2015). The longer isoform p85S6K1 has an additional 23 amino acid extension at the N-terminus of the molecule where the nuclear localization signal is located. Earlier p85S6K1 was described as a predominantly nuclear kinase. However, recent studies revealed it in the cytoplasm of the breast cancer cells and in the primary human fibroblasts using nuclear-cytoplasmic fractionation (Kim et al., 2009; Rosner & Hengstschläger, 2011). Another isoform of S6 kinase p70S6K1 was thought to localize predominantly in the cytoplasm. Treatment of cells with leptomycin B (the nuclear export inhibitor) led to the accumulation of the p70S6K1 in the nucleus. This finding allowed supposition that p70S6K1 shuttles between the cytoplasm and nucleus of the cell (Panasyuk et al., 2006). To date, very little is known about p31S6K1 subcellular localization. It is thought to be nuclear in human normal fibroblasts (Rosner & Hengstschläger, 2011). Overall, S6K1 subcellular localization data have been based predominantly on subcellular fractionation assay or immunocytochemical analysis of recombinantly expressed kinase. Information about nucleocytoplasmic distribution of the endogenous S6K1 is still limited, and mechanisms of its regulation remain elusive.\n\nRecent studies suggest that S6K1 subcellular localization and activation depends on physiological features of different tissues. Immunohistochemical analysis of breast tumors revealed prominent S6K accumulation in the nuclei of carcinoma cells (Filonenko, 2013; Filonenko et al., 2004; Lyzogubov et al., 2005). In other studies, it was shown that nuclear accumulation of S6K1 was indicative of a reduced tamoxifen effect in breast cancer patients, while cytoplasmic localization of S6K1 was associated with better prognosis (Bostner et al., 2015).\n\nMigration of the cancer cells is an important stage of cancer progression that usually lead to the tissue invasion and formation of distant metastases. The recent data suggest that S6K1 could be involved in the regulation of the motility of normal and malignant cells. Knockdown of p70S6K1 or inhibition of S6K1 kinase activity causes a significant decrease in the migration properties of the prostate, breast, and ovarian cancer cells in vitro (Amaral et al., 2016; Ip et al., 2011). Moreover, activation of p70S6K1 in human ovarian carcinoma cells that occurs in response to stimulation by hepatocyte growth factor (HGF) led to increased expression of matrix metalloproteinase 9 (MMP9) and higher migration rate of these cells (Zhou & Wong, 2006). It was shown that p70S6K1 stimulated activation of Cdc42, Rac1, and PAK1 – the known regulators of actin cytoskeleton reorganization (Aslan et al., 2011; Liu et al., 2010). Besides, S6K1 colocalizes with the actin arches at the leading edge of moving mesothelioma cells. Treatment with rapamycin (specific mTOR inhibitor) complicated the formation of actin arches even when cells were stimulated with endothelial growth factor (EGF) (Berven et al., 2004; Liu et al., 2008). However, the link between subcellular localization of S6K1 and its functions in migrating cancer cells is not fully understood.\n\nIn the present research, we studied the subcellular localization of endogenous S6K1 in breast tumor and normal tissue, and in breast adenocarcinoma MCF-7 cells in monolayer culture, 3D multicellular spheroids, and in course of cancer cell migration. We found that nucleocytoplasmic distribution of S6K1 depends greatly on the density of the monolayer culture, and is different in 3D vs 2D cell culture. Moreover, we revealed that S6K1 relocalizes to the nucleus during migration of cancer cells from multicellular spheroids onto growth surface. In addition, we analyzed the possible interaction of S6K1 with a number of transcription factors, involved in the regulation of cell motility. For the first time, we described the colocalization and co-immunoprecipitation of S6K1 and TBR2 (T-box brain protein 2) in breast cancer cells. These data could indicate that during cell migration S6K1 possibly interacts with transcription factors in the cell nucleus, activating genes that are involved in the regulation of cell locomotor activity.\n\n\nMethods\n\nHuman breast adenocarcinoma cell line MCF-7 was obtained from Bank of Cell Lines of the R. E. Kavetsky Institute of Experimental Pathology, Oncology and Radiobiology, NASU (Ukraine). The cells were cultivated in DMEM culture medium (Gibco, USA) supplemented with 10% fetal calf serum (FCS, HyClone, USA), 4 mM glutamine, 50 units/ml penicillin, 50 µg/ml streptomycin at 37°C under 5% CO2. The medium was changed every third day. For immunofluorescence analysis cells were seeded onto sterile glass coverslips.\n\nTo form multicellular spheroids, MCF-7 cells were detached and 1×106 cells were seeded into 100 mm Petri dishes coated with 1% agarose (Sigma-Aldrich, A9045) for 72 h.\n\nFor initiation of spheroid-to-monolayer reversion and cell migration, multicellular spheroids were transferred onto growth surface (glass coverslip) and cultivated for 24 or 72 h. Then outspreaded spheroids were applied to immunofluorescence analysis.\n\nCellular and spheroid morphology was evaluated microscopically using transmitted light (CETI Versus inverted microscope, CETI, Belgium, and Leica DM 1000, Leica Microsystems, Germany).\n\nCultured MCF-7 cells were fixed with 10% formalin for 15 min at room temperature (RT). Thereafter, the cells were permeabilized with 0.2% Triton X-100 in PBS for 10 min. Autofluorescence was decreased by incubation with 10 mM cupric sulphate and 50 mM ammonium acetate, pH 5.0 for 30 min at RT. Non-specific binding was blocked after incubation with 10% FCS in PBS for 30 min at 37°C in a humidifying chamber.\n\nS6K1 subcellular localization was revealed using anti-S6K1-C-terminus rabbit polyclonal antibodies (generated and evaluated earlier (Savinska et al., 2001; if you are interested in obtaining this antibody, please contact the corresponding author)) at 1:100, and anti-phospho-S6K1 (T389) rabbit polyclonal antibodies at 1:20 (Cell Signaling Technology Cat# 9205, RRID:AB_330944). The secondary Fluorescein (FITC)-AffiniPure Goat Anti-Rabbit IgG (H+L) antibody 1:400 (Jackson ImmunoResearch Labs Cat# 111-095-003, RRID:AB_2337972) were applied for 45 min at 37°C in a humidifying chamber.\n\nDouble immunofluorescence analysis was performed by addition of the primary antibody mix: anti-S6K1-C-terminal mouse monoclonal antibodies (generated earlier (Pogrebnoy et al., 1999); if you are interested in obtaining this antibody, please contact the corresponding author) at 1:100 + anti-TBR2 rabbit polyclonal antibodies at 1:100 (Abcam Cat# ab23345, RRID:AB_778267), or + anti-ERG rabbit monoclonal antibodies at 1:50 (Dako, Cat#M7314), or anti-CDX2 rabbit monoclonal antibodies at 1:100 (Abcam Cat# ab76541, RRID:AB_1523334), overnight at +4°C in a humidifying chamber. The secondary Fluorescein (FITC)-AffiniPure Donkey Anti Mouse IgG (H+L) antibody (Jackson ImmunoResearch Labs Cat# 715-095-150, RRID:AB_2340792) at 1:400, and Rhodamine (TRITC)-AffiniPure Donkey Anti-Rabbit IgG (H+L) antibody (Jackson ImmunoResearch Labs Cat# 711-025-152, RRID:AB_2340588) at 1:400 were applied for 45 min at 37°C in a humidifying chamber. Samples were embedded into Mowiol medium (Sigma-Aldrich, USA) containing 2.5% DABCO (Sigma-Aldrich), 0.5 % DAPI (Sigma-Aldrich).\n\nAll microscopy studies were performed using Leica DM 1000 fluorescent microscope and Zeiss LSM 510 META microscope (Carl Zeiss Microscopy GmbH, Germany). Fluorescence images were analyzed with free software Fiji/ImageJ v1.52b (Fiji, RRID:SCR_002285; Schindelin et al., 2012). Figures were generated with the FigureJ plugin (Mutterer & Zinck, 2013) in Fiji/ImageJ v1.52b.\n\nFor quantitative characterization of colocalization Pearson coefficient and Manders coefficients (M1 and M2) analysis was performed on background-subtracted images using JACoP plugin (Bolte & Cordelières, 2006) in Fiji/ImageJ v1.52b. Pearson coefficient (Rr) and Manders coefficients (M1 and M2) were expressed as mean value +/-SD, the experiment was duplicated, n=5 for each condition. To validate and describe the obtained degree of colocalization pre-defined image sets from Colocalization Benchmark Source were used. Obtained values of the colocalization coefficients were used to find the closest benchmark.\n\nHistological samples of human mixed ductal/lobular carcinoma of the breast and surrounding conditionally normal tissue were obtained from 10 patients within the framework of the cooperation agreement between the National Cancer Institute and the Institute of Molecular Biology and Genetics of the National Academy of Sciences of Ukraine. This study has been approved by the Committee on Biological & Medical Ethics of the National Cancer Institute of Ukraine (approval number - № 67, 25.03.2015). Written informed consent was obtained from all patients for the use of their tissues in research.\n\nSections of human breast cancer and surrounding tissues or multicellular spheroids were deparaffinized in xylene and rehydrated in a series of graded alcohols. For the antigen retrieval, slides were placed in citrate buffer (10 mM citric acid, pH 6.0) and transferred to the microwave. Sections were boiled two times for 5–7 min. Then, sections were treated with 0.2% Triton X-100 for 10min. Endogenous peroxidase was quenched with 3% H2O2 in PBS for 30 min. After blocking of non-specific staining with 10% FCS in PBS, sections were incubated with anti-S6K1-C-terminal rabbit polyclonal antibodies (1:100) overnight at +4°C, and thereafter they were incubated with peroxidase-conjugated secondary antibodies (1:100; Promega Cat# W4011, RRID:AB_430833) for 1 hour at 37°C. The reaction was developed with 3,3’diaminobenzidine (Sigma-Aldrich) solution.\n\nPredicted sites of TBR2 phosphorylation by S6K1 were revealed using Group-based Prediction System v2.1 (Xue et al., 2011; GPS, RRID:SCR_016374). The sequence of human TBR2 for this analysis was taken from the National Center for Biotechnology Information, NCBI Reference Sequence: NP_001265111.1.\n\nAnti-S6K1 mouse monoclonal antibodies (Pogrebnoy et al., 1999) were immobilized on protein A/G PLUS Agarose beads (Santa Cruz Biotechnology) overnight at +4 C.\n\nMCF-7 cells were washed with ice-cold phosphate-buffered saline and extracted with lysis buffer, containing 20 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1% Triton X-100, 5mM EDTA, 50 mM sodium fluoride, 5 mM β-glycerophosphate,10 mM sodium pyrophosphate, 1 mM sodium orthovanadate and a mixture of protease inhibitors (Roche Molecular Diagnostics, France). Cell lysates were centrifuged at 13 000 rpm for 20 min at 4°C. Endogenous S6K1 was precipitated by adding 1000µg of total cell lysates to the immobilized antibodies and incubating overnight at 4°C. Immune complexes were washed three times with lysis buffer, boiled for 5 min in Laemmli sample buffer, and used for immunoblot analysis. As a control, protein A/G PLUS Agarose beads were incubated with monoclonal antibodies or cell lysates alone.\n\nSamples were separated by 10% SDS-PAGE and transferred onto PVDF membrane (Millipore, Billerica, MA). The membrane was blocked with 5% skim milk in PBST (140 mM NaCl, 2.6 mM KCl, 10 mM Na2HPO4, 1.8 mM KH2PO4, 0,05% Tween-20, pH 7.4) for 1 h at RT, and then incubated with anti-TBR2/Eomes rabbit antibodies at 1:500 (Abcam Cat# ab23345, RRID:AB_778267) or anti-S6K1 C-terminal rabbit polyclonal antibodies 1:3000 overnight at 4°C. After washing three times with PBST, HRP-conjugated goat anti-rabbit IgG (Jackson ImmunoResearch Labs Cat# 111-035-144, RRID:AB_2307391) in 1:10 000 dilution were incubated with the membrane for 1 h at RT. Finally, the membrane was developed using an enhanced chemiluminescence system (Fluco) and then exposed to Agfa X-ray film.\n\n\nResults and discussion\n\nThere are several mediated and direct approaches to determine protein function in cells. The first includes cell models without direct interaction between the studied protein and the reporter marker. The function of a studied protein is evaluated as a correlation with a reporter marker, which in turn gives a characteristic to a particular cellular process (Nizheradze, 2006). The second group of approaches is based on closer interactions between the agent and the reporter marker. One of them consists of the determination of studied protein belonging to the specific subcellular compartment and next revealing its protein-partners with known functions. The difference in subcellular localization of the protein in different cell types may suggest the realization of different function.\n\nFor the first step of the present study, subcellular distribution of S6K1 was analyzed in normal and cancer cells of a histological section of human breast cancer and normal tissues. Our results support previous data about preferential nuclear localization of S6K1 in breast malignant cells (Bostner et al., 2015; Filonenko et al., 2004) and mainly cytoplasmic one in conditionally normal surrounding tissues (Figure 1A, B).\n\n(A) Immunohistochemical detection of S6K1 in human conditionally normal breast tissue. Magnification 400x. (B) Immunohistochemical analysis of S6K1 subcellular localization in human breast cancer tissue. Magnification 400x. Arrows point out to the positive reaction in the nuclei of the cells. (C) Immunohistochemical detection of S6K1 in MCF-7 multicellular spheroids. Magnification 200x. (D) Immunofluorescence image of S6K1 subcellular distribution (green) in 40% confluent MCF-7 cell monolayer. Arrows point out to the positive reaction in the nuclei of the cells. DNA was counterstained with DAPI (blue). Scale bars correspond to 20 µm. The data are representative of three independent experiments.\n\nAs was shown earlier, 3D cell culture systems are preferable for tumor growth study (Bingel et al., 2017). That’s why we analyzed the intracellular distribution of S6K1 in multicellular spheroids of MCF-7 cells and in monolayer culture. We detected a bright signal of S6K1 in the cytoplasm and its absence in the nuclei of cells of the multicellular spheroids (Figure 1C). In contrast to a spheroid culture, S6K1 distribution in MCF-7 cells monolayer culture at 40–60% confluence revealed bright nuclear staining in interphase cells with the moderate cytoplasmic signal (Figure 1D).\n\nA significant difference in S6K1 localization in monolayer and spheroid cultures can be explained by differences in cell growth conditions in two distinct cultures. First of all, the reason for such alterations could be explained by a cascade of intracellular events induced by cell-matrix adhesion or intercellular interactions. One can assume that the S6K1 is involved in such intracellular rearrangement. To clarify this issue, S6K1 subcellular localization in MCF-7 cells cultured at different densities was analyzed by the present study. The immunofluorescence analysis revealed the displacement of S6K1 from the nucleus to the cytoplasm when the cell culture density increased (Figure 2A–E). At the lowest MCF-7 cell density, S6K1 was observed predominantly in the nuclei of cultured cells whereas at the highest density the main signal was concentrated in the cytoplasm.\n\nMCF-7 cells were seeded onto glass coverslips in the density 10 000 cells/well (A), 30 000 cells/well (B), 50 000 cells/well (C), 70 000 cells/well (D), 100 000 cells/well (E), and cultivated for 48 hours. Then cells were fixed and stained with anti-S6K1 (green). White arrows point to the prominent positive reaction in the nuclei of the cells, yellow arrows point to the decreased staining in the nuclei. Scale bars are 20 µm. The images are representative of three independent experiments. (F) MCF-7 cells were cultivated to form a super-confluent monolayer. Then fragments of the monolayer were gently detached by short incubation with trypsin, transferred on the coverslip and left for 48 hours to grow. After it the fragments were fixed and stained with anti-S6K1 (green). White arrows point to the positive reaction in the nuclei of the cells at the leading edge of the monolayer fragment; yellow arrows point to the decreased staining in the nuclei of the cells at the center of the monolayer fragment. Magnification 400x.\n\nBesides, to test the possible dependence of S6K1 subcellular localization on cell density the following approach was introduced. After reaching 90% confluence layer of MCF-7 cells, the monolayer was detached from growth surface by short treatment with trypsin (w/o EDTA), and replaced in fresh culture medium. After cultivation for 48 h such fragments of monolayer demonstrated high cell density in the center and spread out cells at the edges. Immunofluorescence analysis revealed that cells’ spreading was accompanied by the alterations in S6K1 localization (Figure 2F). Densely packed cells in the center of the fragments demonstrated cytoplasmic reaction for S6K1, while spread out cells had preferentially nuclear staining.\n\nSince cell spreading can be considered as a stage of migration, the previous data led to the hypothesis of a relation between the initiation of cell migration and the relocalization of S6K1. Among the variety of cell migration models, it is necessary to point out those that are based on 3D cell cultures (Metzger et al., 2017). There are a lot of data concerning the similarity of multicellular spheroid organization and structure of solid malignant tissue (Rodrigues et al., 2018). Besides, the transformation of 3D multicellular spheroids in 2D cell colonies can be realized only by activation of cultured cell migration. So, in the present study, S6K1 distribution was examined on migrating MCF-7 cells after replacing MCF-7 multicellular spheroids onto the growth surface (Figure 3). We applied the immunofluorescence analysis of cultured cells after 24 and 72 hours after initiation of cell migration from MCF-7 spheroids. Obtained data suggest that significant relocalization of S6K1 from the cytoplasm into the nuclei in course of locomotor function realization took place (Figure 4A, B). The cells, which were still in 3D condition, had positive cytoplasm and negative nuclei, as well as cells of spheroid at histological sections regardless of their remoteness from the edge of the spheroid (Figure 1C). During MCF-7 spheroid transformation in monolayer, spreading cells demonstrated strong accumulation of S6K1 in the nuclei (Figure 4B).\n\nTo generate multicellular spheroids, MCF-7 cells were seeded in the Petri dishes coated with 1% agarose, and cultivated for 72 hours. To analyze cell migration, obtained spheroids were transferred onto glass coverslips and cultured for 24 or 72 hours. Microphotographs of the spheroids general view were taken by Leica DM1000 (Leica, Germany) in transmitted light. Magnification 200x.\n\n(A) Immunofluorescence analysis of S6K1 subcellular localization (green) in the MCF-7 spheroid reversed for 24 hours. Arrows point to the predominantly nuclear distribution of the S6K1 in the migrating cells. DNA was counterstained with DAPI (blue). Scale bars are 20 µm. (B) Immunofluorescence analysis of S6K1 subcellular localization (green) in the MCF-7 cells at the leading age of spheroid reversed for 72 hours. Arrows point to the predominantly nuclear distribution of the S6K1 in the migrating cells. DNA was counterstained with DAPI (blue). Scale bars are 20 µm. The images are representative of three independent experiments.\n\nThe most often used marker of S6K1 activation is Thr389 phosphorylation mediated by mTOR (Romanelli et al., 2002). Thus, we analyzed phosphorylation status of S6K1 in MCF-7 cells during spheroid transformation into monolayer by immunofluorescence analysis. Overall, the pattern of phospho-S6K1 distribution was similar to that observed for total S6K1 (Figure 5). Namely, in the central part of the spheroid, S6K1 was mainly observed in the cytoplasm (however some of the nuclei were positive), whereas all cells of the leading edge demonstrated predominant nuclear kinase localization (Figure 5A). Also, strong nuclear localization of phospho-S6K1 (Thr389) was revealed in monolayer culture of MCF-7 cells (Figure 5B).\n\n(A) Confocal image of MCF-7 spheroid-to-monolayer reversion model. Cells were stained with anti-phospho-S6K1 (T389) (green). DNA was counterstained with DAPI (blue). Scale bars correspond to 20 µm. Arrows point to the positive reaction in the nuclei of the migrating cells at the leading edge of the spheroid. (B) Confocal image of monolayer culture of the MCF-7 cells stained with anti-phospho-S6K1 (T389) (green). DNA was counterstained with DAPI (blue). Scale bars correspond to 20 µm. The images are representative of two independent experiments.\n\nObtained data suggested that activation of cell locomotor function is accompanied by cytoplasm/nuclear shuttling of S6K1; however, the biological sense of the event is not yet clear. The possible explanation could be the implication of S6K1 in the regulation of transcription factors affecting expression of genes that control cell migration.\n\nAs it was mentioned above, several transcription factors are known as targets of S6K, but their nuclear localization in relation to cell migration was not reported. That’s why, we analyzed subcellular distribution of several transcription factors, which are mTOR/S6K signaling regulated and activated in migrating cells either in a cancer tissue or in the process of organism development. Among them the mammalian transcription factor CDX2, which plays a key role in intestinal development and differentiation. It was determined that reduced expression of CDX2 is important in colon tumorigenesis through mTOR-mediated chromosomal instability (Aoki et al., 2003). Fusion of another transcription factor ERG and androgen-responsive TMPRSS2 serine protease is an important feature of prostate cancer. A strong correlation has been revealed between TMPRSS2-ERG fusion and activation of mTOR/S6K pathway (Faraj et al., 2013; King et al., 2009). The third studied transcription factor was T-box transcription activator Eomesodermin (or TBR2) (Conlon et al., 2001). Earlier it was regarded as a target for anticancer therapy. It was detected that siRNA knockdown of Eomesodermin in human hepatocellular carcinoma significantly affected anchorage-independent cell growth (Gao et al., 2014). Besides, it is involved in lymphocyte differentiation. It should be noted that TOR signaling is involved in modulation of TBR2 activity; it was demonstrated that TOR signaling was the central regulator of transcriptional programs by regulation of expression of transcription factors T-bet and Eomesodermin, that determined effector or memory cell fates in CD8+ T cells (Cui et al., 2016).\n\nImmunofluorescence analysis of subcellular distribution of S6K1 and mentioned transcription factors in MCF-7 cells revealed that ERG transcription factor either was present in scant quantities or not determined at all in MCF-7 cells (Dataset 4; (Kosach et al., 2018d)). CDX2 was determined as positive dots predominantly in the nuclei, CDX-2 and S6K1 colocalization was not detectable by confocal microscopy (Dataset 4; (Kosach et al., 2018d)). TBR2/Eomesodermin positive granules were observed in the cytoplasm as well as in nuclei (Figure 6A, B). In both cases, partial but bright colocalization of TBR2 and S6K1 was revealed. Moreover, in low-density monolayer, when S6K1 localized mainly in the nuclei, TBR2 was observed predominantly in the nuclei as well. At high-density monolayer, S6K1 was redistributed in the cytoplasm, and TBR2 repeated the pattern of immunofluorescent reaction (Figure 6A, B).\n\n(A) Immunofluorescence image of low density monolayer culture of the MCF-7 cells double stained with anti-S6K1 (green) and anti-TBR2 (magenta). DNA was counterstained with DAPI (blue). Scale bars correspond to 20 µm. Arrows point to the regions of S6K1 and TBR2 colocalization. (B) Immunofluorescence image of high density monolayer culture of the MCF-7 cells double stained with anti-S6K1 (green) and anti-TBR2 (magenta). DNA was counterstained with DAPI (blue). Scale bars correspond to 20 µm. Arrows point out to the regions of S6K1 and TBR2 colocalization. The images are representative of two independent experiments.\n\nFor quantitative characterization of S6K1 and TBR2 colocalization, Pearson coefficient (Rr) and Manders coefficient (M1 and M2) analysis was performed on background-subtracted images using JACoP ImageJ plugin (Bolte & Cordelières, 2006). M1 shown the colocalization of S6K1 with TBR2, whereas M2 expressed the pool of TBR2 colocalizing with S6K1. Colocalization analysis of S6K1 and TBR2 in low density monolayer revealed Pearson coefficient Rr= 0.55 +/- 0.113, M1= 0.999 +/-0.01, M2= 0.84 +/-0.087. To validate and describe the obtained degree of colocalization pre-defined image sets from Colocalization Benchmark Source were used. The closest benchmark was CBS007RGM that corresponded to 60% colocalization, thus indicating the medium level of colocalization between TBR2 and S6K1. A slightly lower but reliable colocalization of S6K1 and TBR2 was observed in a monolayer with a high density. Namely Pearson coefficient was Rr=0.47 +/- 0.064, Manders coefficients were M1=0.995 +/-0.004 and M2=0.62+/-0.187. So, a slightly higher level of S6K1 and TBR2 colocalization was revealed in MCF-7 cells grown in low density monolayer, when S6K1 and TBR2 localized mainly in the nuclei.\n\nThe application of immunoprecipitation confirmed the interaction of S6K1 and TBR2 (Figure 7). Protein complexes containing S6K1 were extracted from cultured MCF-7 cell lysate using anti-S6K1 antibodies and then blotted with antibodies to TBR2. Obtained results revealed protein complex formation of S6K1 and TBR2 suggesting the possibility of TBR2 regulation via S6K1 mediated phosphorylation.\n\nEndogenous S6K1 was precipitated with anti-S6K1 mouse monoclonal antibodies immobilized on protein A/G PLUS Agarose beads (Santa Cruz Biotechnology). As a control, protein agarose beads were incubated with monoclonal antibodies or cell lysates alone. Immune complexes were analyzed by immunoblotting with anti-TBR2 rabbit antibodies (Abcam, ab23345) or anti-S6K1 C-terminal rabbit polyclonal antibodies. The data are representative of two independent experiments.\n\nFurther computational prediction of phosphorylation sites in TBR2 (GPS 2.1) indeed revealed several sites, and three of them (Tyr421, Tyr423, Ser646) can be phosphorylated by S6K1 with a high score (Figure 8). Interestingly, both Tyr421 and Tyr423 are located in the DNA binding domain and one can assume that their phosphorylation could be related to the binding ability of the transcription factor to the targeted DNA. Another phosphorylation site (Ser646) is located within transcription activation domain at C-terminus of TBR2, which is involved in transcription activation. So, S6K1 can be involved in the regulation of TBR2 transcription activity. However, further research is needed to find if S6K1 phosphorylates TBR2 in vitro and in vivo.\n\nFree software Group-based Prediction System v2.1 was used for bioinformatics analysis. It revealed that TBR2 contained three sites that could be phosphorylated by S6K1 with a high probability (A). Two of them, T421 and T423, situates in the DNA binding domain of the TBR2. Third site S646 is within transcription activation domain at C-terminus of TBR2 (B). Possible phosphorylation of these residues by S6K1 could significantly influence the TBR2 activity. However, this finding warrants further research.\n\nConcerning TBR2 targets, in the course of organism development, Eomesodermin can induce virtually the entire spectrum of mesodermal genes in all types of mesodermal cells, which could appear in malignant cells of non-mesodermal origin (Reim et al., 2017; Russ et al., 2000).\n\nConsidering the multiplicity of S6K1 substrates, phosphorylation of the TBR2 transcription factor is not the only reason for the movement of the kinase from the cytoplasm into the nucleus of migrating cells. However, the proposed interaction can partially explain the accumulation of kinase in the nucleus of moving cells. The existence of several kinase isoforms may explain the partial colocalization of the kinase and the transcription factor (Amaral et al., 2016). In addition to the previously known classical nuclear substrates of S6K1, in case of breast cancer, it is necessary to note that this kinase can activate estrogen receptor-α, which is a nuclear transcription factor by its phosphorylation at Ser167 in a ligand-independent manner (Yamnik & Holz, 2010). Besides, recent data indicate that S6K1 is targeted by histone acetyltransferases p300 and p300/CBP-associated factor (PCAF). The significance of this acetylation is not fully clear, but by analogy with S6K2, it is assumed that S6K1 is involved in the regulation of the transcription process (Fenton et al., 2010). Summing up, there are a number of data confirming the nuclear localization of S6K1, but the role that S6K1 performs in the nucleus of migrating malignant cells require further investigation.\n\n\nConclusions\n\nFor the first time, this study revealed S6K1 relocalization from the cytoplasm to the nuclei in migrating cells using the model of spheroids MCF-7 cells transformation into monolayer culture. Such relocalization could be linked to the S6K1 driven activation of transcription factors responsible for cell locomotion. Particularly, colocalization and interaction of S6K1 and transcription factor TBR2 were revealed using confocal microscopy and co-immunoprecipitation. In addition, bioinformatics analysis of phosphorylation sites in TBR2 supports a prediction about S6K1 mediated phosphorylation and regulation of TBR2.\n\n\nData availability\n\nF1000Research: Dataset 1. Unedited images that were used in Figure 1 and Figure 2, showing S6K1 subcellular localization in breast normal tissue, cancer tissue, and in MCF-7 cells monolayer., 10.5256/f1000research.15447.d214430 (Kosach et al., 2018a).\n\nF1000Research: Dataset 2. Unedited images from Figure 3., 10.5256/f1000research.15447.d214431 (Kosach et al., 2018b).\n\nF1000Research: Dataset 3. Unedited images that were used in Figure 4 and Figure 5, showing S6K1 and phospho-S6K1 (T389) subcellular localization during MCF-7 cell migration., 10.5256/f1000research.15447.d214432 (Kosach et al., 2018c).\n\nF1000Research: Dataset 4. Unedited images of S6K1 colocalization with transcription factors TBR2 (Figure 6), ERG, and CDX2., 10.5256/f1000research.15447.d214433 (Kosach et al., 2018d).\n\nF1000Research: Dataset 5. Unedited western blot images of co-immunoprecipitation of S6K1 and TBR2 used in Figure 7., 10.5256/f1000research.15447.d214434 (Kosach et al., 2018e).", "appendix": "Grant information\n\nThis work was supported by National Academy of Sciences of Ukraine grants (0115U003745) and (0115U001403 to A.Khoruzhenko).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nWe would like to thank Prof. Valeriy Filonenko for support, expertise, and discussions that greatly assisted the research. We also thank Dr. Volodymyr Shablii for antibodies against transcription factors (TBR2, CDX2, ERG), and Dr. Serhiy Karakhim for help in obtaining the confocal images at Light Microscopy Facility of the O.V. Palladin Institute of Biochemistry of NAS of Ukraine.\n\n\nReferences\n\nAmaral CL, Freitas LB, Tamura RE, et al.: S6Ks isoforms contribute to viability, migration, docetaxel resistance and tumor formation of prostate cancer cells. BMC Cancer. 2016; 16: 602. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAoki K, Tamai Y, Horiike S, et al.: Colonic polyposis caused by mTOR-mediated chromosomal instability in Apc+/Delta716 Cdx2+/- compound mutant mice. Nat Genet. 2003; 35(4): 323–330. PubMed Abstract \| Publisher Full Text\n\nAslan JE, Tormoen GW, Loren CP, et al.: S6K1 and mTOR regulate Rac1-driven platelet activation and aggregation. Blood. 2011; 118(11): 3129–3136. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBen-Hur V, Denichenko P, Siegfried Z, et al.: S6K1 alternative splicing modulates its oncogenic activity and regulates mTORC1. Cell Rep. 2013; 3(1): 103–115. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBerven LA, Willard FS, Crouch MF: Role of the p70S6K pathway in regulating the actin cytoskeleton and cell migration. Exp Cell Res. 2004; 296(2): 183–195. PubMed Abstract \| Publisher Full Text\n\nBingel C, Koeneke E, Ridinger J, et al.: Three-dimensional tumor cell growth stimulates autophagic flux and recapitulates chemotherapy resistance. Cell Death Dis. 2017; 8(8): e3013. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBolte S, Cordelières FP: A guided tour into subcellular colocalization analysis in light microscopy. J Microsc. 2006; 224(Pt 3): 213–232. PubMed Abstract \| Publisher Full Text\n\nBostner J, Karlsson E, Eding CB, et al.: S6 kinase signaling: tamoxifen response and prognostic indication in two breast cancer cohorts. Endocr Relat Cancer. 2015; 22(3): 331–343. PubMed Abstract \| Publisher Full Text\n\nConlon FL, Fairclough L, Price BM, et al.: Determinants of T box protein specificity. Development. 2001; 128(19): 3749–3758. PubMed Abstract\n\nCui N, Su LX, Wang H, et al.: mTOR Modulates Lymphocyte Differentiation through T-bet and Eomesodermin in Response to Invasive Pulmonary Aspergillosis in Rats. Chin Med J (Engl). 2016; 129(14): 1704–1710. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFaraj SF, Albadine R, Chaux A, et al.: Activation of mammalian target of rapamycin signaling pathway markers in minute adenocarcinoma of the prostate. Urology. 2013; 82(5): 1083–1089. PubMed Abstract \| Publisher Full Text\n\nFenton TR, Gwalter J, Ericsson J, et al.: Histone acetyltransferases interact with and acetylate p70 ribosomal S6 kinases in vitro and in vivo. Int J Biochem Cell Biol. 2010; 42(2): 359–366. PubMed Abstract \| Publisher Full Text\n\nFilonenko VV: PI3K/mTOR/S6K signaling pathway - new players and new functional links. Biopolym Cell. 2013; 29(3): 207–214. Publisher Full Text\n\nFilonenko VV, Tytarenko R, Azatjan SK, et al.: Immunohistochemical analysis of S6K1 and S6K2 localization in human breast tumors. Exp Oncol. 2004; 26(4): 294–299. PubMed Abstract\n\nGao F, Xia Y, Wang J, et al.: Integrated analyses of DNA methylation and hydroxymethylation reveal tumor suppressive roles of ECM1, ATF5, and EOMES in human hepatocellular carcinoma. Genome Biol. 2014; 15(12): 533. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIp CK, Cheung AN, Ngan HY, et al.: p70 S6 kinase in the control of actin cytoskeleton dynamics and directed migration of ovarian cancer cells. Oncogene. 2011; 30(21): 2420–2432. PubMed Abstract \| Publisher Full Text\n\nKim D, Akcakanat A, Singh G, et al.: Regulation and localization of ribosomal protein S6 kinase 1 isoforms. Growth Factors. 2009; 27(1): 12–21. PubMed Abstract \| Publisher Full Text\n\nKing JC, Xu J, Wongvipat J, et al.: Cooperativity of TMPRSS2-ERG with PI3-kinase pathway activation in prostate oncogenesis. Nat Genet. 2009; 41(5): 524–526. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKosach V, Shkarina K, Kravchenko A, et al.: Dataset 1 in: Nucleocytoplasmic distribution of S6K1 depends on the density and motility of MCF-7 cells in vitro. F1000Research. 2018a. http://www.doi.org/10.5256/f1000research.15447.d214430\n\nKosach V, Shkarina K, Kravchenko A, et al.: Dataset 2 in: Nucleocytoplasmic distribution of S6K1 depends on the density and motility of MCF-7 cells in vitro. F1000Research. 2018b. http://www.doi.org/10.5256/f1000research.15447.d214431\n\nKosach V, Shkarina K, Kravchenko A, et al.: Dataset 3 in: Nucleocytoplasmic distribution of S6K1 depends on the density and motility of MCF-7 cells in vitro. F1000Research. 2018c. http://www.doi.org/10.5256/f1000research.15447.d214432\n\nKosach V, Shkarina K, Kravchenko A, et al.: Dataset 4 in: Nucleocytoplasmic distribution of S6K1 depends on the density and motility of MCF-7 cells in vitro. F1000Research. 2018d. http://www.doi.org/10.5256/f1000research.15447.d214433\n\nKosach V, Shkarina K, Kravchenko A, et al.: Dataset 5 in: Nucleocytoplasmic distribution of S6K1 depends on the density and motility of MCF-7 cells in vitro. F1000Research. 2018e. http://www.doi.org/10.5256/f1000research.15447.d214434\n\nLiu L, Chen L, Chung J, et al.: Rapamycin inhibits F-actin reorganization and phosphorylation of focal adhesion proteins. Oncogene. 2008; 27(37): 4998–5010. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiu L, Luo Y, Chen L, et al.: Rapamycin inhibits cytoskeleton reorganization and cell motility by suppressing RhoA expression and activity. J Biol Chem. 2010; 285(49): 38362–38373. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLyzogubov V, Khozhaenko Y, Usenko V, et al.: Immunohistochemical analysis of Ki-67, PCNA and S6K1/2 expression in human breast cancer. Exp Oncol. 2005; 27(2): 141–144. PubMed Abstract\n\nMagnuson B, Ekim B, Fingar DC: Regulation and function of ribosomal protein S6 kinase (S6K) within mTOR signalling networks. Biochem J. 2012; 441(1): 1–21. PubMed Abstract \| Publisher Full Text\n\nMetzger W, Rother S, Pohlemann T, et al.: Evaluation of cell-surface interaction using a 3D spheroid cell culture model on artificial extracellular matrices. Mater Sci Eng C Mater Biol Appl. 2017; 73: 310–318. PubMed Abstract \| Publisher Full Text\n\nMutterer J, Zinck E: Quick-and-clean article figures with FigureJ. J Microsc. 2013; 252(1): 89–91. PubMed Abstract \| Publisher Full Text\n\nNizheradze K: Concanavalin A, but not glycated albumin, increases subendothelial deposition of von Willebrand factor in vitro. Endothelium. 2006; 13(4): 245–248. PubMed Abstract \| Publisher Full Text\n\nPanasyuk G, Nemazanyy I, Zhyvoloup A, et al.: Nuclear export of S6K1 II is regulated by protein kinase CK2 phosphorylation at Ser-17. J Biol Chem. 2006; 281(42): 31188–31201. PubMed Abstract \| Publisher Full Text\n\nPogrebnoy PV, Kukharenko AP, Tykhonkova IA, et al.: Generation and characterization of monoclonal antibodies to p70S6 kinase alpha. Exp Oncol. 1999; 21(3–4): 232–238. Reference Source\n\nReim I, Frasch M, Schaub C: T-Box Genes in Drosophila Mesoderm Development. Curr Top Dev Biol. 2017; 122: 161–193. PubMed Abstract \| Publisher Full Text\n\nRodrigues T, Kundu B, Silva-Correia J, et al.: Emerging tumor spheroids technologies for 3D in vitro cancer modeling. Pharmacol Ther. 2018; 184: 201–211. PubMed Abstract \| Publisher Full Text\n\nRomanelli A, Dreisbach VC, Blenis J: Characterization of phosphatidylinositol 3-kinase-dependent phosphorylation of the hydrophobic motif site Thr389 in p70 S6 kinase 1. J Biol Chem. 2002; 277(43): 40281–40289. PubMed Abstract \| Publisher Full Text\n\nRosner M, Hengstschläger M: Nucleocytoplasmic localization of p70 S6K1, but not of its isoforms p85 and p31, is regulated by TSC2/mTOR. Oncogene. 2011; 30(44): 4509–4522. PubMed Abstract \| Publisher Full Text\n\nRuss AP, Wattler S, Colledge WH, et al.: Eomesodermin is required for mouse trophoblast development and mesoderm formation. Nature. 2000; 404(6773): 95–99. PubMed Abstract \| Publisher Full Text\n\nRuvinsky I, Meyuhas O: Ribosomal protein S6 phosphorylation: from protein synthesis to cell size. Trends Biochem Sci. 2006; 31(6): 342–348. PubMed Abstract \| Publisher Full Text\n\nSavinska LO, Kijamova RG, Pogrebnoy PV, et al.: Comparative characterization of S6 kinase α and β isoforms expression in mammalian tissues. Biopolym Cell. 2001; 17(5): 374–379. Publisher Full Text\n\nSchindelin J, Arganda-Carreras I, Frise E, et al.: Fiji: an open-source platform for biological-image analysis. Nat Methods. 2012; 9(7): 676–682. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSong J, Richard S: Sam68 Regulates S6K1 Alternative Splicing during Adipogenesis. Mol Cell Biol. 2015; 35(11): 1926–1939. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTavares MR, Pavan IC, Amaral CL, et al.: The S6K protein family in health and disease. Life Sci. 2015; 131: 1–10. PubMed Abstract \| Publisher Full Text\n\nXue Y, Liu Z, Cao J, et al.: GPS 2.1: enhanced prediction of kinase-specific phosphorylation sites with an algorithm of motif length selection. Protein Eng Des Sel. 2011; 24(3): 255–260. PubMed Abstract \| Publisher Full Text\n\nYamnik RL, Holz MK: mTOR/S6K1 and MAPK/RSK signaling pathways coordinately regulate estrogen receptor alpha serine 167 phosphorylation. FEBS Lett. 2010; 584(1): 124–128. PubMed Abstract \| Publisher Full Text\n\nZhou HY, Wong AS: Activation of p70S6K induces expression of matrix metalloproteinase 9 associated with hepatocyte growth factor-mediated invasion in human ovarian cancer cells. Endocrinology. 2006; 147(5): 2557–2566. PubMed Abstract \| Publisher Full Text" }	[ { "id": "37832", "date": "28 Sep 2018", "name": "Olivier E. Pardo", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe study by Kosach et al. clearly shows for the first time differential nucleo-cytoplasmic localisation of S6K1 in response to cell density and migration. It also provides some information about the type of role that S6K1 may play in the nucleus and the biological processes that may be associated.\n\nThe results shown are convincing and the quality of the data is sufficient to support the conclusions of the authors. Therefore, it is the opinion of this reviewer that the present manuscript should be accepted. However, the authors may want to address the comments below so as to improve the quality/impact of the manuscript:\n\nAt the end of the fourth paragraph of the Introduction, the authors refer to EGF as “endothelial growth factor”. This should be corrected to “epithelial growth factor”. The value of the first paragraph of the Results section (There are several... of different function.”) is not clear and should be deleted. On Page 13 of the manuscript, the authors refer to 2 residues on TRB2, Tyr421 and Tyr423, as potential sites phosphorylated by S6K1. S6K2 is a Ser/Thr kinase and therefore unable to phosphorylate Tyr residues. Are the authors referring to Thr sites instead? Please clarify. The authors argue that interaction of S6K1 with TRB2 may be involved in the migration of cells out of microspheres onto 2D layers. This is an interesting proposition and the manuscript would benefit from an experiment being performed using siRNAs to TRB2 and showing that this impacts either S6K1 distribution or the movement of cells out of the microspheres. The style of the paper is poor in places and could be improved prior to indexing.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4272", "date": "17 Dec 2018", "name": "Viktoriia Kosach", "role": "Author Response", "response": "We thank the reviewer for expert opinion and insightful comments on our study. We appreciate these comments, and we have tried to answer step by step. Please, find below our response to the reviewers’comments. At the end of the fourth paragraph of the Introduction, the authors refer to EGF as “endothelial growth factor”. This should be corrected to “epithelial growth factor”. We agree. It was erratum concerning EGF, and we have fixed it. The value of the first paragraph of the Results section (There are several... of different function.”) is not clear and should be deleted. We have deleted mentioned paragraph. On Page 13 of the manuscript, the authors refer to 2 residues on TRB2, Tyr421 and Tyr423, as potential sites phosphorylated by S6K1. S6K1 is a Ser/Thr kinase and therefore unable to phosphorylate Tyr residues. Are the authors referring to Thr sites instead? Please clarify. Yes, of course, S6K1 is Ser/Thr protein kinase, and we wrote about Threonine 421 and Threonine 423 residues on TRB2, but, unfortunately, there was erratum in the text. We have fixed it in version 2. The authors argue that interaction of S6K1 with TRB2 may be involved in the migration of cells out of microspheres onto 2D layers. This is an interesting proposition and the manuscript would benefit from an experiment being performed using siRNAs to TRB2 and showing that this impacts either S6K1 distribution or the movement of cells out of the microspheres. This is a very interesting suggestion. Also, to our mind, interaction of S6K1 and TBR2 could be just one from the possible reasons of S6K1 subcellular redistribution in course of MCF-7 cell migration. The level of their colocalization is statistically significant, but not very high, so, under TBR2 down-regulation we can obtain partial alteration of S6K1 relocalization that will be difficult to assay. Therefore, in the present work, we so far only point to the revealed colocalization. And we hope to develop this area of research in future. The style of the paper is poor in places and could be improved prior to indexing. The manuscript was proofread by a person with advanced level of scientific English. We believe that this greatly improved the style of the text." } ] }, { "id": "38191", "date": "05 Oct 2018", "name": "Yegor Vassetzky", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe MS by Kosach et al. describes an interesting phenomenon, nucleocytoplasmic shuttling of S6K1 in MCF-7 cells. The authors established a relationship between MCF-7 monolayer density and subcellular localization of the kinase. The work is worthy of publication, but the MS should be significantly shortened as is contains some redundant information and speculations. Proofreading by a native English-speaker is also recommended.\nMajor point: The putative phosphorylation of TBR2 by S6K1 is too speculative and not convincing to use it as one of the major conclusions of the papers. It should be either confirmed experimentally or removed from the abstract and the main results of the paper.\n\nMinor points: Material and Methods: p4 Please remove “if you are interested in obtaining this antibody, please contact the corresponding author“; p4 The anti-ERG and anti-CDX antibodies are not used in the main Figures. Please move information about these antibodies form the Materials and Methods section to legends of the supplementary figures;\nResults and discussion: P5 Please remove the first paragraph or move it to the introduction section. P5 “For the first step of the present study” is redundant Figure 1,5 legend. Please replace “positive reaction” with “staining” P6 The following phrase is unclear: “Since cell spreading can be considered as a stage of migration, the previous data led to the hypothesis…” Figure 6. It is unclear whether the images represent confocal sections or reconstituted 3D images. Colocalization of S6K1 and TBR2 should be analyzed quantitatively, eg using ImageJ tools for colocalization analysis. It is also unclear whether TBR2 shuttles between the nucleus and the cytoplasm similarly to S6K1. P9. S6K1 is a serine threonine kinase and the putative phosphorylation of TBR2 by S6K1 should concern threonines rather than tyrosines. I also propose to remove the bioinformatic part and the Figure 8 as they are too speculative. Indeed, many other kinases may potentially phosphorylate TBR2 basing on the bioinformatic analysis, and the authors did not even show whether TBR2 is indeed phosphorylated. Figure 7. TBR2 has a molecular weight of 84 kDa of TBR2, but the major band has a lower molecular weight. This should be explained in the Results section.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4273", "date": "17 Dec 2018", "name": "Viktoriia Kosach", "role": "Author Response", "response": "We thank the reviewers for deep and knowledgeable revision of this manuscript. Please, find below our response to the reviewers’ comments point by point. Major point:The putative phosphorylation of TBR2 by S6K1 is too speculative and not convincing to use it as one of the major conclusions of the papers. It should be either confirmed experimentally or removed from the abstract and the main results of the paper. We have withdrawn the mentioned statements from the conclusions and the abstract, and left as a hypothesis in the text of our article. 2. Minor points p4 Please remove “if you are interested in obtaining this antibody, please contact the corresponding author“; We could not remove it, as this sentence is a recommendation of the F1000Research Editorial board, and it informs the other researchers, where they could obtain these antibodies to replicate the study. p4 The anti-ERG and anti-CDX antibodies are not used in the main Figures. Please move information about these antibodies form the Materials and Methods section to legends of the supplementary figures; We agree. We have moved mentioned information to the legend of Dataset 4. Results and discussion: P5 Please remove the first paragraph or move it to the introduction section. P5 “For the first step of the present study” is redundant Figure 1,5 legend. Please replace “positive reaction” with “staining” P6 The following phrase is unclear: “Since cell spreading can be considered as a stage of migration, the previous data led to the hypothesis…” We agree with all comments. We have corrected them in version 2 according to reviewers’ queries. Figure 6. It is unclear whether the images represent confocal sections or reconstituted 3D images. Colocalization of S6K1 and TBR2 should be analyzed quantitatively, eg using ImageJ tools for colocalization analysis. It is also unclear whether TBR2 shuttles between the nucleus and the cytoplasm similarly to S6K1. The images represent confocal sections. Colocalization of S6K1 and TBR2 was analysed using ImageJ JaCoP plugin, as it is indicated in article. As well as S6K1, we observed TBR2 nucleo/cytoplasmic relocalization that depended on cell density. Moreover, TBR2 shuttled between cytoplasm and nuclei similarly to S6K1 during spheroid to monolayer reversion. But quantitatively the level of colocalization of S6K1 and TBR2 we detected on MCF-7 monolayer at different density, because the cells are in more similar condition reliable for image analysis than in course of 3D spheroid transformation into 2D monolayer colony. P9. S6K1 is a serine threonine kinase and the putative phosphorylation of TBR2 by S6K1 should concern threonines rather than tyrosines. I also propose to remove the bioinformatic part and the Figure 8 as they are too speculative. Indeed, many other kinases may potentially phosphorylate TBR2 basing on the bioinformatic analysis, and the authors did not even show whether TBR2 is indeed phosphorylated. We agree that there was erratum concerning tyrosin and threonine. We have fixed it. As co-immunoprecipitation revealed the existence of S6K1-TBR2 protein complex in MCF-7 cells, we considered the theoretical probability of TBR2 phosphorylation by S6K1 kinase. Of course, this does not exclude the presence of other effectors of this transcription factor, but gives us the prerequisites for hypothesis. We removed this statement from the conclusions of the article. Figure 7. TBR2 has a molecular weight of 84 kDa of TBR2, but the major band has a lower molecular weight. This should be explained in the Results section. The antibody supplier also state that anti-TBR2 antibodies detects two bands in immunoblot analysis. We have found information that 4 splice isoforms of TBR2 with lower molecular weight are known today (https://www.uniprot.org/uniprot/O95936)." } ] } ]	1	https://f1000research.com/articles/7-1332
https://f1000research.com/articles/7-420/v1	03 Apr 18	{ "type": "Opinion Article", "title": "The possible implications of advances in genetic testing technologies for Genetic Counsellors working with families of children with developmental disorders", "authors": [ "Flora M. Joseph" ], "abstract": "New genetic testing technologies such as microarrays and whole exome sequencing mean the diagnostic potential for a child with a development disorder is greatly increased over traditional testing techniques. With this increased potential comes increased expectations from families and professionals about the answers a diagnosis will provide. However, limitations remain and a proportion of individuals will continue to remain undiagnosed. In addition, some individuals will receive novel or very rare diagnoses about which very little is known in terms of prognosis and effective treatments. In this paper, I present an argument for why these families would benefit from additional Genetic Counsellor support and how Clinical Genetics services in the UK could provide this support. I acknowledge that resources are limited, but as demands on services increase and interactions with families become shorter, I argue that these resources should be fought for, for the benefit of these families.", "keywords": [ "genetic counseling", "genetic services", "genetic testing", "intellectual disability", "developmental disabilities", "psychosocial support systems", "uncertainty", "psychological adaptation" ], "content": "Introduction\n\nProjects such as the Deciphering Developmental Disorders (DDD) study, and the 100,000 Genomes project, have used whole exome or genome sequencing to pinpoint disease causing mutations, increasing the chance of making a diagnosis in individuals with developmental disorders. This type of technology is a realistic possibility for clinical NHS care in the UK within the not-too-distant future (Hazelton & Petchey, 2015). In this paper, I explore the implications this could have for Genetic Counsellors supporting families of children with developmental disorders as they go through the diagnostic journey.\n\nDevelopmental disorders may manifest in any area of development (e.g. growth, congenital malformations, seizures) but the most common phenotype is developmental delay (DD) or intellectual disability (ID) (Deciphering Developmental Disorders Study, 2015) which are common referral indicators to Clinical Genetic Services. ID is defined by the World Health Organisation as having an Intelligence Quotient (IQ) below 70 and can be classified as mild, moderate, severe or profound (World Health Organisation, 2015). It is estimated to affect approximately 1–3% of the population (Shaffer, 2005). The term ‘Intellectual Disability’ is generally only used when a person is of an age when an IQ test is valid and reliable, which is typically from 5 years old (Shavell et al., 2003). Prior to this the term DD is used. DD may present itself in any of the following areas: gross/fine motor, speech/language, cognition, social/personal, and activities of daily living (Shavell et al., 2003). If a child has delay in two or more of these areas they are said to have Global Developmental Delay (GDD). The estimated prevalence of GDD is about the same as ID at approximately 1–3%. It is estimated that 5–10% of all children have some sort of DD, therefore DD or even GDD may not necessarily lead to ID, but children with ID often had DD in their early years (Shavell et al., 2003).\n\n\nParental responses to having a child with a developmental disorder\n\nIt is natural for parents to hope for a healthy child. The point when a developmental disorder may become apparent can range from a prenatal scan, to birth, to later on as the child grows. This can be a sudden or a gradual realisation (Heiman, 2002; Lewis et al., 2010). Initial reactions to this realisation are likely to be negative such as disbelief, anger, denial, and grief (Graungaard & Skov, 2007; Heiman, 2002; Kearney & Griffin, 2001; Lewis et al., 2010). From this initial reaction, parents must go through a process of adaptation of “replacing the hopes and expectations…with the realities of their child’s actual prognosis” (pg 186, Barnett et al., 2003). This process has been likened to that of a ‘journey’ that at times is like a ‘rollercoaster’ full of highs and lows (Lewis et al., 2010).\n\nFor parents to adjust appropriately to the reality of having a disabled child, they must adopt a number of coping strategies. These may be practically or emotionally-directed (Graungaard & Skov, 2007; Lewis et al., 2010; Rosenthal et al., 2001). Practical coping strategies may be information gathering e.g. about disease prognosis, treatments or research opportunities. Emotional coping strategies may be talking with friends and family or accessing support services. Some of the studies cited earlier went on to explore parental feelings after they have gone through this period of adaptation. They found that many parents have positive and optimistic feelings about their child’s future (Heiman, 2002; Kearney & Griffin, 2001; Lewis et al., 2010).\n\nAlthough the majority of parents appear to adapt well to their new reality, some parents and families are less successful. Traits of less successful adaptation may be: unrealistic appreciation of their child’s weaknesses and limitations; continued feelings of self-pity and guilt; searching for a ‘magical solution’; and feelings of rejection or over-protection of their child, sometimes at the expense of other family members (Kandel & Merrick, 2007). This inability to cope and adapt can have a negative effect on mental health (e.g. stress, depression), relationships and functioning (Barnett et al., 2003).\n\nA developmental disorder can leave families with many questions, some of which can only be answered by obtaining an accurate diagnosis. In this way, searching for a diagnosis can form part of the process of coping and adaptation. Below is a list of reasons given for seeking a diagnosis that has been amalgamated from a range of studies. Most of these studies are of parents of children with developmental disorders, but one includes responses from adults with genetic disorders (Hazelton & Petchey, 2015). A range of methods were used including semi-structured interviews (Lewis et al., 2010; Makela et al., 2009; Rosenthal et al., 2001) and surveys (Hazelton & Petchey, 2015; Limb et al., 2010; Madeo et al., 2012).\n\nTo provide information about progression and prognosis for their child to make life plans and help form realistic expectations for the future\n\nTo be aware of recurrence risk for future pregnancies and whether pre-natal testing or carrier testing is available for relatives\n\nTo guide clinical management e.g. whether any additional clinical surveillance is recommended or whether any treatments, therapies or diets are known to be ineffective/harmful for that condition\n\nTo have ‘a label’. Both positive and negative associations were attributed to this. For example, a benefit is having a term to explain why their child is different from other children. However, a concern is that it may cause their child to be stereotyped and people, such as teachers, to have reduced expectations of their abilities\n\nTo improve access to support services (e.g. education, health or social services). Although this should be based on need regardless of diagnosis, this has been reported to be easier when a diagnosis is known (Rosenthal et al., 2001)\n\nTo improve access to peer support of families in a similar situation as themselves. Without a diagnosis, there can be increased feelings of isolation\n\nTo have ‘an answer’. This can provide psychological relief to keep from wondering why this has happened and increase perceived feelings of control\n\n\nDiagnostic tools employed for individuals with developmental disorders\n\nTraditionally, those fulfilling the clinical features of a recognised syndrome may be diagnosed on clinical examination alone. As technology has advanced, other diagnostic tools have become available, such as metabolic studies, EEG, CT/MRI imaging, cytogenetic studies (e.g. karyotype, FISH studies, etc), and single-gene targeted testing (Rauch et al., 2006; Shavell et al., 2003). Until the advent of more recent technologies (array CGH and next generation sequencing), these tools gave a diagnostic yield for children with GDD/ID in the region of 50–70% (Daily et al., 2000; Rauch et al., 2006). Therefore, up to half of children with GDD/ID remained undiagnosed.\n\n‘Molecular karyotyping’ or ‘array Comparative Genomic Hybridisation’ (array CGH) is a technique that came into clinical practice across the UK during the first decade of the 21st century (Rauch et al., 2006). Array CGH identifies sub-microscopic genetic imbalances across the genome. It can identify copy number variants (CNVs) in the number of genes present which may affect health or development. It allows much more detailed cytogenetic analysis and in most parts of the UK has become a first-line test, superseding traditional cytogenetic studies. In cases where traditional cytogenetic studies have not found a significant result, array CGH gives an average diagnostic yield of 10% (excluding variants of uncertain significance (VUSs)), and this yield rises to 14% for resolutions below 1 Mb (Rauch et al., 2006; Sagoo et al., 2009). As our knowledge advances, it is hoped the classification of VUSs will improve which could increase this diagnostic yield.\n\nThe Deciphering Developmental Disorders (DDD) study was a UK-based study sequencing the exomes of children with undiagnosed developmental disorders (Firth et al., 2011). Exomes are the coding regions of our genes, and account for about 1% of all our genetic material (Wang et al., 2013). The DDD study hoped that a molecular diagnosis could be found by comparison of a child’s exome with their parents’ (Firth et al., 2011). A diagnostic yield of 31% was reported from the first 1133 trios (children and both parents) recruited to the study, and 12 new developmental genes were discovered (Deciphering Developmental Disorders Study, 2015). An American-based study sequencing clinical exomes also found a diagnostic yield of 31% for trios, and 22% for proband-only cases (Lee et al., 2014). The diagnostic yield is expected to increase as new genes are described and further analysis takes place.\n\nAs genetic testing becomes more accessible and diagnostic rates increase, I wonder how this will affect the role of the Genetic Counsellor in supporting these families. Even though diagnostic rates will improve, there will undoubtedly be individuals who remain undiagnosed. In addition, there will be a growing number of individuals who are given a novel molecular diagnosis, about which very little may be known. While these families may receive answers to some of their questions, others will remain unanswered (such as information about prognosis and availability of peer-support from other families who have a child with the same condition). I wonder whether the role of the Genetic Counsellor is currently sufficient to meet the needs of these families, or whether adaptations to the role could better meet the need.\n\n\nSupport needs of families adapting to a child with a developmental disorder\n\nRosenthal et al. (2001) and Lewis et al. (2010) each conducted semi-structured interviews with parents of children with developmental disorders either in the USA (Rosenthal group) or the UK (Lewis group). The purpose of these studies was to find out what impact a lack of diagnosis had on parental adjustment and coping. There was a wide range in the length of time that parents had been aware of their child’s difficulties in both studies, which provides useful information about how coping and adaptation may change over time. Similar reactions to the initial recognition of a problem were reported by parents whether a child had a known diagnosis or not (Lewis et al., 2010). However, in the absence of a diagnosis, additional challenges were noted which could result in a longer period of adjustment and adaptation (Barnett et al., 2003; Heiman, 2002; Lewis et al., 2010; Rosenthal et al., 2001). These included having increased uncertainty about prognosis and a reduced ability to make life plans (e.g. whether their child had a reduced life expectancy, or whether they would be able to live independently as an adult). In addition, the number of investigations arranged, with the aim of seeking a diagnosis, could be both emotionally and physically draining.\n\nA lack of diagnosis means there is greater uncertainty. Madeo et al. (2012) looked at the effect of uncertainty on parental coping and adaptation to raising a child with a developmental disorder. Lipinski et al. (2006) also looked at factors associated with parental uncertainty and perceived control, and the role they played in coping and adaptation for parents of children with a rare chromosome disorder (prevalence of 1/120,000 or lower), which provides useful information about families with novel or very rare conditions. Both studies used a mixed-methods survey (either paper or computer-based) and had relatively large sample sizes (266 and 363 respectively). Madeo et al. (2012) discussed that uncertainty can sometimes aid coping, as it leaves room for optimism of a positive outcome. However, in the majority of cases, both studies found that uncertainty perpetuated a feeling of lacking control over their child’s condition, which was linked to poorer coping and a longer process of adaptation, as reported by Rosenthal et al. (2001) and Lewis et al. (2010). Both Lipinski et al. (2006) and Madeo et al. (2012) found factors that were associated with lower perceived control were being less optimistic about the future and perceiving their child’s condition as more severe. Lipinski et al. (2006) found younger parents felt greater uncertainty but this was not replicated by Madeo et al. (2012), although the latter study population did not have a very wide age range. These studies help to explain why a lack of diagnosis, or the diagnosis of a very rare or novel condition, often results in an extended period of coping and adaptation and therefore why additional support may be appropriate.\n\nAs one parent from the Genetic Alliance UK patient charter commented “We always imagined that getting a diagnosis would be the final piece of the puzzle and the end of the journey, but it now feels as if we are at the very beginning of a new journey” (pg 9, Hazelton & Petchey, 2015). Another parent, whose child was found to have a unique unbalanced translocation, used the term ‘non-diagnosis’ as there was no prognostic information available and commented “It’s like being told something in a foreign language. It wasn’t a relief because I didn’t understand it” (pg 810, Lewis et al., 2010). For these parents the diagnosis had not brought all the answers they had hoped for.\n\nAn example seen through my own clinical practice was when I met parents of a 22 year old man with an overgrowth syndrome and moderate learning difficulties. He had been enrolled in the Childhood Overgrowth (COG) study run by the Institute of Cancer Research in the UK and recently identified as having a de novo pathogenic mutation is the gene DNMT3A. At the time the working name for this new syndrome was ‘DNMT3A overgrowth syndrome’ (Tatton-Brown et al., 2014) but is now called Tatton-Brown-Rahman syndrome (OMIM #615879). In clinic the parents had a strong emotional reaction to finally receiving a diagnosis. The mother commented that, with the lack of a diagnosis, it had always played on her mind whether the cause was due to something she had done during pregnancy, and finally she could put this ‘nagging doubt’ to rest. These parents asked what this meant for prognosis. The oldest reported individual from this study was 29, only 7 years older than this young man. With regards to prognosis, this can only be speculated based on the information already known about this gene. For DNMT3A, it is known to be one of the most commonly mutated genes in acute myeloid leukaemia. Therefore, the authors suggest there may be an increased cancer risk for individuals with a germline mutation, but further study is required to assess this. Although the parents had a very positive reaction to receiving a diagnosis, the uncertainties they were left with raised new questions and concerns which left them with an element of worry for the future. This exemplifies one of the key challenges for novel or rare diagnoses.\n\nParents described how the desire for a diagnosis diminished over time, as the child grew older, however often it never completely went away (Lewis et al., 2010; Rosenthal et al., 2001). In part, this may be due to becoming more familiar with their child’s condition as they grow, and forming a clearer idea of what the future may be like. It addition, this may be due to the realisation that a diagnosis will not change who their child is. However, there are times that re-kindle the desire for a diagnosis e.g. when the child is approaching adulthood and applying for additional support (e.g. supported housing), or when siblings are reaching reproductive age, and additional support could also be appropriate at these times.\n\n\nThe role of the Genetic Counsellor\n\nBiesecker (pg 327, 2001) discussed the goals of genetic counselling and stated that “contemporary genetic counselling should strive to… facilitate clients’ ability to use genetic information in a personally meaningful way that minimises psychological distress and increases personal control”. Families are for the most part resilient, but the process of coping and adaptation still benefits from psychological support, regardless of whether a child receives a diagnosis or not (Barnett et al., 2003; Lewis et al., 2010). Seymour Kessler described two models of practice for Genetic Counsellors, a ‘teaching model’ and a ‘counselling model’ (Kessler, 1997). Kessler suggests that a hybrid approach is adopted incorporating elements of both so that the counselee has received the appropriate information but has also had time and space for discussion of the consequences and personal reflection. However, a number of studies have examined modern practice and found that the ‘teaching model’ is more often adopted, with the main purpose of information provision (Lerner et al., 2014; Meiser et al., 2008; Roter et al., 2006; Walser et al., 2017). Austin et al. (2014) reviewed the available literature and found that a ‘counselling model’ with the aim of addressing the psychosocial concerns, is reported to be associated with increased knowledge retention, reduced anxiety and higher satisfaction with decision-related outcomes. The authors go on to suggest that Genetic Counsellors focus more on this style in their practice. This is supported by Lipinski et al. (2006) who looked at the perceived helpfulness of genetic counselling, for parents of children with developmental disorders. The authors found that it was perceived as more helpful when parents were helped to increase a sense of perceived control over their child’s condition (Lipinski et al., 2006) which fits more with a ‘counselling model’ of practice. A study from the USA looking at the result-giving appointment of exome sequencing highlighted that these appointments were information-heavy and often missed opportunities to build relationships with patients (Walser et al., 2017). The counselling element tended to be neglected due to time restraints and so an on-going relationship with these families would be particular important, to improve understanding, reduce misconceptions, address frustration and disappointment, and improve satisfaction.\n\nGenetic Counsellors have historically had the resources to offer long-term support to families in the UK. However, as demands on Genetic Services have increased, time restraints have limited the amount of contact Genetic Counsellors are able to offer families, and relationships have become shorter-term. Therefore, even though the need for support still exists, families may not request it as they may not recognise where this form of support is best sought (Lipinski et al., 2006). Genetic testing and genetic understanding is infiltrating many areas of healthcare, and the role of Genetic Counsellors as ‘information providers’ is becoming less specialised (Austin et al., 2014) however Genetic Counsellors have a unique set of skills and can play an important role in providing psychological support for families (Lipinski et al., 2006; Middleton et al., 2017). Austin et al. (2014) propose genetic counselling is remodelled “as a time-limited, highly circumscribed psychotherapeutic encounter”. Whereas there is some support for this (Wynn, 2016), in practical terms, due to increased demands and limited resources, from my own experience I feel there is doubt about whether this sort of service is viable in the UK at the moment. I would like to see the profession working towards this sort of model for families of children with developmental disorders and I describe below how this could look in practice.\n\n\nImplications for Genetic Counselling practice\n\nWith the advancement of new genetic technologies described above, diagnostic rates will increase and families will receive more of the answers they seek. I see two key responsibilities for Genetic Counsellors in regards to advanced genetic testing. Firstly, in managing expectations about the limitations of genetic testing and secondly in identifying those families who could benefit from additional support and having the capacity to offer this.\n\nWhen a child with an undiagnosed developmental disorder is referred to Clinical Genetics, the initial contact will often be an information gathering exercise (of family and medical histories) to aid the Consultant Geneticist when considering a diagnosis. This may be in an appointment with a Genetic Counsellor, but may also be with a Family History Coordinator via telephone or by postal questionnaire. Through this process, an assessment may be made about how well the family are adapting to their child’s condition and which families would most benefit from additional support. The main aims of this initial contact, besides factual information gathering, could be:\n\nEliciting concerns and managing expectations\n\nAssessing perceived control\n\nAssessing sources of support\n\nEliciting concerns and managing expectations. For all individuals referred to Clinical Genetics seeking a diagnosis, it is important to address these factors in an initial consultation. For families of children with a developmental disorder, it may be that some of their concerns are hindering their ability to adapt to their child’s condition (Graungaard & Skov, 2007; Rosenthal et al., 2001). Families who place a greater significance on finding a diagnosis may struggle more if one is not made. However, by exploring their reasons for seeking a diagnosis, the Genetic Counsellor may be able to help them find resolution to their concerns, even in the absence of a diagnosis.\n\nAs Genetic Counsellors cannot know where the diagnostic path will lead, it is important to prepare families for each eventuality. It should be highlighted that, even if a diagnosis is found, it may not necessarily bring all the answers they seek. The current diagnostic rate of approximately 50–70% could be given (although this depends whether the child in question has already had any genetic testing, e.g. karyotype or array CGH). However, it is important to find a careful balance to manage expectations both about the possibility of obtaining a diagnosis, and the answers that a diagnosis may or may not provide. Depending on their background, some Genetic Counsellors may feel the need for further training about the limitations of advanced genetic testing technologies to appropriately counsel families.\n\nAssessing perceived control. As Lipinski et al. (2006) and Madeo et al. (2012) found, lower perceived control is associated with being less optimistic about the future and perceiving their child’s condition as more severe. Therefore, families with these indications may benefit from additional support. One suggestion is to use the Revised Life Orientation Test (LOT-R) which rates 10 statements to assess optimism (Madeo et al., 2012). Asking families about their perceptions of the severity of the child’s disease provides a point of reference for when they are assessed by a Consultant Geneticist. Madeo et al. (pg 9, 2012) suggest that “if there exists significant differences in the parent’s and clinician’s perceptions further clarification may reduce the perceived uncertainty of the situation”.\n\nAnother indicator may be the time elapsed since recognition that their child has a developmental disorder. Rosenthal et al. (2001) and Lewis et al. (2010) found the desire for finding a diagnosis decreased with time. Therefore, families nearer the beginning of their journey may benefit more from additional Genetic Counsellor support (Lipinski et al., 2006).\n\nAssessing sources of support. By asking what sources of support the family have around them, or who they talk to about their child, families may be identified that are experiencing isolation and are in greater need of additional intervention.\n\nFrom this initial contact, it may be appropriate to offer additional support with a Genetic Counsellor to families with lower levels of perceived control or a lack of social support. In addition, once all available tests are exhausted, any families that receive a novel or very rare diagnosis, or no diagnosis may also benefit from additional support. As many Genetic Services are already under great pressure with current demands, adding additional support may not currently be viable. I would hope that the value Genetic Counsellors could bring in delivering this kind of service may be used to apply for additional funding and resources so that it could and perhaps a specific clinic set up (or clinic slots within a clinic).\n\nFor families identified as needing additional support, a time-limited intervention (e.g. 1–3 Genetic Counsellor consultations) could be offered. The aims would either be i) in helping them to cope with an uncertain future in the absence of a diagnosis, or ii) to readjust to having a diagnosis for their child, and if this is for a very rare or novel condition, helping them to cope with the uncertainties this brings.\n\nSome reasons for seeking a diagnosis may be addressed even in the absence of a diagnosis, for example, accessing peer-support.\n\nSignposting to other sources of support. Feelings of isolation and not knowing where to turn for support were often cited as reasons for seeking a diagnosis (Lewis et al., 2010; Rosenthal et al., 2001). Parents report a lack of information about what educational, social and psychological help is available (Heiman, 2002). If a Genetic Counsellor is aware of the local groups and resources available, these can be provided. Resources could be provided that contain a broad range of information, such as the roles of different healthcare providers, about educational support and benefits for children with developmental disorders (e.g. the support group Unique (www.rarechromo.org) has a useful booklet called ‘After diagnosis: What happens next? The early years’).\n\nSupport groups can provide peer-support for families in a similar situation. In the UK support groups such as ‘Syndromes Without A Name (SWAN) UK’ (www.undiagnosed.org.uk) has been formed specifically for families of children without a diagnosis.\n\nSupport groups can also be a useful source of information about relevant research projects (Limb et al., 2010). Patients report that they rarely hear of research opportunities relating to their condition from clinicians (Limb et al., 2010). This is understandable to an extent as it is challenging for clinicians to stay abreast of all relevant opportunities when dealing with multiple conditions. Therefore, this should be highlighted to patients as a benefit of being part of a support group. Many families who were recruited to the DDD study first heard of it through SWAN UK (Hazelton & Petchey, 2015).\n\nFor families affected by very rare or novel conditions, such groups may not exist and feelings of isolation can persist (Rosenthal et al., 2001). In the UK, some more general support groups exist such as Rare Disease UK (www.raredisease.org.uk), and Unique (www.rarechromo.org/). Unique was originally set up for families affected by rare chromosomal abnormalities. However, they have now broadened their spectrum to include families affected by very rare single gene disorders, where a more specific support group does not yet exist (personal correspondence, 2015). For these families, the challenges can be very similar to those with rare chromosomal conditions and the peer-support reduces feelings of isolation, which can be immensely empowering (Limb et al., 2010).\n\nHaving a label for their child’s condition. Another concern that could be discussed in the absence of a diagnosis, is having a term or label for their child to use with other professionals, or with friends and family e.g. ‘developmental delay’, ‘a SWAN child’ (after the support group Syndromes Without A Name) (Lewis et al., 2010). As a Genetic Counsellor, or in partnership with the Consultant Geneticist, families could be helped to come up with terminology they can use.\n\nIncreasing perceived control in uncertain situations. For the questions that cannot be answered, e.g. prognosis or recurrence risk, families will continue to have a degree of uncertainty, which may perpetuate a feeling of lacking control. As Lipinski et al. (2006) found, genetic counselling was perceived as more helpful when parents are helped to increase a sense of perceived control over their child’s condition. Madeo et al. (2012) suggests Genetic Counsellors may help families by identifying areas where they do have some control. The authors suggest training in interventions, such as Coping Effectiveness Training “in which individuals identify the controllable and uncontrollable aspects of their situation and are assisted in identifying coping strategies that are predicted to best match the controllability of the stressor” (Madeo et al., 2012).\n\nPractical-focussed strategies, such as information gathering, may give an increased perception of control (Madeo et al., 2012). Lewis et al. (2010) provides some other suggestions given during interviews with parents. Examples of these are (Lewis et al., 2010; Madeo et al., 2012):\n\nKeeping a diary to monitor their child’s condition and progress. This will not only serve as a useful tool when talking to health care providers, but can also act as a reminder of the progress their child has made\n\nDeveloping ‘a passport’ of their child’s likes/dislikes, what they can/can’t do, and their medical problems can aid communication with professionals\n\nLearning about medicine and treatments that are being offered for their child. In this way parents can feel they are making more informed decisions about management\n\nBecoming experts and advocates for their child’s condition. Rosenthal et al. (2001) found that parents who felt informed about their child’s problems felt empowered to act as advocates for them in obtaining support services (e.g. educational).\n\nContributing to fundraising or research for their child’s condition or for rare diseases in general. Rosenthal et al. (2001) found a keenness from parents to engage in activities such as these, so that their child’s disability may benefit others\n\nSetting up a blog or website about their child. This can be both therapeutic and create opportunities for networking and advocacy\n\nHowever, when there is a high level of uncertainty, practical coping strategies may not be successful and may reduce perceived control due to a lack of available information. In these instances, emotion-focussed strategies may increase perceived feelings of control as “one’s internal state may be more amenable to change than the situation itself” (pg 239, Lipinski et al., 2006). These coping strategies may be talking with friends and family, accessing support services, retaining hope and focussing on the positives. Lipinski et al. (2006) reported that parents would have liked Genetic Counsellors to have more hope and encouragement. This is not limited to hope that a diagnosis will be made, but also hope for the future, even in the absence of a diagnosis (e.g. in what support their child may obtain and hope for prognosis) (Graungaard & Skov, 2007).\n\nIn time it is likely that new information/tests will become available, as research continues (Lipinski et al., 2006). It would be important to inform families of the option of a review appointment in Clinical Genetics if any new symptoms/features arise, or in a few years to see if any new tests are available. This will also hopefully lessen any feelings of abandonment.\n\n\nConclusion\n\nIn the age of advanced genetic technologies, expectations have never been higher about the diagnostic potential for individuals with developmental disorders. However, a diagnosis will not necessarily provide all of the answers sought. In addition, the search for a diagnosis could be hindering a family’s ability to accept their child’s condition. Without appropriate support, the process of coping and adaptation could be prolonged or even unsuccessful. It is important that Genetic Counsellors recognise the limitations of these new technologies and continue to support families irrespective of whether a diagnosis is made or not. Genetic Counsellors have the skills and opportunity to provide the support required to address the uncertainties families face, help identify areas where peer support can be found and increase perceived control, facilitating adaptation to improve individual and family functioning. Genetic Services in the UK may currently not have the resources to facilitate an extended support service such as this due to the demands they are already under, however I believe this type of service is the very essence upon which the Genetic Counselling profession was built and is in danger of being lost to the patients cost therefore should be fought for.", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declare that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nThis paper was completed as part of a portfolio for professional registration with the Genetic Counsellor Registration Board (GCRB). I would like to thank Dr Nicki Taverner for her supervision and guidance in writing this paper. I would also like to thank the GCRB and the All Wales Medical Genetics Service for their support.\n\n\nReferences\n\nAustin J, Semaka A, Hadjipavlou G: Conceptualizing genetic counseling as psychotherapy in the era of genomic medicine. J Genet Couns. 2014; 23(6): 903–909. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBarnett D, Clements M, Kaplan-Estrin M, et al.: Building new dreams – Supporting parents’ adaptation to their child with special needs. Infants Young Child. 2003; 16(3): 184–200. Publisher Full Text\n\nBiesecker BB: Goals of Genetic Counseling. Clin Genet. 2001; 60(5): 323–330. PubMed Abstract \| Publisher Full Text\n\nDaily DK, Ardinger HH, Holmes GE: Identification and evaluation of mental retardation. Am Fam Physician. 2000; 61(4): 1059–1067, 1070. PubMed Abstract\n\nDeciphering Developmental Disorders Study: Large-scale discovery of novel genetic causes of developmental disorders. Nature. 2015; 519(7542): 223–8. PubMed Abstract \| Publisher Full Text\n\nFirth HV, Wright CF, DDD Study: The Deciphering Developmental Disorders (DDD) study. Dev Med Child Neurol. 2011; 53(8): 702–703. PubMed Abstract \| Publisher Full Text\n\nGraungaard AH, Skov L: Why do we need a diagnosis? A qualitative study of parents’ experiences, coping and needs, when the newborn child is severely disabled. Child Care Health Dev. 2007; 33(3): 296–307. PubMed Abstract \| Publisher Full Text\n\nHazelton A, Petchey L: Genome Sequencing: What do patients think? Patient Charter. Genetic Alliance UK. 2015. Reference Source\n\nHeiman T: Parents of children with disabilities: Resilience, Coping and Future Expectations. J Dev Phys Disabil. 2002; 14(2): 159–171. Publisher Full Text\n\nKandel I, Merrick J: The child with a disability: parental acceptance, management and coping. ScientificWorldJournal. 2007; 7: 1799–1809. PubMed Abstract \| Publisher Full Text\n\nKearney PM, Griffin T: Between joy and sorrow: being a parent of a child with developmental disability. J Adv Nurs. 2001; 34(5): 582–592. PubMed Abstract \| Publisher Full Text\n\nKessler S: Psychological Aspects of Genetic Counseling. IX. Teaching and Counseling. J Genet Couns. 1997; 6(3): 287–295. PubMed Abstract \| Publisher Full Text\n\nLee H, Deignan JL, Dorrani N, et al.: Clinical exome sequencing for genetic identification of rare Mendelian disorders. JAMA. 2014; 312(18): 1880–1887. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLerner B, Roberts JS, Shwartz M, et al.: Distinct communication patterns during genetic counseling for late-onset Alzheimer's risk assessment. Patient Educ Couns. 2014; 94(2): 170–179. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLewis C, Skirton H, Jones R: Living without a diagnosis: The parental experience. Genet Test Mol Biomarkers. 2010; 14(6): 807–815. PubMed Abstract \| Publisher Full Text\n\nLimb L, Nutt S, Sen A: Experience of rare diseases: An insight from patients and families. Rare Disease UK. 2010. Reference Source\n\nLipinski SE, Lipinski MJ, Beisecker LG, et al.: Uncertainty and perceived personal control among parents of children with rare chromosome conditions: the role of genetic counseling. Am J Med Genet C Semin Med Genet. 2006; 142C(4): 232–240. PubMed Abstract \| Publisher Full Text\n\nMadeo AC, O’Brien KE, Bernhardt BA, et al.: Factors associated with perceived uncertainty among parents of children with undiagnosed medical conditions. Am J Med Genet A. 2012; 158A(8): 1877–1884. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMakela NL, Birch PH, Friedman JM, et al.: Parental perceived value of a diagnosis for intellectual disability (ID): a qualitative comparison of families with and without a diagnosis for their child’s ID. Am J Med Genet A. 2009; 149A(11): 2393–402. PubMed Abstract \| Publisher Full Text\n\nMeiser B, Irle J, Lobb E, et al.: Assessment of the content and process of genetic counseling: a critical review of empirical studies. J Genet Couns. 2008; 17(5): 434–451. PubMed Abstract \| Publisher Full Text\n\nMiddleton A, Marks P, Bruce A, et al.: The role of genetic counsellors in genomic healthcare in the United Kingdom: a statement by the Association of Genetic Nurses and Counsellors. Eur J Hum Genet. 2017; 25(6): 659–661. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPersonal Correspondence: Conversation with Beverly Searle, Chief Executive Officer and Sarah Wynn, Information Officer, when they gave a talk at our Genetics Department. 27th February 2015. 2015.\n\nRauch A, Hoyer J, Guth S, et al.: Diagnostic yield of various genetic approaches in patients with unexplained developmental delay or mental retardation. Am J Med Genet A. 2006; 140(19): 2063–2074. PubMed Abstract \| Publisher Full Text\n\nRosenthal ET, Biesecker LG, Biesecker BB: Parental attitudes toward a diagnosis in children with unidentified multiple congenital anomaly syndromes. Am J Med Genet. 2001; 103(2): 106–114. PubMed Abstract \| Publisher Full Text\n\nRoter D, Ellington L, Erby LH, et al.: The Genetic Counseling Video Project (GCVP): Models of Practice. Am J Med Genet C Semin Med Genet. 2006; 142C(4): 209–220. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSagoo GS, Butterworth AS, Sanderson S, et al.: Array CGH in patients with learning disability (mental retardation) and congenital anomalies: updated systematic review and meta-analysis of 19 studies and 13,926 subjects. Genet Med. 2009; 11(3): 139–146. PubMed Abstract \| Publisher Full Text\n\nShaffer LG: American College of Medical Genetics guideline on the cytogenetic evaluation of the individual with developmental delay or mental retardation. Genet Med. 2005; 7(9): 650–654. PubMed Abstract \| Free Full Text\n\nShavell M, Ashwal S, Donley D, et al.: Practice parameter: evaluation of the child with global developmental delay: report of the Quality Standards Subcommittee of the American Academy of Neurology and The Practice Committee of the Child Neurology Society. Neurology. 2003; 60(3): 367–380. PubMed Abstract \| Publisher Full Text\n\nTatton-Brown K, Seal S, Ruark E, et al.: Mutations in the DNA methyltransferase gene DNMT3A cause an overgrowth syndrome with intellectual disability. Nat Genet. 2014; 46(4): 385–388. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWalser SA, Werner-Lin A, Mueller R, et al.: How do providers discuss the results of pediatric exome sequencing with families? Per Med. 2017; 14(5): 409–422. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWang Z, Liu X, Yang BZ, et al.: The role and challenges of exome sequencing in studies of human diseases. Front Genet. 2013; 4: 160. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWorld Health Organisation: International Classification of Disease-10 Version: 2015. 2015; accessed 14th December 2017. Reference Source\n\nWynn J: Genomic Testing: a Genetic Counselor's Personal Reflection on Three Years of Consenting and Testing. J Genet Couns. 2016; 4: 691–697. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "32695", "date": "22 May 2018", "name": "Maria Soller", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThank you for giving me the opportunity to review this manuscript.\nI think that this article highlights several important aspects. New genomic technologies will add to the complexity of genetic disorders. Both pre-test and post-test counselling are of great importance to understand what information the new technologies can bring as well of course the limitations of the tests and results.\nI do also think as the writer states that patients as well as relatives might need even more support related to the results (both when no diagnosis is set or when a very rare diagnosis is found, there is limited knowledge is present).\nIt is thus important to be able to help and support the patients in that situation and as the author states in a citation; receiving the result can be not the end but the start of the journey. Genetic counsellors’ roles might also change and develop in this genomic area.\nSo as a subject I think this is an important paper. However, I have some major issues regarding this paper, which need major revisions prior to possible indexing:\nTitle I think the title is long and not very clear. I think it needs to be shortened, given a more focused topic.\nAim For me the aim (focus) of the manuscript is not completely clear. What I can see there are two different focuses: 1) What support do these parents need regarding acceptance and coping 2) How can genetic counsellors take part in this area\nI think the manuscript should benefit from either focus on one of those aims OR to clarify that there are 2 aims/focuses.\nStructure The manuscript should benefit from changing the structure of the text. I think it is a little unstructured moving back and forward between things.\nIntroduction I think the introduction is too long and to detailed on the different definitions on ID. It is not necessary for the aims of the manuscript.\nI would also change the order of the subtitles of the introduction. After the first part on the definition and prevalence I think it would be more relevant of having the section on \"Diagnostic Tools\". I think that this part also can be shortened.\nThen I think that the section on parental responses and support needs to be put together as one section. There is a bit of overlap between them. Check for overlaps. I do think the section on the different coping issues is very good and relevant.\nAt the end of the left column page 3 I would change the focus from \"how it will affect the role of genetic counsellors.....\" to how it would affect the need for support and genetic counselling and if so, how genetic counsellors can participate in this area of support.\nRole of genetic counsellors I think it is a very interesting approach that GCs can play a role here. However it lacks discussions on what unique skills GCs have to take this role. What is included in GC education? The author talks in a sentence about that GCs need to be more prepared to do this and that the current knowledge is not sufficient and that further education might be needed. I would like to see a development of the discussion on this and references to studies on the education and role of genetic counsellors.\nExample I do believe that it is good to show things by examples and I think the citations from parents is a good ways of showing parents thoughts and experiences. However, for me the example of childhood overgrowth stands a little on it is own and does not add very much here. I recommend to take that out.\nConclusion I think the conclusion could be clearer and should benefit from clarifying the aim/aims of the study.\nIn summary, I think it is an interesting paper within an interesting area, but which needs major revision to be indexed.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Partly\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? No", "responses": [ { "c_id": "3678", "date": "30 May 2018", "name": "Flora Joseph", "role": "Author Response", "response": "Thank you very much for reviewing my article. You make some valuable points and I will look to revise the article and incorporate them." } ] }, { "id": "34843", "date": "28 Jun 2018", "name": "Vicki Wiles", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nComments by Sue Kenwrick, Principal Genetic Counsellor The author sets out some of the potential implications for families with a child affected by a developmental disorder who are undergoing broad genetic testing such as array CGH or exome analysis. She proposes an increased role for genetic counsellors (GCs) in helping families adapt to genetic test results and proposes a model where GCs hold clinics prior to genetic testing to help manage patient expectations and to identify individual family needs. The background overview of the impact of having a child with a developmental disorder, the search for a diagnosis and research into factors that help or hinder adjustment is thought provoking and well structured. The author should also mention large projects doing whole genome as well as exome analysis for completion (such as 100,000 genome project). She sets out well the complex needs of families in these situations and the need to pay attention to emotional needs (counselling) as well as educational needs (teaching) of parents going through testing. In particular, she highlights the need for management of expectations surrounding a test and a need to signpost/refer parents to additional resources/support groups that may further facilitate adjustment and coping. Clearly, these families would benefit from incorporation of more of the ‘counselling’ aspect of their consultations in order to optimise their adjustment. However, there is more than one way this could be achieved. Personally I think the idea of introducing a ‘preconsultation’ with a GC prior to broad genetic testing by another clinician (as suggested in the authors model) is a retrograde step for the profession. I think this would be a great strain on limited GC availability at a time when we are taking steps to deal with increase in referrals by doing more genetic counselling by telephone. For many genetic services models already exist for pregathering of information by family history questionnaire for certain conditions and some of this is done by coordinators that are not trained in genetic counselling. This non-GC workforce would not be equipped to ask about family coping strategies or perceived control as seems to be implied in the authors model (under initial contact). While it is valid and laudable to highlight GC input and support for families with a new diagnosis or uncertain result, I don’t think this shift in service provision provides the answer. Rather, any provider (whether Clinical Geneticist or GC) offering a genetic investigation should be equipped to discuss managing expectations and support the family by signposting to resources or offering additional consultations, or referrals as required. Depending on the centre, It is already common for GCs to see families after a ‘genomic’ result to facilitate understanding and assess the families further needs and this could be an open offer or case by case. Additional training is something that might be required for some providers. Another part of the solution could be multimedia resources that facilitate the process of testing and follow-up (e.g. online videos about having a genetic test and how it doesn’t always give a clear answer or improved follow-up literature for families). There is no doubt, however, that, as the author points out, GCs are skilled at managing the psychosocial aspects of genetic counselling and there will be a need to increase the workforce as more complex genetic investigations and results are provided. For readers outside clinical genetics services, I think the author should distinguish between the terms genetic counselling and genetic counsellor, as this can be confusing for those in other specialties. Genetic counselling is done by different types of provider (Clinical Geneticists, GCs and, in some specialties, by specialist nurses who have had genetics training). Genetic counsellors, however, are a body of allied health professionals from science or nursing background, ‘usually’ not MDs, who are trained in genetic counselling. Overall, as this is an opinion piece and raises interesting issues surrounding broad genetic testing. I approve publication with reservation based on a limited exploration or potential solutions and significant reservation about the model proposed.\n\nComments by Vicki Wiles, Consultant Genetic Counsellor\nThis review and opinion piece by a Genetic Counsellor from the Wales Genetic Counselling service raises interesting questions about the use of Genetic Counselling resources. The support and counselling issues for families with children with disorders of developmental are well described and the summary of research papers supports the argument that Genetic Counsellors are well trained in counselling skills, with an understanding of genetic concepts and testing and are therefore equipped to help families with children with undiagnosed disorders. I would agree that there is often a need for support that a GC can provide but this support might also be provided by the family Health Visitor, a paediatric Clinical Nurse Specialist or the Consultant Geneticist working with the family. I would have liked to see more discussion on how genetic counselling varies across England and the devolved Countries, with an acknowledgement that some services have continued to do pre-clinic work sometimes including home visits, with longer term involvement, while other services have moved to more autonomous working for Genetic Counsellors focussing more on genetic testing where there is known diagnosis in the family. Funding and workforce differences in England have influenced the type of work that genetic counsellor’s focus on and I would have welcomed reference to this. The article would benefit from a broader health resources overview to set the author’s views in context including the financial side to her argument, which might well underpin her views. As genomics moves more into mainstream medicine there will be many, possibly conflicting, demands on genetic counsellor’s time with patients and so this article is timely and merits publication to stimulate debate.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Partly\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Partly", "responses": [] }, { "id": "34842", "date": "09 Jul 2018", "name": "Andrew Cuthbert", "expertise": [ "Reviewer Expertise genetic counselling", "psychiatry" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nTitle Too long and tentative, be succinct and give a definite message to attract the reader.\n\nAim Needs a clearer definition of the aims. Is it to define the challenges faced by parents of children with DDs, diagnosed or otherwise, as well as the how well genetic counsellors address these issues? Key aim is a justification for extending genetic counsellor’s role in the post genomic diagnosis period, make this very clear early on. Needs to consider and compere alternative models of family support.\n\nStructure Somewhat confused, needs clearly define and describe each topic and refrain from commentary therein, rather to have clearly subtitled commentary paragraphs where appropriate, either as a standalone section or at the end of each topic, and be consistent.\n\nIntroduction Unnecessary detail on definitions of DD and ID, some of which is inaccurate. Should think about the audience who may not be familiar with jargon such as exome sequencing. It would help to describe typical referral pathways to clinical genetics, most referral for DD/ID are from paediatric neurology services who frequently request array tests and return results prior to visiting genetics. How will this change, if at all, as genome sequencing is commissioned into the NHS. IS there a different perspective in the US, Europe, elsewhere?\n\nThe description of diagnostic measures and yields is misleading. The focus ought, to be, briefly, on the yield (actually 15-23%) of current methods in use in UK services (or wherever the author feels this is relevant). Clinical microarray was originally commissioned in around 2007 (NHS England), ie it’s not so modern. Then should describe potential impacts of integrating exome/genome sequencing with increased genomic diagnoses (around 70%) and how this will impact current GC practice.\n\nNeeds a paragraph on the psychosocial imperatives of a genetic diagnosis/non-diagnosis and the role of GCs (or others) in counselling for this. Following this present a concluding paragraph outlining the commentary’s aims and objectives – and why there is, arguably, any need to change current practice.\n\nGenetic Counsellors and Genetic Counselling Genetic counsellors are not the sole providers of genetic counselling (needs also to acknowledge clinical geneticists have been counselling families for as long as GCs). The commentary should be cognisant of the role of nurse specialists who provide counselling, ie that genetic counselling is not the exclusive the domain of trained GCs.\n\nSince developments in genome technology are inevitably influencing developments in service design and delivery, in the UK and other countries, the author should consider their impact on GC roles in supporting families post-diagnosis - as pre-test counselling becomes increasingly obsolete as has been happening for several years due primarily to technological developments and their superior diagnostic accuracy. Argue whether any change in service delivery might be justifiable.\n\nSupport The author should recognise and comment that a majority, probably vast majority, of families engage support from places other than clinical genetics, especially over the medium-long term. Genetics services role in supporting families is circumscribed and predominantly short-term.\n\nThe nature of support is vaguely described, what exactly is the support given in a genetic counselling encounter (by a genetic counsellor or other practitioner) which would facilitate improved adjustment and where is the justification to develop a specialist post-test/post-genomic diagnosis support service as distinct from routine follow-up. A significant Would this be a genetic counsellor led service? How would this differ from current open-door policy where parents are able to access follow-up contact with the clinician/practitioner. The author could consider how signposting and information content and provision could be improved, particularly how to ensure support group contact is increased, training for rare disease support groups to tackle complexities associated with single gene diagnoses for DD/ID after exome/genome sequencing, as with 100k genomes. Arguably, well informed peer/third sector support is more accessible, immediate and broader in relevance than highly specialised genetic counsellors/clinical specialists and perhaps better placed to facilitate longer term ‘community’ support, family acceptance and adjustment.\n\nPsychological support for families with children with severe/life limiting illness, more generally, isn’t dependent on the disease specialist being trained to provide such additional ‘support/care’. Arguably this applies to aspects of rare disorders care ad management. Moreover, given the extent and complexity of mental health risks, impacts and needs of children with DD/ID, combined with the psychological and social struggles of their parents/carers, who experience high levels of anxiety, depression and trauma (as the author acknowledges), are there alternative models of ‘support’ which are more rigorous and wide-ranging than conventional genetics follow-up, for which mental health and voluntary sector providers would be better placed to provide? A clear definition of the purpose, content and contextual relevance of ‘support’ needs to be given, as forms the basis of the authors justification for a significant change in service.\n\nConclusion This is a thought provoking commentary on a subject of increasing importance and at a time when genomic medicine’s relevance and promise to improve lives is subject to intense scrutiny, presented at a time when certain aspects clinical genetics services are experiencing perceived existential threats. Despite its shortcomings, need for structural correction and lack of coherent justification for what would be a significant change to current service provision which is counter to the evolving emphasis of genetic counselling training programmes, this kind of commentary is welcome and strongly supported. A revised version would be very welcome.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Partly\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Partly", "responses": [] } ]	1	https://f1000research.com/articles/7-420
https://f1000research.com/articles/7-1939/v1	17 Dec 18	{ "type": "Research Article", "title": "Cortical auditory evoked potentials and hemispheric specialization of speech in individuals with learning disability and healthy controls: A preliminary study", "authors": [ "Mayur Bhat", "Hari Prakash Palaniswamy", "Arivudai Nambi Pichaimuthu", "Nitha Thomas", "Mayur Bhat", "Arivudai Nambi Pichaimuthu", "Nitha Thomas" ], "abstract": "Background: Dichotic listening (DL) technique is a behavioral non-invasive tool which is used in studying hemispheric lateralization. Previous studies using behavioral DL have hypothesized that individuals with learning disabilities (LD) exhibit a lack of cortical specialization for processing speech stimulus. However, there is no event related potential (ERP) evidence, hence the main objective of the study is to explore hemispheric asymmetry using cortical auditory evoked potential (CAEPs) in normal hearing adults and also to compare the same in children with LD and healthy controls. Methods: CAEPs were recorded in 16 normal hearing young adults, eight right-handed children with LD and their age matched controls. Two stop constants (/Pa/ – voiceless, bilabial, stop: /Ta/ - voiceless, alveolar, stop) were chosen for this experiment and presented in each ear and dichotically in two different orders (/pa-ta/, /ta-pa/). ERPs were processed using a standard pipeline, and electrodes readings over the left and right hemispheres were averaged to create left and right regions of interest (ROI). The CAEPs were analyzed for mean amplitude and peak latency of P1-N1-P2 components. Results: The current study results suggest no statistically significant difference between the two stimulus in monaural condition and absence of order effect in dichotic condition. In healthy controls the CAEP latencies were shorter over the left hemisphere in both monaural and dichotic conditions in adults and control children. However, it was very evident that such a difference was lacking in children with LD. Conclusions: Hemispheric asymmetry can be detected using CAEPs for speech stimulus. The measures are consistent and void of stimulus or order effect. Taken together, the findings of current study, both monaural and dichotic condition illustrates the hemispheric differences in processing speech stimuli in normal hearers. Absence of latency differences between hemispheres in children with LD indicate a lack of hemispheric asymmetry.", "keywords": [ "CAEPs", "hemispheric asymmetry", "dichotic listening", "learning disability" ], "content": "Introduction\n\nThe human brain is comprised of two hemispheres, and both hemispheres differ from each other in terms of anatomy as well as physiology. Among the two hemispheres, one is more active and demonstrates superior performance on specific tasks. This phenomenon is referred to as brain dominance or hemispheric asymmetry. This brain dominance seems to be related to the handedness of the person. In humans, brain dominance or asymmetry seems to be established early in fetal development. Dichotic listening (DL) is one of the conventional methods to study this cerebral dominance effect (Ahonniska et al., 1993). In this test, two different auditory signals are presented to two ears independently, and the listeners are expected to recognize the signals presented to both ears. The DL test is sensitive to hemisphere differences to specific sounds (Brancucci et al., 2008). The principle of this test is that speech is lateralized to the left hemisphere (Tervaniemi & Hugdahl, 2003), resulting in the individual preferring to repeat the stimulus presented to the right ear more often than the left ear. The vice versa is true when it comes to non-speech stimulus (Brancucci et al., 2008). These effects are termed as right ear advantage (REA), and left ear advantage (LEA) respectively and highly correlates with the Wada-test (Hugdahl et al., 1997). Since, it is a simple, effective and non-invasive equivalent of the Wada-test, it has been widely used in the assessment of various clinical populations. Therefore the DL technique has been used widely as a measure of cortical processing and auditory perception for several decades (Hugdahl et al., 1997).\n\nDL technique has been found to be useful in studying language lateralization in children with early focal brain damage (Brizzolara et al., 2002; Carlsson et al., 1992; Chilosi et al., 2005; Isaacs et al., 1996), aphasia (Bavosi & Rupp, 1984; Johnson et al., 1977; Johnson et al., 1978; Pettit & Noll, 1979; Selnes et al., 1983), and stuttering (Brady & Berson, 1975; Blood & Blood, 1986; Curry & Gregory, 1969; Foundas et al., 2004; Gruber & Powell, 1974; Robb et al., 2013; Slorach & Noehr, 1973; Strub et al., 1987). The DL test is particularly useful in the assessment of children with learning disability. The DL tests have been used to reveal cerebral dominance deficits (van den Noort et al., 2008), subtypes of children with dyslexia (Cohen et al., 1992), developmental changes in language lateralization (Porter & Berlin, 1975), and bilateral hemispheric processing deficits (Obrzut & Mahoney, 2011) in children with LD.\n\nTraditionally, DL has been assessed using behavioral methods. However, electrophysiological methods, especially cortical auditory evoked potentials (CAEPs), would help in understanding the neurophysiology of dichotic listening. Since the left hemisphere is dominant for speech and language function, CAEPs in DL is evidenced by larger amplitudes and shorter latencies over the left hemisphere (Bayazit et al., 2009; Eichele et al., 2005; Friedrich et al., 2017; Haaland, 1974; Morrell & Salamy, 1971). However, these studies have either focused on latency (Eichele et al., 2005) or amplitude (Haaland, 1974; Morrell & Salamy, 1971), but not both. Hence, the complete cortical dynamics underlying processing is not yet known. Analyzing both these measures will provide information on the strength of cortical activation, as well as the efficiency of neural conduction.\n\nDichotic study involves the presentation of one stimulus to right ear and other stimulus to left ear. Previous studies have reported cortical changes using dichotic, but have not explored stimulus effect or order effect (Bayazit et al., 2009; Eichele et al., 2005; Friedrich et al., 2017; Haaland, 1974; Morrell & Salamy, 1971). The physiological responses elicited using different stimulus may differentially influence evoked ERPs, since these are obligatory responses to external stimuli. It is important to rule out stimulus specific effects before commenting on cortical asymmetry using this paradigm. Hence, the current study aimed at studying the stimulus effect in monotic (/pa/ vs. /ta/) and order effect in dichotic condition (/pa-ta/ vs. /ta-pa/). The study also aimed at comparison of monaural vs. dichotic processing differences in the same individual. This will provide important insight on how cortical processing of dichotic listening differs from that of monaural listening.\n\nPrevious studies using behavioral DL have hypothesized that individuals with LD exhibit a lack of cortical specialization for processing speech stimulus. To date there is no literature evidence for this using ERPs. A handful of studies have utilized CAEPs to study DL in children with LD, and revealed a comparable amplitude between hemispheres indicating decreased cortical asymmetry for speech stimulus (Brunswick & Rippon, 1994). Hence studying cortical processing of dichotic listening using ERPs will further validate these findings. In this view, there is a definite need for a study to establish the monaural and dichotic auditory processing differences as reflected by CEAPs in healthy individuals and those with LD.\n\n\nMethods\n\nThe study was carried out at the Department of Speech & Hearing, School of Allied Health Science, and Manipal. The study began on 1st August 2016 and continued till 20th August 2017. The study protocol was approved by Institutional Ethics Committee (IEC), Kasturba Hospital, Manipal (IEC 460/2016).\n\nThis study was a prospective observational study where 16 normal young adults (18–25 years), eight normal learning right-handed children (7–15 years) for the control group and eight right-handed individuals with LD (7–15 years) were recruited. Healthy volunteers were either students at the School of Allied Health Sciences, who were recruited through advertisement through notice board or members of the public who visited the department. They had the study explained to them and were recruited if interested in participating. All the Volunteers were provided with participant information sheet which had complete details of the study. Written informed consent was taken from all interested individuals prior to participation in the study.\n\nAll the participants were screened for the presence of hearing loss (Pure Tone Audiometry done using duly calibrated Madsen Astera (American National Standard Institute S3.43-1996) should be <15dBHL (decibels Hearing Level) for both air conduction and bone conduction tests, and middle ear dysfunction (Tympstar middle ear analyzer, Grason-Stadler Inc., MN, USA). All the tests were carried out by investigators (audiologist) at Dept. of Speech and Hearing, School of Allied Health Sciences). Individuals who were diagnosed with a learning disability at the Department of Psychology, SOAHS, Manipal University, Manipal and concented to participate in the study were included in the experimental group. Edinburg’s handedness inventory (Oldfield, 1971) was administered to the participants, and only right-handed individuals were selected for the study because handedness is considered as a major variable that affects cortical asymmetry. (Delorme & Makeig, 2004)\n\nTwo speech sounds (/Pa/ – voiceless, bilabial, stop: /Ta/ - voiceless, alveolar, stop,), were selected as stimulus to elicit ERP. Syllables (/Pa/ and /Ta/) were used as a stimulus for both monaural and dichotic paradigms. Also, similar stimuli has been shown to be effective in eliciting LLR and used in studying cerebral asymmetry in the literature (Lawson & Gaillard, 1981). The above syllables were recorded using a standard microphone kept at a 6cm distance from the mouth (Extended data (Palaniswamy, 2018b)). A normal native Kannada speaker was asked to produce these two syllables with normal intensity and normal intonation. The dichotic stimulus was prepared using Adobe Audition version 1.0, where stimulus /pa/ was stored in the right channel and /ta/ was stored in the left channel to create a single / pa-ta/ dichotic stimuli, and vice-versa to create /ta-pa/ dichotic stimuli. Duration of the stimulus was trimmed so that both the stimulus had the same duration.\n\nAll the measurements were carried out in an acoustically treated room. The ‘SOUND’ module of Stim system (Version 2) was used for stimulus presentation with inserts at an intensity of 70dBSPL. CEAPs were recorded using the ‘Acquire’ module of the SynAmps2 amplifier (Compumedics NeuroScan, Abbotsford, Australia). A 32 channel electrode cap used with combined mastoid as reference. Impedance at all electrode sites was maintained below 5k Ohms. Raw EEG recording were acquired with a bandpass filters set between 0 and 100Hz with a sampling rate of 1000/sec. The obtained EEGs were analyzed offline using a filter from 1 – 30 Hz, and artifact rejection was also be done offline (EEGLab version 13_6_5b (Delorme & Makeig, 2004)).\n\nThe stimulus was presented in 3 conditions.\n\n1. A monaural condition in which stimulus (/pa/ and /ta/) was presented to the right ear only\n\n2. A monaural condition in which stimulus (/pa/ and /ta/) was presented to the left ear only.\n\nIn both, the monaural conditions patient will be asked to watch a silent movie and ignore the stimulus presented to the ear.\n\n3. Dichotic passive attention condition in which the stimulus (/pa-ta/ and /ta-pa/) was presented to both ears simultaneously and the patient will be asked not to pay attention to the stimulus.\n\nThe current study used two stimuli (/pa/ and /ta/) to record CAEPs. In monaural condition, two stimuli were presented to each ear one at a time to check whether the stimulusaffects CAEPs. Similarly, in dichotic condition, /pa/-/ta/ was presented to the right and left rear respectively, and the reverse of it, i.e., /ta/-/pa/ was used to check if reversal of stimulus order had any effect on CAEPs. Thus a total of 6 conditions were obtained from a single participant.\n\nRaw EEG data were imported to EEGLab version 13_6_5b (Delorme & Makeig, 2004), a free software commonly used for analyzing EEG/ERP signals offline, which runs on MATLAB (2010a). The following preprocessing steps were done serially on each data to obtain a final average waveform. After editing channel locations (BESA 4 shell dipfit spherical model) bad channels and bad blocks were visually inspected and interpolated using spherical interpolation method in the command line in MATLAB. The data was then subjected to high pass filtering with a cut-off frequency of 1kHz. Bin based epochs were extracted using ERPLAB version 6.14 (Luck, 2014) between -200 to 800ms timelocked to stimulus onset, and then were baseline corrected for the prestimulus duration (-200 to 0ms). Independent component analysis (ICA) was done to decompose multivariate ERP waveform into their subcomponent based on their source using the ‘runica’ command in EEGLAB, then analysed using MARA 1.1 (Multiple Artefact Rejection Algorithm) (Winkler et al., 2011) which automatically removes the components with artifacts based on several parameters. Post artifact rejection, the waveforms were low pass filtered with a cutoff frequency of 30Hz and then rereferrened to common average. All the epochs were averaged in ERPLAB.\n\nA region of interest (ROI) is one of the prescribed methods of analyzing ERPs where few neighboring electrodes that represents a particular anatomical area for a specific purpose are selected for analysis rather than a single electrode. The basis of this type of analysis is mainly on certain assumptions. The first reason is to explore one's data. It is often useful to see the activity in areas of interest plotted for each condition or plotted against other variables. The second reason is to control Type I error by restricting the number of statistical tests to a few ROIs. The third reason is to limit testing to a specific brain region that is defined functionally by some information (Poldrack, 2007).\n\nIn the current study, two 2 ROIs with three electrodes in each hemisphere were selected, here in after synonymously referred to as right and left hemisphere electrodes. Left ROI was an average of 3 electrodes (C3, FC3, and CP3), and the right hemisphere ROI was an average of homologs of the these three electrodes (C4, FC4, CP4). For example, the latency of N1 component from C3, FC3, and CP3 are 108ms, 110ms, 112ms respectively; then the left hemispheric ROI is 110ms. The right and left hemispheric ROI were obtained for all the three conditions. Hence monaural right condition included monaural right ear right hemisphere electrodes (MonoR RH), and monaural right ear left hemisphere electrodes (MonoR LH). Monaural left conditions included monaural left ear right hemisphere (MonoL RH) electrodes, and monaural left ear left hemisphere electrodes (MonoL LH). Dichotic conditions included dichotic right hemisphere electrodes (DI RH) and dichotic left hemisphere electrodes (DI LH).\n\nGrand mean average waveform across participants was used as a reference to decide the latency range of measurement. In the current study, for all the conditions across groups, P1 mean amplitude and peak latency was measured between 40 to 80 msec. Similarly, 90 to 140 msec, and 170 to 220 msec windows were used for N1 and P2 respectively. These mean amplitudes and peak latency measures for right and left ROIs were automatically measured using the measurement toolbox of ERPLAB for each participant and the output was written in .txt format then later exported to MS Excel 2016 and SPSS version 15 (SPSS Inc., Chicago)\n\n\nResults\n\nAll the data were first tested for normality using Shapiro-Wilk’s test, and the results showed that the latencies and amplitudes of all the components were normally distributed.\n\nSince the use of two stimuli is inevitable in dichotic listening, it was a must to rule out any stimulus effect on CAEPs. In monaural condition, these two stimuli (/pa/ vs. /ta/) did not result in significant latency or amplitude difference in both right and left ear (Table 1). Similarly, in dichotic condition comparison of two stimuli in a different order (/pa-ta/ vs. /ta-pa/) also did not lead to any significant difference (Table 1) which ruled out order effect. Given this, data was combined across stimuli for rest of the analysis. ANOVA with repeated measures (3×2×3) was carried out to check for main effect and interaction effect.\n\nResults showed significant main effect between groups on P1 latency (F (2, 29) =23.50, p<0.001, ŋ2= 0.618). Post hoc analysis revealed the shortest latency in adults with normal hearing compared to the other two children groups, which was statistically significant (p<0.001). Though children with normal hearing had shorter latencies than children with LD, the latency difference did not reach significance (p= 0.08).\n\nFurther there was a significant main effect of hemisphere on P1 latency (F (1, 29) =31.8, p<0.001, ŋ2= 0.523) and there was a significant interaction between hemisphere and group (F (1, 29) =3.2, p=0.04, ŋ2= 0.184). These results were analyzed further by combining the condition and running a paired ‘t’ test on the data for different groups. Results showed a significantly shorter latency over the left hemisphere when compared to the right hemisphere in adults with normal hearing (p<0.001), and children with normal hearing (p<0.001). However, such hemispheric difference was not significant in children with LD (p=0.08) (Table 2). There was no significant main effect of conditions on P1 latency (F (2, 58) =1.609, p=0.20, ŋ2= 0.053) (Figure 1 a, b and c).\n\nLH – Left Hemisphere, RH – Right Hemisphere\n\nGraphical representation of P1 mean latency across (a) dichotic condition (b) Monaural Left conditions (c) Monaural Right condition in all three groups. The error bar represents +/- standard deviation. LH – Left Hemisphere, RH – Right Hemisphere.\n\nResults also revealed that there is not any main effect of either groups (F (1, 29) =1.9, p=.08, ŋ2=0.12), condition (F (1.7, 51.5) =3.1, p=.06, ŋ2=0.09), or hemispheres (F (1, 29) =0.049, p=0.82, ŋ2=0.002) on P1 amplitude (Table 2) (Figure 2 a, b and c).\n\na) Graphical representation of P1 mean amplitude across (a) dichotic condition (b) Monaural Left conditions (c) Monaural Right condition in all three groups. The error bar represents +/- standard deviation. LH – Left Hemisphere, RH – Right Hemisphere.\n\nResults showed a significant main effect of group on N1 latency (F (2, 29) =4.1, p=0.02, ŋ2 =0.223). Post hoc results showed no significant difference between any of the groups (p = 0.066), though similar a developmental pattern as P1 was seen in N1 latency also.\n\nFurther there was a significant main effect of hemisphere on N1 latency too (F (1, 29) =19.2, p<0.001, ŋ2 =0.399), and also a significant interaction between hemisphere and group (F (2, 29) =5.4, p=0.01, ŋ2 = 0.27). These results were analyzed further by combining the condition and running a paired ‘t’ test on the data for different groups. Similar to P1, N1 latency showed a significantly shorter latency over left hemisphere when compared to right hemisphere for adults with normal hearing (p<0.001), and children with normal hearing (p<0.001). However, such hemispheric difference was not significant in children with LD (p=0.716) (Table 3).\n\nLH – Left Hemisphere, RH – Right Hemisphere\n\nThere was also a significant main effect of condition on N1 Latency (F (2, 58) =5.9, p=0.04, ŋ2 =0.16). Post hoc results showed significant latency difference between dichotic and monaural left condition (p= 0.04) were N1 latency was shorter in dichotic condition compared to the monoaural left condition. Such significance was not seen in any of other combinations (DI vs. MR and ML vs. MR) (Figure 3 a, b and c).\n\nGraphical representation of N1 mean latency across (a) dichotic condition (b) Monaural Left conditions (c) Monaural Right condition in all three groups. The error bar represents +/- standard deviation. LH – Left Hemisphere, RH – Right Hemisphere.\n\nResults showed no significant main effect of either group (F (1, 29) =2.8, p=0.07, ŋ2=0.162) or condition (F (1.5, 44.4) =2.1, p=0.132, ŋ2=0.06) on N1 amplitude. But there was a significant main effect of hemisphere on N1 amplitude (F (1, 29) =11.2, p=0.002, ŋ2=0.276), and also there was significant interaction between condition and hemisphere (F (1.9, 56.1) =8.5, p=0.001, ŋ2= 0.227). Further analysis revealed significantly larger amplitude over the left hemisphere in both the dichotic and monaural right condition (p<0.001), and no such latency difference in the monaural left condition (p=0.893) (Table 3) (Figure 4 a, b and c).\n\nGraphical representation of N1 means amplitude across (a) dichotic condition (b) Monaural Left conditions (c) Monaural Right condition in all three groups. The error bar represents +/- standard deviation. LH – Left Hemisphere, RH – Right Hemisphere.\n\nResults showed no significant main effect of either group (F (2, 29) = 3.2, p=0.053, ŋ2 =0.184), or condition (F (2, 58) = 0.79, p=0.42, ŋ2 =0.027) on P2 latency.\n\nFurther, there was a significant main effect of hemispheres on P2 latency (F (1, 29) = 4.2, p=0.04, ŋ2 =0.28), and there was no significant interaction between hemisphere and group (F (1, 29) = 1.309, p=0.286, ŋ2 =0.083) (Table 4) (Figure 5 a, b and c).\n\nLH – Left Hemisphere, RH – Right Hemisphere\n\nGraphical representation of P2 mean latency across (a) dichotic condition (b) Monaural Left conditions (c) Monaural Right condition in all three groups. The error bar represents +/- standard deviation. LH – Left Hemisphere, RH – Right Hemisphere.\n\nANOVA results showed no significant main effect of either group (F (1, 29) =2.2, p=0.1.7, ŋ2=0.133), hemisphere (F (1, 29) =1.42, p=0.264, ŋ2=0.04) or condition (F (1.6, 46.4) =0.73, p=0.486, ŋ2=0.02) on P2 amplitude (Table 4) (Figure 6 a, b and c).\n\nGraphical representation of P2 mean amplitude across (a) dichotic condition (b) Monaural Left conditions (c) Monaural Right condition in all three groups. The error bar represents +/- standard deviation. LH – Left Hemisphere, RH – Right Hemisphere.\n\n\nDiscussion\n\nThis article explores the hemispheric asymmetry in three groups using CAEPs in a dichotic and monotic paradigms. While the preliminary aim of the study understands the neurophysiology of dichotic processing, the fact that monaural differences in CEAPs itself are not well understood. Hence it is worthwhile to discuss these findings in detail for the sake of better understanding of typical auditory processing.\n\nIn monaural stimulus condition, the results confirmed that the stimulus effect was negligible since the latencies evoked by stops in the current study (\|pa\| and \|ta\|) were comparable. Further, it was observed that the latencies of P1, N1 and P2 components in the left hemisphere were shorter in latency compared to right irrespective of the ear stimulation (Figure 7, Figure 8, Figure 10 and Figure 11).\n\nWaveforms clearly depict shorter latency over left hemisphere than right hemisphere. LH – Left Hemisphere, RH – Right Hemisphere.\n\nWaveforms clearly depict shorter latency over left hemisphere than right hemisphere. LH – Left Hemisphere, RH – Right Hemisphere.\n\nSimilarly, dichotic condition resulted in insignificant order effect, and the CEAP components P1, N1,and P2 had significantly shorter latency over the left hemisphere compared to the right (Figure 9 and Figure 12), essentially the same results as in monaural stimulus condition. It is difficult to compare earlier studies using CEAPs in dichotic listening tasks since all the CAEP components were not studied. Nevertheless, N1 latency in the left temporal electrode was shown to have 5 ms shorter latency than that of the homologues of the right (Eichele et al., 2005). Another recent study reported left central electrodes were 8 ms shorter than that of the homologues of the right region (Friedrich et al., 2017). They hypothesized that, under high perceptual load, the N1 predicts perceptual preferences (Eichele et al., 2005). However, in the current study a similar effect was seen even in monotic listening conditions. Hence, it can be said that there could be a common perceptual preference mechanism for both monaural and dichotic listening conditions and could be interpreted in the light of hemispheric specialization for speech processing.\n\nWaveforms clearly depict shorter latency over left hemisphere than right hemisphere. LH – Left Hemisphere, RH – Right Hemisphere.\n\nWaveforms clearly depict shorter latency over left hemisphere than right hemisphere. LH – Left Hemisphere, RH – Right Hemisphere.\n\nWaveforms clearly depict shorter latency over left hemisphere than right hemisphere. LH – Left Hemisphere, RH – Right Hemisphere.\n\nWaveforms clearly depict shorter latency over left hemisphere than right hemisphere. LH – Left Hemisphere, RH – Right Hemisphere.\n\nSeveral behavioral and imaging studies in the literature have unanimously suggested that, the left hemisphere is specialized for processing speech and language related information (Ci et al., 2016; Hinkley et al., 2016; Ishikawa et al., 2017; Morrell & Salamy, 1971; O’Grady et al., 2016; Witelson & Pallie, 1973). Though the pathway and the neural substrates are fundamentally similar, due to unknown reasons, it is proven that the left auditory cortex is characterized to have a specialized function for speech stimuli (Corina et al., 1992; Witelson & Pallie, 1973).\n\nCAEP amplitude is a variable measure as a whole (van Hedel et al., 2007). In the current study, P1 and P2 amplitude in monaural conditions were larger over the left hemisphere when compared to the right. Similar results were seen for N1 amplitude too, except in the monaural left condition. In the dichotic condition, all the CAEP components showed larger amplitude over the left hemisphere than right. Previous studies done using structured magnetic resonance imaging methods has shown similar results (Dos Santos Sequeira et al., 2006). However, In the current study, none of the amplitude measures reached significance, this may be due to the low sample size in the current study or due to the inherent variance of amplitude measures.\n\nIn children with LD, there was no significant latency difference between hemispheres in both monaural and dichotic stimulus conditions (Figure 4.7, 4.8 and 4.9 Figure 13, Figure 14 and Figure 15). This finding is very consistent between the P1, N1 and P2 components of CEAP. Similar findings were reported in earlier studies using neuroimaging studies on dichotic listening, where these individuals showed symmetrical activation of the bilateral auditory cortex (Illingworth & Bishop, 2009; Njemanze, 1991). Concerning amplitude, there was no significant trend.\n\nWaveforms clearly depict no significant latency difference over left hemisphere and right hemisphere. LH – Left Hemisphere, RH – Right Hemisphere.\n\nWaveforms clearly depict no significant latency difference over left hemisphere and right hemisphere. LH – Left Hemisphere, RH – Right Hemisphere.\n\nWaveforms clearly depict no significant latency difference over left hemisphere and right hemisphere. LH – Left Hemisphere, RH – Right Hemisphere\n\nIn 1978, Galaburda, Geschwind, and colleagues hypothesized that patients with learning disabilities was associated with disruptions in brain asymmetry (Galaburda et al., 1978). Few authors have argued that there could be atypical asymmetry in these individuals (Cohen et al., 1992; Foster et al., 2002; Hugdahl et al., 1997), and others have stated that there is reduced normal asymmetry (Martínez & Sánchez, 1999; Koltuska & Grabowska, 1975). The current study findings are in line with the latter hypothesis rather than the former. Previous studies have reported reduced cortical asymmetries in brain regions including planamtemporale asymmetry (Foster et al., 2002; Galaburda et al., 1978; Galaburda et al., 1985; Leonard & Eckert 2008), corpus callosum abnormalities that of the larger total callosal areas and larger posterior (splenial) areas (Duara et al., 1991), smaller anterior-most regions (genu) (Hynd et al., 1995), larger posterior third of the callosum including the isthmus and splenium (Rumsey et al., 1996). Apart from these, several other structures also have been reported to lack asymmetry including parietal areas (Habib & Robichon, 1996), the posterior region of the inferior frontal gyrus (Galaburda et al., 1985; Hynd et al., 1990), and Broca’s area (Robichon et al., 2000).\n\nThough the current study findings could easily attribute to the established anatomical deficits that are associated in children with LD, the functional asymmetry/ deficits in decoding phonological information cannot be completely ruled out. Since speech processing is well differentiated from nonspeech stimulus right from the brainstem, as earlier evidence suggest (Abrams et al., 2006; Ibañez et al., 1989), the observed lack of asymmetry could be a combination of both functional phonological decoding deficits as well as the structural deficits. Further, a method that elucidates the speech-specific processing from non-speech processing is at this moment warranted.\n\nIn the current study, there were significant differences between normal adults and individuals with LD in terms latency of CEAP components (P1 and N1), were the latencies were shortest in adults with normal hearing, shorter in children with normal hearing, and prolonged in children with LD. Previous ERP studies on these individuals have shown mixed results. Though few authors have observed no latency difference in auditory late latency response (ALLR) using click stimulus except for P1 (Purdy et al., 2002), other studies suggest that individuals with LDs often have prolonged latency when compared with controls in all ALLR components (Frizzo, 2015; Kumar & Gupta, 2014). The delay may be because of altered cortical functions (Pinkerton et al., 1989), short attention span (Picton et al., 1978), or deficits in auditory cortical information synchronization associated to auditory attention factors (Leppänen & Lyytinen, 1997).\n\nTaken together, the findings of current study, both monaural and dichotic condition elucidates the hemispheric differences in processing speech stimuli in normal hearers. At the same time, these effects are either suppressed or absent in LDs. However, there is no previous evidence to support these findings. Hence the results have to be interpreted with caution and open for exploration.\n\n\nConclusion\n\nThe current study method is unique in comparison to previous CEAP studies, and is consistent in indicating cerebral asymmetry in normal hearers and LDs. The study failed to categorize learning disability subjects based on their specific learning disability. A lack of behavioral dichotic listening tests supplementing the electrophysiological findings can be considered as one of the major drawbacks of the current study. Overall results indicate that shorter latency and larger amplitude in the left hemisphere irrespective of the ear of presentation may indicate left hemispheric preferences for speech stimulus in normal, but lack of this difference suggests void of hemispheric asymmetry in individuals LDs. However, there is no previous evidence to support these findings. Hence the results have to be interpreted with caution and are open for exploration. Hence, based on this preliminary evidence, it can be suggested that CEAPs can be used as one of the tools to study cerebral asymmetry. Latencies of CEAP components are more sensitive to hemisphere specific difference than amplitude.\n\n\nData availability\n\nUnderlying data is available from Figshare\n\nFigshare: Dataset 1. P1, N1 and P2 Latency for all the group across all the condition\n\nhttps://doi.org/10.6084/m9.figshare.7358387.v1 (Palaniswamy, 2018a)\n\nFigshare: Dataset 2. P1, N1 and P2 Amplitude for all the group across all the condition\n\nhttps://doi.org/10.6084/m9.figshare.7358396.v1 (Palaniswamy, 2018b)\n\nFigshare: Dataset 3. Stimulus effect and order effect for Latency and Amplitude\n\nhttps://doi.org/10.6084/m9.figshare.7358417.v1 (Palaniswamy, 2018c)\n\nAll data is available under a CC0 1.0 Universal license\n\nStimuli used to evoke Late Latency Response\n\nFigshare: Extended data. Stimuli used for Evoking ALLR https://doi.org/10.6084/m9.figshare.7358438.v1 (Palaniswamy, 2018d)\n\nLicense: CC0 1.0 Universal", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nAbrams DA, Nicol T, Zecker SG, et al.: Auditory brainstem timing predicts cerebral asymmetry for speech. J Neurosci. 2006; 26(43): 11131–11137. PubMed Abstract \| Publisher Full Text\n\nAhonniska J, Cantell M, Tolvanen A, et al.: Speech perception and brain laterality: the effect of ear advantage on auditory event-related potentials. Brain Lang. 1993; 45(2): 127–146. PubMed Abstract \| Publisher Full Text\n\nBavosi R, Rupp RR: Dichotic abilities in children, normal adults and aphasic adults for open- and closed-context words. J Aud Res. 1984; 24(4): 265–278. PubMed Abstract\n\nBayazit O, Oniz A, Hahn C, et al.: Dichotic listening revisited: trial-by-trial ERP analyses reveal intra- and interhemispheric differences. Neuropsychologia. 2009; 47(2): 536–545. PubMed Abstract \| Publisher Full Text\n\nBlood IM, Blood GW: Relationship between specific disfluency variables and dichotic listening in stutterers. Percept Mot Skills. 1986; 62(1): 337–338. PubMed Abstract \| Publisher Full Text\n\nBrady JP, Berson J: Stuttering, dichotic listening, and cerebral dominance. Arch Gen Psychiatry. 1975; 32(11): 1449–1452. PubMed Abstract \| Publisher Full Text\n\nBrancucci A, Penna SD, Babiloni C, et al.: Neuromagnetic functional coupling during dichotic listening of speech sounds. Hum Brain Mapp. 2008; 29(3): 253–264. PubMed Abstract \| Publisher Full Text\n\nBrizzolara D, Pecini C, Brovedani P, et al.: Timing and type of congenital brain lesion determine different patterns of language lateralization in hemiplegic children. Neuropsychologia. 2002; 40(6): 620–632. PubMed Abstract \| Publisher Full Text\n\nBrunswick N, Rippon G: Auditory event-related potentials, dichotic listening performance and handedness as indices of lateralisation in dyslexic and normal readers. Int J Psychophysiol. 1994; 18(3): 265–275. PubMed Abstract \| Publisher Full Text\n\nCarlsson G, Hugdahl K, Uvebrant P, et al.: Pathological left-handedness revisited: dichotic listening in children with left vs right congenital hemiplegia. Neuropsychologia. 1992; 30(5): 471–481. PubMed Abstract \| Publisher Full Text\n\nChilosi AM, Pecini C, Cipriani P, et al.: Atypical language lateralization and early linguistic development in children with focal brain lesions. Dev Med Child Neurol. 2005; 47(11): 725–730. PubMed Abstract\n\nCi H, van Graan A, Gonzálvez G, et al.: Mandarin functional MRI Language paradigms. Brain Behav. 2016; 6(10): e00525. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCohen M, Hynd G, Hugdahl K: Dichotic listening performance in subtypes of developmental dyslexia and a left temporal lobe brain tumor contrast group. Brain Lang. 1992; 42(2): 187–202. PubMed Abstract \| Publisher Full Text\n\nCorina DP, Vaid J, Bellugi U: The linguistic basis of left hemisphere specialization. Science. 1992; 255(5049): 1258–1260. PubMed Abstract \| Publisher Full Text\n\nCurry FK, Gregory HH: The performance of stutterers on dichotic listening tasks thought to reflect cerebral dominance. J Speech Hear Res. 1969; 12(1): 73–82. PubMed Abstract \| Publisher Full Text\n\nDelorme A, Makeig S: EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. J Neurosci Methods. 2004; 134(1): 9–21. PubMed Abstract \| Publisher Full Text\n\nDos Santos Sequeira S, Woerner W, Walter C, et al.: Handedness, dichotic-listening ear advantage, and gender effects on planum temporale asymmetry--a volumetric investigation using structural magnetic resonance imaging. Neuropsychologia. 2006; 44(4): 622–636. PubMed Abstract \| Publisher Full Text\n\nDuara R, Kushch A, Gross-Glenn K, et al.: Neuroanatomic differences between dyslexic and normal readers on magnetic resonance imaging scans. Arch Neurol. 1991; 48(4): 410–416. PubMed Abstract \| Publisher Full Text\n\nEichele T, Nordby H, Rimol LM, et al.: Asymmetry of evoked potential latency to speech sounds predicts the ear advantage in dichotic listening. Brain Res Cogn Brain Res. 2005; 24(3): 405–412. PubMed Abstract \| Publisher Full Text\n\nFoster LM, Hynd GW, Morgan AE, et al.: Planum temporale asymmetry and ear advantage in dichotic listening in Developmental Dyslexia and Attention-Deficit/Hyperactivity Disorder (ADHD). J Int Neuropsychol Soc. 2002; 8(1): 22–36. PubMed Abstract \| Publisher Full Text\n\nFoundas AL, Corey DM, Hurley MM, et al.: Verbal dichotic listening in developmental stuttering: subgroups with atypical auditory processing. Cogn Behav Neurol. 2004; 17(4): 224–232. PubMed Abstract\n\nFriedrich P, Ocklenburg S, Heins N, et al.: Callosal microstructure affects the timing of electrophysiological left-right differences. NeuroImage. 2017; 163: 310–318. PubMed Abstract \| Publisher Full Text\n\nFrizzo AC: Auditory evoked potential: a proposal for further evaluation in children with learning disabilities. Front Psychol. 2015; 6: 788. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGalaburda AM, LeMay M, Kemper TL, et al.: Right-left asymmetrics in the brain. Science. 1978; 199(4331): 852–856. PubMed Abstract \| Publisher Full Text\n\nGalaburda AM, Sherman GF, Rosen GD, et al.: Developmental dyslexia: four consecutive patients with cortical anomalies. Ann Neurol. 1985; 18(2): 222–233. PubMed Abstract \| Publisher Full Text\n\nGruber L, Powell RI: Responses of stuttering and non-stuttering children to a dichotic listening task. Percept Mot Skills. 1974; 38(1): 263–264. PubMed Abstract \| Publisher Full Text\n\nHaaland KY: The effect of dichotic, monaural and diotic verbal stimuli on auditory evoked potentials. Neuropsychologia. 1974; 12(3): 339–345. PubMed Abstract \| Publisher Full Text\n\nHabib M, Robichon F: Parietal lobe morphology predicts phonological skills in developmental dyslexia. Brain and Cognition. 1996. Reference Source\n\nHinkley LB, Marco EJ, Brown EG, et al.: The contribution of the corpus callosum to language lateralization. J Neurosci. 2016; 36(16): 4522–4533. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHugdahl K, Carlsson G, Uvebrant P, et al.: Dichotic-listening performance and intracarotid injections of amobarbital in children and adolescents. Preoperative and postoperative comparisons. Arch Neurol. 1997; 54(12): 1494–1500. PubMed Abstract \| Publisher Full Text\n\nHynd GW, Hall J, Novey ES, et al.: Dyslexia and corpus callosum morphology. Arch Neurol. 1995; 52(1): 32–38. PubMed Abstract \| Publisher Full Text\n\nHynd GW, Semrud-Clikeman M, Lorys AR, et al.: Brain morphology in developmental dyslexia and attention deficit disorder/hyperactivity. Arch Neurol. 1990; 47(8): 919–926. PubMed Abstract \| Publisher Full Text\n\nIbañez V, Deiber MP, Fischer C: Middle latency auditory evoked potentials in cortical lesions. Critical of interhemispheric asymmetry. Arch Neurol. 1989; 46(12): 1325–32. PubMed Abstract \| Publisher Full Text\n\nIllingworth S, Bishop DV: Atypical cerebral lateralisation in adults with compensated developmental dyslexia demonstrated using functional transcranial Doppler ultrasound. Brain Lang. 2009; 111(1): 61–65. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIsaacs E, Christie D, Vargha-Khadem F, et al.: Effects of hemispheric side of injury, age at injury, and presence of seizure disorder on functional ear and hand asymmetries in hemiplegic children. Neuropsychologia. 1996; 34(2): 127–137. PubMed Abstract \| Publisher Full Text\n\nIshikawa T, Muragaki Y, Maruyama T, et al.: Roles of the Wada Test and Functional Magnetic Resonance Imaging in Identifying the Language-dominant Hemisphere among Patients with Gliomas Located near Speech Areas. Neurol Med Chir (Tokyo). 2017; 57(1): 28–34. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJohnson JP, Sommers RK, Weidner WE: Dichotic ear preference in aphasia. J Speech Hear Res. 1977; 20(1): 116–29. PubMed Abstract \| Publisher Full Text\n\nJohnson JP, Sommers RK, Weidner WE: In response to dichotic ear preference in aphasia: another view. J Speech Hear Res. 1978; 21(3): 601–603. Publisher Full Text\n\nKoltuska B, Grabowska A: Instability of hemispheric asymmetry in dyslexic children. language. 1975.\n\nKumar P, Gupta RK: Cortical processing of speech in children with Dyslexia. Scientific paper presented in 46 th ISHACON held in ….2014.\n\nLawson EA, Gaillard AW: Evoked potentials to consonant-vowel syllables. Acta Psychol (Amst). 1981; 49(1): 17–25. PubMed Abstract \| Publisher Full Text\n\nLeonard CM, Eckert MA: Asymmetry and dyslexia. Dev Neuropsychol. 2008; 33(6): 663–681. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLeppänen PH, Lyytinen H: Auditory event-related potentials in the study of developmental language-related disorders. Audiol Neurootol. 1997; 2(5): 308–340. PubMed Abstract \| Publisher Full Text\n\nLuck SJ: An introduction to the event-related potential technique. 2014. Reference Source\n\nMartínez JA, Sánchez E: Dichotic listening CV lateralization and developmental dyslexia. J Clin Exp Neuropsychol. 1999; 21(4): 519–534. PubMed Abstract \| Publisher Full Text\n\nMorrell LK, Salamy JG: Hemispheric asymmetry of electrocortical responses to speech stimuli. Science. 1971; 174(4005): 164–166. PubMed Abstract \| Publisher Full Text\n\nNjemanze PC: Cerebral lateralization in linguistic and nonlinguistic perception: analysis of cognitive styles in the auditory modality. Brain Lang. 1991; 41(3): 367–380. PubMed Abstract \| Publisher Full Text\n\nO’Grady C, Omisade A, Sadler RM: Language lateralization of a bilingual person with epilepsy using a combination of fMRI and neuropsychological assessment findings. Neurocase. 2016; 22(5): 436–442. PubMed Abstract \| Publisher Full Text\n\nObrzut JE, Mahoney EB: Use of the dichotic listening technique with learning disabilities. Brain Cogn. 2011; 76(2): 323–331. PubMed Abstract \| Publisher Full Text\n\nOldfield RC: The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia. 1971; 9(1): 97–113. PubMed Abstract \| Publisher Full Text\n\nPettit JM, Noll JD: Cerebral dominance in aphasia recovery. Brain Lang. 1979; 7(2): 191–200. PubMed Abstract \| Publisher Full Text\n\nPicton TW, Woods DL, Proulx GB: Human auditory sustained potentials. II. Stimulus relationships. Electroencephalogr Clin Neurophysiol. 1978; 45(2): 198–210. PubMed Abstract \| Publisher Full Text\n\nPinkerton F, Watson DR, McClelland RJ: A neurophysiological study of children with reading, writing and spelling difficulties. Dev Med Child Neurol. 1989; 31(5): 569–581. PubMed Abstract \| Publisher Full Text\n\nPoldrack RA: Region of interest analysis for fMRI. Soc Cogn Affect Neurosci. 2007; 2(1): 67–70. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPorter RJ Jr, Berlin CI: On interpreting developmental changes in the dichotic right-ear advantage. Brain Lang. 1975; 2(2): 186–200. PubMed Abstract \| Publisher Full Text\n\nPalaniswamy HP: P1, N1 and P2 Latency across all the group and condition. figshare. Dataset. 2018a.\n\nPalaniswamy HP: P1, N1 and P2 Amplitude across all the group and condition. figshare. Dataset. 2018b.\n\nPalaniswamy HP: Stimulus effect and order effect for Latency and Amplitude. figshare. Fileset. 2018c.\n\nPalaniswamy HP: Stimuli used for Evoking ALLR. figshare. Fileset. 2018d.\n\nPurdy SC, Kelly AS, Davies MG: Auditory brainstem response, middle latency response, and late cortical evoked potentials in children with learning disabilities. J Am Acad Audiol. 2002; 13(7): 367–82. PubMed Abstract\n\nRobb MP, Lynn WL, O’Beirne GA: An exploration of dichotic listening among adults who stutter. Clin Linguist Phon. 2013; 27(9): 681–693. PubMed Abstract \| Publisher Full Text\n\nRobichon F, Levrier O, Farnarier P, et al.: Developmental dyslexia: atypical cortical asymmetries and functional significance. Eur J Neurol. 2000; 7(1): 35–46. PubMed Abstract \| Publisher Full Text\n\nRumsey JM, Casanova M, Mannheim GB, et al.: Corpus callosum morphology, as measured with MRI, in dyslexic men. Biol Psychiatry. 1996; 39(9): 769–775. PubMed Abstract \| Publisher Full Text\n\nSelnes OA, Knopman DS, Niccum N, et al.: Computed tomographic scan correlates of auditory comprehension deficits in aphasia: a prospective recovery study. Ann Neurol. 1983; 13(5): 558–566. PubMed Abstract \| Publisher Full Text\n\nSlorach N, Noehr B: Dichotic listening in stuttering and dyslalic children. Cortex. 1973; 9(3): 295–300. PubMed Abstract \| Publisher Full Text\n\nStrub RL, Black FW, Naeser MA: Anomalous dominance in sibling stutterers: evidence from CT scan asymmetries, dichotic listening, neuropsychological testing, and handedness. Brain Lang. 1987; 30(2): 338–350. PubMed Abstract \| Publisher Full Text\n\nTervaniemi M, Hugdahl K: Lateralization of auditory-cortex functions. Brain Res Brain Res Rev. 2003; 43(3): 231–246. PubMed Abstract \| Publisher Full Text\n\nvan den Noort M, Specht K, Rimol LM, et al.: A new verbal reports fMRI dichotic listening paradigm for studies of hemispheric asymmetry. NeuroImage. 2008; 40(2): 902–911. PubMed Abstract \| Publisher Full Text\n\nvan Hedel HJ, Murer C, Dietz V, et al.: The amplitude of lower leg motor evoked potentials is a reliable measure when controlled for torque and motor task. J Neurol. 2007; 254(8): 1089–1098. PubMed Abstract \| Publisher Full Text\n\nWinkler I, Haufe S, Tangermann M: Automatic classification of artifactual ICA-components for artifact removal in EEG signals. Behav Brain Funct. 2011; 7: 30. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWitelson SF, Pallie W: Left hemisphere specialization for language in the newborn. Neuroanatomical evidence of asymmetry. Brain. 1973; 96(3): 641–646. PubMed Abstract \| Publisher Full Text" }	[ { "id": "42063", "date": "29 Jan 2019", "name": "Sandeep Maruthy", "expertise": [ "Reviewer Expertise Audiology", "Auditory evoked potentials", "Immittance audiometry" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI am happy to have reviewed the article 'Cortical auditory evoked potentials and hemispheric specialization of speech in individuals with learning disability and healthy controls: A preliminary study'. The need for the study is well supported by the literature, authors have used cogent method and the results are reported well in light of hemispheric asymmetry in children with Learning disability. The study has the potential to make significant contributions to the understanding of hemispheric specialization in speech processing. However, I have few important concerns which I would like the authors to address, in order to bring more clarity and precision in the presentation of their findings.\nAbstract\nIt is not appropriate to state that there are no ERP studies. You may state that ‘there is dearth of studies’. ‘However, there is no event related potential (ERP) evidence, hence the main objective of the study is to explore hemispheric asymmetry using cortical auditory evoked potential (CAEPs) in normal hearing adults and also to compare the same in children with LD and healthy controls.’ Split this into two sentences. ‘main objective of the study is to explore hemispheric asymmetry using cortical auditory evoked potential (CAEPs) in normal hearing adults and also to compare the same in children with LD and healthy controls’. I am not convinced with this statement. I think the main objective of this study is to assess hemispheric specialization in LD and the compare the same with a group of control subjects. The stimuli are syllables not consonants. The syllables should be uniformly reported in IPA. P1, N1 and P2 are independent components. Do not report as ‘P1-N1-P2’. ‘The current study results suggest no statistically significant difference between the two stimulus in monaural condition and absence of order effect in dichotic condition’. This is in which group? ‘In healthy controls’, who are the healthy controls? Overall, the abstract needs to be rewritten with the help of an English expert. In Dichotic tests, the two stimuli are presented simultaneously, not independently. ‘CAEPs in DL is evidenced by larger amplitudes and shorter latencies over the left hemisphere’. Please specify the stimulus used in these studies (speech versus non-speech). Authors need to justify why the findings of independent studies on latency and amplitude cannot be taken together to understand the cortical dynamics. Why does a new study with both together need to be carried out? ‘Stimulus-specific’ instead of ‘stimulus specific’.\n\nMethods\nAvoid using the term ‘healthy volunteers’. How did the authors ensure that the control participants did not have auditory processing deficits? In this study, it is not sufficient to ensure that they have normal hearing sensitivity. Syllables to be reported in IPA. How was LD diagnosed? Give more information about the degree of impairment. 6cm distance from whom? What does ‘Extended data’ refer to? What do authors mean by ‘normal native speaker’? Write it as ‘Speaker of Kannada’ not ‘Kannada speaker’. Define Kannada. I think the authors mean neutral tone and not ‘normal intonation’. Why were participants not tested with behavioural DL? If tested, they could have been sure of normal binaural integration in their control participants. Further, relating the ERP findings with behavioural DL would have shed more light into the underlying mechanisms. What was the final duration of the stimuli? Was any kind of quality judgement done for the stimuli? Was normalization carried out for the two stimuli to keep the intensity equivalent?\n\nConsider providing waveforms of the stimuli, maybe with the time alignment as presented in the experiment. CEAPs to be changed to CAEPs. ‘A 32 channel electrode cap used with combined mastoid as reference’. Add ‘was’ in between. The spacing between the digit and the corresponding unit (For example, 100 Hz) is not correct in most places. ‘artifact rejection was also be done offline’ - please rephrase. ‘The stimulus was presented in 3 conditions’. Please rephrase this because you presented more than one stimulus in the third condition. ‘In both, the monaural conditions patient will be asked to watch a silent movie and ignore the stimulus presented to the ear.’ Needs to be changed to past-tense. Same comment for the next sentence. Why is it ‘passive attention condition’? The authors, I believe, did not have an active attention condition. As I understand, the stimulus paradigm has not changed the order of the stimuli but only the ear to which they were delivered. In terms of time, in dichotic condition, /pa/ and /ta/ were presented simultaneously. In such a case, it should not be termed as '. The paragraphs on stimulus effect and the order effect needs to be rewritten to bring better clarity to the readers. I suggest that the section on ‘stimulus’ be shifted before the section on ‘LLR recording’ What was the total number of stimulus presentations? What was the task of the participants, particularly during dichotic listening? If they were passive, what were they doing during the recording? timelocked should read as ‘time-locked’. Low pass should read as low-pass. Similarly for high pass. ‘In the current study, two 2 ROIs with three electrodes in each hemisphere were selected, here in after synonymously referred to as right and left hemisphere electrodes’. Split this into two sentences. ‘Here in after’ should read as ‘Hereafter’. ‘the latency of N1 component from C3, FC3, and CP3 are 108ms, 110ms, 112ms respectively; then the left hemispheric ROI is 110ms’. Start the sentence with an ‘if’. How was ‘mean amplitude’ measured?\n\nResults\nI do understand that there was no significant difference between /pa/ and /ta/ in their latency or amplitude. That does not mean that there were no differences at all. Therefore, I am not convinced that the LLRs elicited for two different stimuli are clubbed together. This is likely to increase the variability in the data and mask subtle differences if any in the future analysis. I do see changes in the latency and amplitude LLR for different stimuli in the grand average waveform. Therefore, instead of comparing the mean amplitudes and mean latencies, a point-to-point comparison of the waves would have given a better idea of the differences between the stimuli. ‘significant main effect between groups’ should be written as ‘significant main effect of group’. ‘Though children with normal hearing had shorter latencies than children with LD, the latency difference did not reach significance’. Did children with LD have hearing loss? Please use the name of the groups uniformly.\n\n‘p’ is not reported uniformly. In some places the exact ‘p’ is reported while in others, it reports p as <0.01. Please bring uniformity. The names of groups in tables and figures are not acceptable. Participants with LD cannot be termed as ‘LDs’. Other groups cannot be termed as ‘normal groups’. Figure 1 is expected to represent data of P1, but it has latency of P2. Giving the same data in tables as well as figures will add redundancy. Please give only one. The title of the figures and tables needs to be lot more descriptive and precise. Number of decimal points given in the table could be restricted to 2. Consider clubbing together all the grand average waves in one panel if the journal permits.\n\nDiscussion\n\nRephrase ‘While the preliminary aim of the study understands the neurophysiology of dichotic processing, the fact that monaural differences in CEAPs itself are not well understood’. ‘CEAPs’ to be changed to ‘CAEPs’ - throughout the article.\n\nConclusions\n‘The study failed to categorize learning disability subjects based on their specific learning disability’. This was not the objective anyway and no analysis was meant to do this. Why is this coming here? Again, avoid using the term LDs. Conclusion will be better if it is precise and short. Present the most salient contributions for your research here. This can be followed by caveats and limitations. Please write the implications of your findings.\n\nGeneral Comments\nThe manuscript needs a thorough editing of grammar, punctuation and spellings.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "43254", "date": "22 Feb 2019", "name": "Hemanth Narayan Shetty", "expertise": [ "Reviewer Expertise Hearing aids", "electrophysiology", "speech perception", "tinnitus" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors have assessed hemispheric asymmetry using. CAEPs. /pa/ and /ta/ stimuli in three conditions (mono R, L and dichotic (two orders)) which were used to record CAEPs on normal hearing adults, young adults children with age matched learning disabled children. All participants were right handed individuals. Results revealed a significant hemispheric asymmetry was observed in normal adults and adult children but not in LD children, irrespective of stimuli and conditions (MonL L hemisphere/ right hemisphere; monR R hemisphere/ left hemisphere and DI RH and DI LH).\n\nSpecific comments\nIntroduction Recent studies were reviewed to strengthen their need. Research questions and hypothesis are missing. I felt purpose needs to still be strengthened on importance of CAEPs and hemispheric asymmetry. In addition, consider to mention the objectives of the study.\n\nMethod Research design is missing. Justification for stimuli specifically for hemispheric asymmetry is needed. Authors have told that in previous studies similar stimuli were used to assess hemispheric asymmetry thus we also used. This explanation is not correct. Consider to justify. Is the stimuli is normalized? If yes, how have you normalized it? Since CAEP is exogenous potentials it is preferred to give acoustic characteristics of both stimuli (Spectrogram and spectra). These stimuli were presented at 70 dB HL. How these stimuli are calibrated? Consider to write the procedure of calibration. Authors have used high intensity to deliver the stimuli, is there any specific reason?\n\nROI is well explained. Ocular channel is activated? If yes, please specify. I know authors have removed bad channels and bad blocks were visually inspected. An eye blink induces artifacts and has an amplitude similar to that of a response.\nResults Authors have used appropriate statistical analyses to prove the aim of study. A repeated measure ANOVA with between subject factors as group was used to assess hemispheric asymmetry in each component of CAEP (peak latency and amplitude). I feel it is two way repeated measure (? condition* 2 stimuli) with between subject factor as groups. Thus, I suggest authors to mention the factors (323). In P1, N1 and P2 latencies what is the post hoc test used? Sometimes authors have used condition and hemispheric asymmetry interchangeably. Consider to maintain the same term through out the manuscript.\n\nStimuli* condition result is mentioned and what about the interaction effect of stimuli* condition* group. Though main effect of ‘group’ is not significant but interaction of stimuli and condition may have an effect on group. Consider to give the result of stimuli* condition* group. In condition there was a significant difference and authors have used paired sample t test. There are three groups (normal, young adult and LD). If these three groups are considered then authors should use alpha corrections. Anyway a significant difference has come but it is prepared to use alpha corrected.\n\nIn the same figures the waveforms of NH and LD for different condition is required rather than representing individually.\n\nDiscussion should have been in the heading of a) hemispheric difference on each component of CAEPs\nLatency Amplitude\n\nb) Between groups on on each component of CAEPs\nLatency Amplitude\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "43444", "date": "03 Apr 2019", "name": "Lauren Petley", "expertise": [ "Reviewer Expertise Electrophysiology", "neuroimaging", "attention", "cognition", "hearing", "hearing loss", "brain injury", "listening difficulties." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nOverview\nIn this article, Bhat and colleagues aim to characterize the cortical processing of dichotically-presented speech tokens, as reflected by cortical auditory evoked potentials (CAEPs). Several stimulus- and subject-related factors are considered as possible contributors to the observed CAEP morphology: stimulus type (/pa/ or /ta/), spatial order, hemisphere of measurement, and age. To explore these factors, three experimental conditions were administered to adult and pediatric participants, consisting of a monaural right presentation of each speech sound, a monaural left presentation of each sound, and two dichotic configurations of the unmixed speech sounds. Another objective is to examine the differences in CAEP morphology between healthy individuals and those with learning disabilities. This goal is motivated by previous findings that dichotic listening is a useful test to examine cerebral dominance deficits, developmental changes in language lateralization, and bilateral hemispheric processing deficits in children with learning disabilities.\n\nIs the work clearly and accurately presented and does it cite the current literature?\n\nThere is a notable gap in the literature review for this paper regarding findings of the last ~ 15 years which have fundamentally changed scientific thinking around the cerebral lateralization of speech processing (e.g., Poeppel, 20031, Poeppel et al., 20122, Friederici & Alter, 20043). The rationale for the study is generally clearly stated and the experimental conditions are logically designed to address the questions that the researchers have posed. However, the authors have failed to provide a convincing argument that the study has significant novelty or importance. Is the aim simply to fill a gap in the scientific reporting of CAEP amplitudes and latencies to dichotically-presented stimuli? Is there no greater motivation, perhaps pertaining to the clinical utility of these measures as either a diagnostic, or a method to study the neurophysiological underpinnings of learning disabilities?\nThere are several easily correctable errors in this paper which confuse the reader. First, the healthy adult and pediatric participants are repeatedly referred to as “normal hearing,” but this is not a study of hearing impairment. According to the stated methodology, all participants, including those with learning disabilities, had to pass a screen for normal hearing, therefore all participants, including the disease group, must have normal hearing. Please refer to the control groups as “normal” or “controls”, not “normal hearing.” If the learning-disabled group does not have normal hearing, the methods should be updated. Second, one of the figures is incorrect: below Table 1 of P1 latencies and amplitudes, Figure 1 should depict P1 latencies, but in fact shows P2 latencies. More minor issues include: the description of the amplitude measure as “mean amplitude” in the abstract when a peak amplitude was used, the description of participant demographics in the abstract suggests that only the pediatric participants were right-handed when all participants were right-handed, frequent misspelling of the acronym CAEP as CEAP, use of the acronym LLR before it is defined, and using the word “latency” rather than “amplitude” in the final sentence of the paragraph describing N1 amplitude results. Finally, as a general comment, the paper would benefit from editing for grammatical correctness.\nThe waveform figures in this paper would all benefit from considerable revision. First, none of the figures indicate which channels are represented. This can be inferred from the text but should be stated again in the figure or figure caption. The label “Ch1” above the vertical axis is particularly confusing. The units for voltage and time are also missing. Presenting each pair of waveforms in a separate, large figure makes it impossible to visually examine the effects of age, clinical group, or experimental condition. Please combine these waveforms in a way that facilitates the observation of these effects. The captions for the waveform figures are uninformative and sometimes incorrect. For example, Figure 10 states that “waveforms clearly depict shorter latency over left hemisphere than right hemisphere” when no latency differences are apparent. If anything, the latency of P1 appears to be slightly shorter in the right hemisphere. All figures, both bar graphs and waveforms, would benefit from the addition of symbols to highlight which experimental effects reached statistical significance.\n\nIs the study design appropriate and is the work technically sound?\n\nIf the goal of this study is to support future clinical and experimental use of CAEPs to study cerebral lateralization in the processing of dichotic speech, there are two major flaws in the experimental design. First, there is no justification provided for treating the learning-disabled group as a homogeneous sample despite likely representation from a variety of subtypes (e.g., reading disabilities, attention deficit disorder, and arithmetic disabilities, among others). The importance of subtyping to improve the reproducibility of research using event-related potentials in learning-disabled children has long been acknowledged (e.g., Dool et al. 19934).\nSecond, by failing to collect behavioral dichotic listening data from these participants, the authors provide no link between clinical and experimental research using behavioral and CAEP-based measurements. This is particularly important because behavioral dichotic listening paradigms do not simply reflect cerebral lateralization in speech processing, but also the influence of selective attention (Hugdahl, 20115). The present study uses passive stimulation, except perhaps in the dichotic condition where it appears that attention deployment has not been controlled. Without knowing how the obtained CAEP measures relate to behavioral dichotic listening performance, it is impossible for clinicians or experimental scientists to weigh the relative merits of the two approaches. The purpose of highlighting these shortcomings is not to suggest that the paper is not worthy of publication, but rather to encourage the authors to appropriately address these issues in their introduction, discussion, and presumed future work that builds on these findings. Their brief mention in the paper’s conclusion does not constitute a sufficient acknowledgement of these important flaws.\nA stated objective of this experiment is to compare monaural vs. dichotic processing in the same individuals. However, the methods section suggests that a silent movie was presented under the monaural conditions and not the dichotic condition. Is this indeed the case? Like the monaural conditions, the dichotic condition is passive, so a diversionary task is still desirable to help control the deployment of attention. If a silent movie was not used in the dichotic condition, what was the rationale for excluding it?\nThere may also be a problem with accurate measurement of CAEP peaks in this study. Automated or semi-automated methods to detect and measure CAEP peaks are commonly used and these algorithms must be supplied with search window bounds that are specified by the user. In this study, the search window bounds were selected based on visual evaluation of grand average waveforms across all participants. The selected windows for the P1, N1, and P2 components were 40 – 80, 90 – 140, and 170 – 220 ms, respectively. However, search windows that are selected based on the grand average waveform, particularly if computed across different experimental groups, may not perform well on all subjects. It is common for search window bounds to be subsequently modified to ensure that they truly encompass the peaks of interest at the individual subject level. While it is unknown whether the authors completed this type of inspection, the peak latencies that are reported in Tables 2 – 4 suggest that the selected window bounds for the P1 and P2 components may not have been adequate to detect these peaks at the individual level.\nBy converting the search window bounds to z-scores on the normal distribution defined by the mean and standard deviation of the observed peak latency, one can estimate the percent of participants whose peaks might lie outside the search window. Among healthy adult participants, the upper bound of the P1 search window may have failed to capture the true peak for 7.21% and 7.49% of participants (in both cases, 1 out of 16 participants) for the monaural right and monaural left conditions, respectively, in the right hemisphere. Among healthy pediatric participants, the upper bound may have failed to capture the true P1 peak for 12.92% and 13.57% of participants (both 1 of 8 participants) for the monaural left and dichotic conditions, again both for the right hemisphere. In the monaural right condition for this group, again for the right hemisphere, the observed mean latency (77.75 ms) lies nearly upon the upper bound of the search window (80 ms) and the standard deviation (1.669 ms) is the smallest observed in the study, suggesting a possible ceiling effect imposed by the search window. Thus, the mean latency of the P1 may have been artificially lowered in healthy individuals, specifically for measurements taken over the right hemisphere. In children with learning disabilities, the upper bound of the P1 search window may have failed to capture 31.92% participants (2 of 8) in the monaural left condition, for measurements over the right hemisphere, as well as 20.33% and 29.81% of participants (1 and 2 of 8) for the left and right hemispheres, respectively, under dichotic conditions. Thus, the mean latency for the P1 may have been underestimated across both hemispheres for participants with learning disabilities under conditions of dichotic stimulus presentation.\nWhile the search window selected for the P1 appears to have ended somewhat too early, the search window for the P2 may have started too late, particularly for measurements over the left hemisphere. In healthy adults, the lower bound of the search window may have failed to capture the true component peak for 34.46%, 16.35%, and 11.15% of participants (5, 2, and 2 of 16) in the monaural left, right, and dichotic conditions, all for measurements over the left hemisphere. The lower bound of the window may also have failed to capture 14.01% (1 of 16) participants in the monaural left condition for the right hemisphere. In healthy pediatric participants, the lower bound of the search window may have failed for all measurements over the left hemisphere, for 25.46%, 25.78%, and 24.2% of participants (in all cases, 2 of 8 participants) for the monaural left, right, and dichotic conditions, respectively. Thus, the latency of the P2 in healthy individuals may have been artificially increased, particularly for measurements over the left hemisphere.\nFor children with learning disabilities, the lower bound of the search window for P2 measurement may have failed to capture 15.15% (left hemisphere) and 18.14% (right hemisphere) of participants (in both cases, 1 of 8) in the monaural right condition. Measurement may also have been compromised over both hemispheres in the dichotic condition, in which 21.48% (left hemisphere) and 14.25% (right hemisphere) of participants, both equaling roughly 1 of 8 participants, may have had peaks lying outside the search window. Thus, the mean latency of P2 may have been artificially increased over both hemispheres in the right monaural and dichotic experimental conditions for children with learning disabilities.\nOverall, these statistical results suggest that the employed search window bounds may not have adequately captured the P1 and P2 components. However, there remains some question as to whether the reported search window bounds are in fact correct, due to the presence of latencies that lie outside these bounds in the datasets that are available via Figshare. Please specify whether visual inspection was performed to ensure that component peaks were accurately identified by the peak detection algorithm and ensure that the search window bounds that are reported in the methods section are correct. Finally, please indicate how many trials contributed to the averaged waveforms that were used for peak measurement, as the influence of noise on the observed peaks varies inversely with the number of trials used for averaging.\n\nAre sufficient details of methods and analysis provided to allow replication by others?\n\nOne of the objectives of this study is to examine the utility of CAEPs to dichotic stimuli to identify altered speech processing in children with learning disabilities. What are the clinical demographics of the participants in this group? Please specify what learning disabilities the children were diagnosed with and, if possible, what clinical criteria were used to render the diagnosis. Please also specify which frequencies were tested to screen for normal hearing status.\nIt is helpful that the study stimuli are available via Figshare, but several important details of the speech stimuli are missing from the text. Please provide the durations of the speech tokens and specify how their intensities were matched to one another. It would be helpful to see the speech waveforms as a figure. Please specify the rate of stimulus presentation, how many stimulus trials were presented under each experimental condition, and how many of these trials survived data cleaning to contribute to the averaged waveforms.\nSome details of EEG data processing are erroneously described. The “Data analysis” section states that “… bad channels and bad blocks were visually inspected and interpolated …” but interpolation is a method that applies only to bad channels, not bad blocks of data. It also specifies that, “The data was then subjected to high pass filtering with a cutoff frequency of 1 kHz.” This is implausible because the data was originally acquired with a bandpass of 0 – 100 Hz. It would additionally be helpful to know how many ICA components were rejected per subject by the MARA algorithm.\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\n\nThe first hypothesis tested is that component amplitudes and latencies do not differ between the two speech tokens (/pa/ and /ba/) or between the two spatial orders of these stimuli under dichotic listening conditions. To test this hypothesis, component amplitudes and latencies were compared via a series of t-tests that is summarized in Table 1. Problematically, it is not specified anywhere which electrode was used to obtain these measurements. Furthermore, the authors were concerned with hemispheric differences in speech stimulus processing, yet this analysis does not appear to take hemispheric asymmetry into consideration. What is the reasoning for this approach?\nFollowing the statistical evaluation for stimulus effects, amplitudes and latencies were combined across stimuli. They were then tested by a 3 x 2 x 3 repeated measures ANOVA, but the factors for the ANOVA are not stated when the statistical approach is introduced. The type of test to be used for pairwise post-hoc statistics should also be specified here.\nThe authors are to be commended for reporting means and standard deviations of component amplitudes and latencies, as well as effect sizes. These types of statistics are under-reported in the CAEP literature and doing so will benefit any research groups who wish to build on the results of this study. It is unknown to what extent the statistical results observed here might be impacted by amending the P1 and P2 search windows, if necessary.\nThe N1 statistical results require clarification. The N1 latency ANOVA found a significant main effect of group which was inspected via post-hoc tests, presumably paired t-tests. No significant differences were found between the groups, but only one p-value was reported (p = 0.066). Three pairwise contrasts should have been performed: healthy adults vs. healthy children, healthy adults vs. children with learning disabilities, and healthy children vs. those with learning disabilities. Was the same p-value obtained for all pairwise contrasts? If so, this should be clarified.\nUnlike P1 and N1, there was no significant effect of group on P2 latency (p = 0.053). Significance might be reached if the P2 search window is amended. There is also a typographical error in the p value for the main effect of group on P2 amplitude.\n\nAre all the source data underlying the results available to ensure full reproducibility?\n\nSource data consisting of component latencies and amplitudes for all experimental groups across all experimental conditions are available via Figshare. However, these data tables are perplexing because many of the latencies lie beyond the reported bounds of the peak detection search windows. Were the search window bounds different than those reported?\n\nAre the conclusions drawn adequately supported by the results?\n\nAs noted previously, the discussion section of the paper needs to address the shortcomings of this study’s experimental design. The authors conclude that their results, demonstrating earlier component peak latencies over the left hemisphere than the right, are in line with known hemispheric specialization for language in healthy, normal individuals. This asymmetry was visible for both monaural and dichotic conditions of stimulus presentation and was generally observed in terms of larger component amplitudes over the left hemisphere as well. With respect to the children with learning disabilities, the authors state that their results are in line with the hypothesis that these children have reduced normal asymmetry. This conclusion does not appear to be supported by their results. No significant hemispheric latency effects were observed for the P1 or N1 in this group. Furthermore, inspection of Table 4 clearly indicates that, despite the lack of a significant interaction between hemisphere and group on P2 latency, earlier latencies in the left hemisphere were only observed in the learning disability group in the dichotic condition. Indeed, the authors later state that, “the observed lack of asymmetry could be a combination of both functional phonological decoding deficits as well as structural deficits.” The discussion should be more consistent in describing the lack of hemispheric asymmetry in CAEP latencies in this group.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/7-1939
https://f1000research.com/articles/7-216/v1	22 Feb 18	{ "type": "Software Tool Article", "title": "Breathe Easy EDA: A MATLAB toolbox for psychophysiology data management, cleaning, and analysis", "authors": [ "John C. Ksander", "Sarah M. Kark", "Christopher R. Madan", "Sarah M. Kark", "Christopher R. Madan" ], "abstract": "Electrodermal activity (EDA) recordings are widely used in experimental psychology to measure skin conductance responses (SCRs) that reflect sympathetic nervous system arousal. However, irregular respiration patterns and deep breaths can cause EDA fluctuations that are difficult to distinguish from genuine arousal-related SCRs, presenting a methodological challenge that increases the likelihood of false positives in SCR analyses. Thus, it is crucial to identify respiration-related artifacts in EDA data. Here we developed a novel and freely distributed MATLAB toolbox, Breathe Easy EDA (BEEDA). BEEDA is a flexible toolbox that facilitates EDA visual inspection, allowing users to identify and eliminate respiration artifacts. BEEDA further includes functionality for EDA data analyses (measuring tonic and phasic EDA components) and reliability analyses for artifact identification. The toolbox is suitable for any experiment recording both EDA and respiration data, and flexibly adjusts to experiment-specific parameters (e.g., trial structure and analysis parameters).", "keywords": [ "respiration artifact", "electrodermal activity", "skin conductance" ], "content": "Introduction\n\nElectrodermal activity (EDA) methods evaluate fluctuations in skin electrical conductance caused by changes in sweat gland production. The sympathetic nervous system innervates palmar and plantar eccrine sweat glands, and changes in skin conductance are thought to measure sympathetic nervous system arousal (Bach et al., 2010). Importantly, EDA recordings are a valuable and popular psychophysiological measurement in studies of affect and cognition (Boucsein et al., 2012).\n\nIt is well known that respiration and EDA influence each other (Schneider et al., 2003). In laboratory settings, researchers often leverage this relationship to check the integrity of a psychophysiology set-up. Asking participants to take a deep breath should produce concurrent deflections in both waveforms, and properly configured recording equipment should detect that response. EDA is typically recorded using electrodes placed on the palmar or plantar surfaces where eccrine sweat glands are densely located. Respiration, typically recorded using a belt secured around the diaphragm, is an oscillatory event that approximates a sine wave with regular breathing. However, irregular respiration, or abnormalities in the respiration waveform (frequency or amplitude), are associated with non-specific changes in the EDA waveform. These physiological respiration-related artifacts can lead researchers to overestimate the presence or magnitude of skin conductance responses (SCRs) in experiments (Schneider et al., 2003).\n\nDespite the strong relationship observed between EDA and respiration traces, prior work has shown that EDA and respiratory signals are not strictly coupled (Rittweger et al., 1996), which may relate to differences in their physiological origin. Physiologically, the emotion-reactive palmar and plantar eccrine sweat glands are maximally innervated by cholinergic (Sato & Sato, 1981) sudomotor fibers leaving the ventral root of the spinal cord (Boucsein, 2012, p. 20). While eccrine sweat glands are modulated by the sympathetic nervous system, the transmission related to EDA is mainly cholinergic, not noradrenalergic (Sato & Sato, 1981; Stern et al., 2001). However, deep breathing has been associated with sudden increases in free-circulating adrenaline, producing sweat responses (Boucsein, 2012, p. 32), which mimic SCRs on EDA recordings. As mentioned above, this relationship is useful for checking psychophysiological signal integrity, but can also bias SCR analyses.\n\nWhile movement-induced EDA artifacts are fairly straightforward to identify (e.g., presence of an unusually steep rise in the waveform), physiologically derived artifacts appear similar to arousal-related waveforms (Boucsein, 2012). Developing methods for identifying respiration-related artifacts has been a challenge for the field of psychophysiological research due to high intersubject and intrasubject variability in respiration activity, yielding a wide range of waveform characteristics (Schneider et al., 2003). A lack of analytical solutions has motivated software development within this field since the early 1990’s, with the goal of improving how researchers inspect and manipulate respiration data (Wilhelm & Roth, 1993).\n\nResearchers are strongly encouraged to account for such respiration-induced EDA artifacts, and subsequently outline those artifact elimination procedures in their manuscripts (Boucsein et al., 2012). This can be challenging, since common artifact-control practices involve researchers visually inspecting their respiration data, which is unfortunately both time-consuming and subjective. Schneider et al. (2003) has provided a useful decision tree for discarding artifact EDA responses based on a set of criteria. However, an easy-to-use and freely available software that expedites visual inspection of respiration data, and allows researchers to quantify their artifact-control procedures is not available. This toolbox might be particularly helpful for researchers identifying respiration artifacts in experiments with longer trial durations, such as viewing video clips or recalling autobiographical memories. In these experiments, the standard stimulus-response latency window for identifying event-related SCRs (e.g., 1–4 seconds) may no longer be suitable, and longer trials almost certainly have a higher probability of respiration-related SCR artifact contamination.\n\nCurrently, there is a need for easy-to-use, flexible, and interoperable software that facilitates EDA artifact elimination via the widely employed and accepted method of visual inspection. We have developed a novel MATLAB toolbox for efficiently eliminating EDA respiration artifacts and analyzing EDA data, which we freely distribute as Breathe Easy EDA or ‘BEEDA’. BEEDA’s streamlined artifact removal interface allows users to quickly identify and clean EDA data, expediting EDA analysis without compromising analysis integrity. Additionally, BEEDA’s integrated EDA analysis functionality allows users to seamlessly analyze cleaned EDA data within the toolbox. Furthermore, the toolbox includes inter-rater reliability (IRR) analyses so that researchers may evaluate the reliability of their artifact-control procedures.\n\nThe BEEDA toolbox is controllable through a graphical user interface (GUI), and requires no programming skill to use. This toolbox may be used either for simple artifact detection, EDA analyses, or for both artifact elimination and subsequent EDA analyses—as illustrated in Figure 1. This flexibility allows users to take advantage of BEEDA’s functionality without restricting the use of complementary software such as Mindware (MindWare Technologies Ltd., Gahanna, OH), Ledalab (Benedek & Kaernbach, 2010), ANSLAB (Wilhelm & Peyk, 2005), or AcqKnowledge (Braithwaite et al., 2013). For instance, one could use BEEDA only for marking artifacts in a dataset, and then use the artifact information file BEEDA produces with an alternative EDA analysis program. Furthermore, BEEDA is suitable for any experiment where both EDA and respiration data were collected, and parameters specific to individual experiments can easily be modified through the GUI (e.g., trial structure and analysis options). This permits a great deal of functional flexibility, without encumbering the toolbox’s usability. Here we describe the toolbox design, workflow, and functionality.\n\n\nToolbox design and workflow\n\nBEEDA’s workflow was designed to offer users situationally-specific functionality within the simplest framework possible. This allows researchers to use the toolbox for their specific goals, without the toolbox adding unnecessary work in the process. As illustrated in Figure 1, the BEEDA workflow begins with loading a dataset and setting a few critical parameters. After that initialization, the GUI main menu (Figure 2) lets researchers tailor their own workflow to their specific needs. This workflow is flexible to include any combination of data visualization, artifact inspection/cleaning, calculating EDA statistics, or performing interrater reliability (IRR) analyses. The degree of overhead imposed by the workflow (e.g., in specifying parameters or manipulating the data) at this stage should only match the requirements of the user. The following sections describe these abilities and their implementation in detail.\n\n(A) Visual experiment summary; the EDA timecourse is plotted in blue, red points mark valid SCR onsets, and vertical green lines mark recording events (e.g. trial onsets). (B) Trial-type window displays each type of recording event imported with the dataset. (C) Current settings.\n\nInitializing the BEEDA toolbox (executing BreatheEasyEDA.m) immediately launches the data loading GUI. This interface allows users to either load data files for a new session, or load data from a previously saved session. If a new session is started, BEEDA copies and reformats raw data files into a MATLAB structure variable (BEEDAdata). The BEEDAdata variable is the toolbox’s primary data structure; all user defined parameters (e.g. analysis settings) and analysis actions (e.g. artifact removal) are written to this BEEDAdata structure. Resuming a previous session reads information from a saved BEEDAdata structure and launches into the main menu.\n\nFor new sessions, basic analysis parameters are also specified in the data loading GUI. These basic settings are: downsampling and Skin Conductance Response (SCR) parameters. Importantly, once downsampling and SCR options are chosen, these settings are permanently fixed for the current BEEDA session (even if the session is saved and resumed). If a downsampling factor is specified, both the EDA and respiration data are immediately downsampled within BEEDAdata. This downsampling functionality is provided because the sampling rate capabilities of modern EDA systems (e.g. >1000 Hz) far exceed the resolution necessary for EDA analyses. Downsampling datasets to lower temporal resolutions can dramatically reduce a dataset’s size, consequently improving BEEDA’s memory and hard disk requirements, computation time, and GUI responsiveness.\n\nThe main menu provides a visual summary of your experiment, trial information, analysis settings, and display settings (Figure 2A). The main menu also allows users to save the current BEEDA session, start the artifact removal interface, run IRR analyses, and export final analysis results.\n\nBefore displaying the experiment summary panel, the EDA data is first smoothed via convolution with a Gaussian kernel (as in Benedek & Kaernbach (2010)). Smoothing removes minor signal noise, which may originate from a variety of sources (e.g. recording equipment or downsampling). Next, valid SCRs are identified based on previously specified threshold and rejection-rate parameters. The experiment summary panel plots the entire experiment’s EDA timecourse, marking onset times for trials, and valid SCRs (Figure 2A). This window provides users with an overview of the experiment’s EDA data, allowing users to easily confirm the indented dataset has loaded correctly.\n\nAll unique trial-types are displayed in the trial-type information window, and the current BEEDA session’s settings are displayed in the setting information window (Figure 2B and 2C). From the main menu, users can easily set a number of session settings: SCR latency tolerances, valid trials for analysis, and display settings (see Interface display options). SCR latency tolerances establish the stimulus time-locked window when SCRs may be appropriately attributed to the preceding stimulus (see Main EDA analysis parameters), typically a 3-second window between 1–4 seconds post-stimulus onset (Boucsein, 2012), but shorter windows have been proposed (e.g., 2 seconds or less; Barry, 1990; Levinson & Edelberg, 1985). Additionally, if end-of-trial events were omitted during an experiment’s data collection, specifying a maximum SCR latency parameter effectively creates these events. Specifying the valid trials for analysis determines which trial-types are available for artifact cleaning and EDA analysis. All unique events recorded during data collection may be declared as valid trial-types; this allows users to disregard inter-trial events, baseline events, or events not corresponding to trials of interest.\n\nThe “Display settings” main menu button (Figure 2) allows users to customize the Artifact Removal Interface. The Expanded trial window parameter controls the additional timecourse data displayed before and after each trial in the artifact removal interface. For instance, setting expanded trial window to 5 (seconds) will display the 5 seconds before every trial and the 5 seconds after every trial. This option may help users evaluate how respiration immediately preceding or following a trial relates to respiration during a trial. More specifically, we found that being presented with the activity surrounding the trial provided a useful context for identifying potential respiration artifacts.\n\nThe Number of trial windows to display parameter controls the number of trials simultaneously displayed in the artifact removal interface. This option may be particularly useful when running the BEEDA toolbox on computers with lower resolution computer monitors, as users can adjust the number of trials in each ARI page to best fit their display configuration.\n\nSelecting “Remove artifacts” from the main menu will launch the Artifact Removal Interface (ARI). The ARI allows users to efficiently clean EDA data via streamlined data presentation and easy to use controls. Users can easily scroll through ‘pages’ of trials, examining each trial for irregular respiration waves, as shown in Figure 3. If problematic respiration waves are identified, users can clean the data with either ‘SCR delete mode’ or ‘drag-delete mode’. Drag-delete mode removes entire time segments of EDA data, whereas SCR delete mode only removes SCRs from analysis consideration. Consequently, drag delete mode is recommended for Skin Conductance Level (SCL) analyses and thorough artifact elimination, whereas SCR delete mode is only recommended for SCR analyses (see EDA analysis functionality).\n\n(A) Event navigation controls. (B) Data manipulation controls and hotkey guide. (C) Respiration timecourse is plotted in blue, red points mark valid SCR onsets, and vertical green lines mark an event’s start and end.\n\nIn the ARI, user defined trials of interest are individually displayed by plotting SCR onset timepoints directly onto the trial’s respiration data (Figure 3C). This presentation simplifies the manual identification of problematic breathing (e.g. Figure 4), and the recommended procedures for EDA respiration artifact scrubbing can be found in Schneider et al. (2003). All user actions (e.g., data cleaning) are immediately applied to BEEDAdata and can be saved through the main menu.\n\nThe presentation simplifies inspecting data for a sudden deep breath (Panel A) or highly irregular breathing pattern (Panel B) preceding an SCR onset.\n\nSelecting “Export final results” in the main menu will analyze the user-defined trials-of-interest and export the analysis results to a Comma Separated Values formatted spreadsheet (.CSV file). This spreadsheet will show trial-wise EDA statistics, in addition to whether or not the trial was flagged for artifacts. A trial will show “flagged for artifacts” if any SCR or data segment was deleted from the trial. In this way, one may simply use BEEDA’s GUI to mark artifacts within an EDA dataset, then use the artifact information output with another EDA analysis software. Similarly, the artifact information output provides an easy means for assessing overall data quality. Experimenters may also directly analyze this output with BEEDA, in order to evaluate how reliably artifacts were identified within a dataset.\n\n\nArtifact inter-rater reliability\n\nTo facilitate the reporting and validity of respiration artifact rejection methods, BEEDA includes inter-rater reliability (IRR) analysis functionality for respiration artifact rejection. Selecting “Inter-rater reliability” in the main menu will perform an artifact IRR analysis directly on exported BEEDA result files. This requires that researchers have cleaned a dataset multiple times, under the same relevant parameters (verified by built-in sanity checks). After specifying these files, users can set the IRR analysis’ scope to match their analysis goals. Specifically, users can limit their analysis to only trials containing SCRs (as defined by SCR threshold parameters) or analyze all trials of interest. This is a critical distinction, as the IRR for SCR-negative trials may give unrepresentative reliability statistics for SCR oriented analyses (i.e. trials without SCRs may not have been inspected). On the other hand, these trials would certainly be considered for SCL analyses. This choice determines T in the subsequent equations.\n\nAfter setting the IRR scope, the pair-wise Cohen’s κ between all raters is calculated and exported to a CSV spreadsheet as a labeled matrix. We used Cohen’s κ implementation (Cohen, 1960) in this context:\n\nFor the set of all trials T we defined two trial classes C as: the absence of any artifact marking, or the presence of any SCR/data-segment deletion. The expected chance agreement, pe, between each pair of raters i and j was:\n\nThe user-guide documentation describes how this analysis and its output (i.e. the labeled Cohen’s κ matrix) are configured in greater detail.\n\n\nEDA analysis functionality\n\nThe BEEDA toolbox features integrated EDA analysis functionality, which may be used with or without prior artifact removal. Selecting the Export final results main menu button will initialize EDA analyses and export the subsequent results as a spreadsheet. These analyses measure tonic and phasic EDA using standard methodology (Boucsein, 2012). Tonic EDA is defined as the slow change in SCLs over a timecourse of interest. BEEDA determines the mean and standard deviation of each trial’s EDA levels, and these statistics are included in the results output. Data segments marked as artifacts using the Drag delete mode are not included in SCL analyses.\n\nPhasic EDA measurements are determined via the trough-to-peak detection of SCRs (Boucsein, 2012). SCRs are quickly changing EDA levels that exceed an amplitude threshold and occur within a response window time-locked to a stimulus. The SCR amplitude is defined as the SCR’s peak EDA level minus the SCR’s initial trough EDA level. Users can explicitly specify an SCR amplitude threshold, and this practice is typical for trough-to-peak SCR detection. Alternately, the amplitude threshold can be flexible and data driven via setting an SCR rejection rate (Kim et al., 2004). In BEEDA, specifying an explicit SCR threshold of 0μS and a rejection rate of 10% emulates the algorithmic SCR thresholding procedure described in Kim et al. (2004). While this thresholding procedure is not typically employed, BEEDA includes this functionality to mirror proprietary EDA analysis software packages which offer similar analysis options (Braithwaite et al., 2013).\n\nFor phasic EDA analyses, BEEDA detects valid SCRs and exports the following statistics for each trial: number of SCRs, average SCR magnitude, cumulative SCR magnitude, and maximum SCR magnitude. SCRs in data segments removed with Drag delete mode, in addition to SCRs marked as artifacts with SCR delete mode, are not included in SCR analyses.\n\n\nEDA analysis statistics\n\nNumber of SCRs: the number of valid SCRs in a trial\n\nAverage SCR magnitude: average trial SCR amplitude\n\nMax SCR magnitude: the largest SCR amplitude within a trial\n\nCumulative SCR magnitude: the sum of all trial SCR amplitudes\n\nSCL(average): mean EDA signal within a trial\n\nSCL(standard deviation): the standard deviation of a trial’s EDA signal\n\n\nMain EDA analysis parameters\n\nSCR threshold: Only EDA responses above this amplitude threshold are considered valid SCRs. Typically an amplitude threshold of .05μS is used, although some researchers advocate for thresholds as low as .01μS (Braithwaite et al., 2013). Schmidt & Walach (2000) recommend that sampling resolution should be taken into account when considering low thresholds, and thresholds lower than .01μS should not be used.\n\nRejection rate: If a rejection rate greater than 0 is specified, trial-wise thresholding is applied according to: Rth = max⁡(R) α where Rth is the trial-specific response threshold, R is the trial’s set of responses and α is the rejection rate. For example, if the rejection rate is 10% and a trial’s largest SCR amplitude is 4μS, SCRs with amplitudes below .4 μS are rejected in that trial.\n\nMin SCR latency: The minimum time after a trial’s start when EDA data can be considered for analyses (i.e. the stimulus response window). Valid SCRs onsets must begin after the specified minimum latency time, and EDA levels before minimum latency time will be excluded from SCL analyses. Benedek & Kaernbach (2010) report that a minimum latency of 1 second post-stimulus is typical.\n\nMax SCR latency: The time after a trial’s start when EDA data cannot be considered for analyses. Valid SCRs must begin before the specified maximum latency, and EDA signal after the maximum latency is excluded from SCL analyses. Benedek & Kaernbach (2010) report that a maximum latency of 3 or 5 seconds post-stimulus is typical.\n\n\nOperation\n\nBEEDA’s system requirements are: Matlab R2014b or newer, and the Matlab Signal Processing Toolbox. Any computer with that prerequisite Matlab software can run BEEDA (e.g. regardless of operating system). However, users are recommended to run BEEDA with Matlab R2015a, since the toolbox was developed and extensively tested with R2015a.\n\nBEEDA was designed for input datasets containing both EDA and respiration recordings. However, suitable input files may also contain placeholder values for either data channel (i.e. for datasets without either respiration or EDA recordings). The toolbox was designed to accept raw data files from Biopac (Biopac Systems Inc., USA) recording systems. The BEEDA user-guide describes how these files are obtained from Biopac systems, and how these files are formatted. Although BEEDA was designed to easily accept files from these widely-used systems, any comparably formatted files are also suitable (i.e. from other recording systems). The acceptable formatting is very basic, and therefore recordings from other systems should not present major issues.\n\n\nUse case\n\nWe have provided a sample dataset1. This data was collected during an emotional-image viewing experiment, and is provided for toolbox demonstration purposes. Documentation for this sample dataset is included with the distribution, and provides further background about the experiment and the data’s structure. We have also provided example analysis output using this dataset as Supplementary File 1. This output shows artifact information from data cleaning, along with the analyses described in the sections on EDA analysis functionality and statistics. The output file is formatted as a .CSV spreadsheet, with easily interpretable column headers.\n\n\nConclusion\n\nBreathe Easy EDA is a novel MATLAB toolbox developed for easy and reliable identification of respiration-related artifacts in EDA data. This software was specifically built to facilitate the methodical considerations of psychophysiology researchers through a simple, flexible, interoperable, and tolerant design. BEEDA’s simplified data presentation allows efficient data inspection and cleaning, without sacrificing functionality in the GUI. In fact, the intuitive interface includes features that are absent from widely used contemporary EDA software, but still essential to researchers (e.g., an “undo” function). The artifact cleaning functionality extends to integrated reliability analyses, providing a simplified means for researchers to establish the consistency of their artifact-control procedures across independent raters. BEEDA’s common output-file format and range of analysis capabilities also allows users to integrate this toolbox in their analysis pipelines without precluding alternate software packages. Furthermore, BEEDA was built to flexibility handle any experiment where both respiration and EDA data were collected, regardless of trial duration or experimental design. In these ways, this software provides researchers with optimized tools for psychophysiology analysis. The toolbox is freely available from http://github.com/johnksander/BreatheEasyEDA, and the user-guide documentation for BEEDA is included with this distribution.\n\n\nSoftware availability\n\nSoftware/source code available from: http://github.com/johnksander/BreatheEasyEDA\n\nArchived source code as at time of publication: https://doi.org/10.5281/zenodo.1168739 (Ksander, 2018).\n\nLicense: GNU General Public\n\n\nNotes\n\n1Figure 2–Figure 4 were produced with this data.", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by NIH grant R01MH080833 and a Searle scholars grant, both awarded to Elizabeth A. Kensinger. CRM was supported by a fellowship from the Canadian Institutes of Health Research (FRN-146793), SK was supported by a National Science Foundation Graduate Research Fellowship Program (DGE1258923) award and NIH-NIMH Ruth L. Kirschstein National Research Service Award (F31MH113304), and JCK was supported by a NIH/NIGMS predoctoral training grant (T32-GM084907). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe would like to thank Elizabeth A. Kensinger and Eric S. Allard for their valuable feedback and contributions to the development of this project.\n\n\nSupplementary material\n\nSupplementary File 1: Example output file.\n\nClick here to access the data.\n\n\nReferences\n\nBach DR, Friston KJ, Dolan RJ: Analytic measures for quantification of arousal from spontaneous skin conductance fluctuations. Int J Psychophysiology. 2010; 76(1): 52–55. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBarry RJ: Scoring criteria for response latency and habituation in electrodermal research: a study in the context of the orienting response. Psychophysiology. 1990; 27(1): 94–100. PubMed Abstract \| Publisher Full Text\n\nBenedek M, Kaernbach C: A continuous measure of phasic electrodermal activity. J Neurosci Methods. 2010; 190(1): 80–91. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBoucsein W: Electrodermal activity. Springer Science & Business Media. 2012. Publisher Full Text\n\nBoucsein W, Fowles DC, Grimnes S, et al.: Publication recommendations for electrodermal measurements. Psychophysiology. 2012; 49(8): 1017–1034. PubMed Abstract \| Publisher Full Text\n\nBraithwaite JJ, Watson DG, Jones R, et al.: A guide for analysing electrodermal activity (EDA) & skin conductance responses (SCRs) for psychological experiments. Psychophysiology. 2013; 49: 1017–1034. Reference Source\n\nCohen J: A coefficient of agreement for nominal scales. Educ Psychol Meas. 1960; 20(1): 37–46. Publisher Full Text\n\nKsander JC: BreatheEasyEDA (Version v1.0.0). Zenodo. 2018. Data Source\n\nKim KH, Bang SW, Kim SR: Emotion recognition system using short-term monitoring of physiological signals. Med Biol Eng Comput. 2004; 42(3): 419–427. PubMed Abstract \| Publisher Full Text\n\nLevinson DF, Edelberg R: Scoring criteria for response latency and habituation in electrodermal research: a critique. Psychophysiology. 1985; 22(4): 417–426. PubMed Abstract \| Publisher Full Text\n\nRittweger J, Lambertz M, Langhorst P: Electrodermal activity reveals respiratory and slower rhythms of the autonomic nervous system. Clin Physiol. 1996; 16(3): 323–326. PubMed Abstract \| Publisher Full Text\n\nSato K, Sato F: Pharmacologic responsiveness of isolated single eccrine sweat glands. Am J Physiol. 1981; 240(1): R44–51. PubMed Abstract \| Publisher Full Text\n\nSchmidt S, Walach H: Electrodermal activity (EDA)-State-of-the-art measurement and techniques for parapsychological purposes. J Parapsychol. 2000; 64(2): 139–164. Reference Source\n\nSchneider R, Schmidt S, Binder M, et al.: Respiration-related artifacts in EDA recordings: introducing a standardized method to overcome multiple interpretations. Psychol Rep. 2003; 93(3 Pt 1): 907–920. PubMed Abstract \| Publisher Full Text\n\nStern R, Ray W, Quigley K: Psychophysiological Recording. (2nd ed.). New York, NY USA: Oxford University Press. 2001. Reference Source\n\nWilhelm F, Peyk P: ANSLAB: Autonomic Nervous System Laboratory (Version 4.0). Available at the SPR Software Repository: http://www.sprweb.org, 2005.\n\nWilhelm F, Roth W: Exam 2.0: a program to visualize, edit and analyze large vectors of data, with application to biomedical engineering. Paper presented at the First Matlab Conference, Boston, USA. 1993." }	[ { "id": "32730", "date": "05 Apr 2018", "name": "Edwin S. Dalmaijer", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors present a Matlab toolbox to assess skin-conductance measurements.Their toolbox sets itself apart from existing analysis options in particular by allowing users to filter out artifacts related to movement and (partially) unrelated physiological signals that arise through e.g. breathing. The authors outline what task-related signal researchers tend to be interested in, and how artifacts can distort these. The authors also highlight a current lack of user-friendly analysis solutions, and therefore present a toolbox that should offer Matlab users a platform to do skin conductance analyses.\nI particularly enjoyed the brief and to-the-point introduction, which manages to clearly introduce the need for the presented software without unnecessary excursions. In addition, I appreciate that the authors took the time to produce this software package and to open it up to fellow researchers. The authors would benefit from a citable publication on their efforts, as unfortunately scientific software is currently under-appreciated in reward frameworks for scientists (i.e. it's hard to cite, and direct citations to software tend to be disregarded).\nThe article is well-written, but summarises the algorithms included in the toolbox rather coarsely. I think the manuscript would benefit from a higher level of detail on the statistical analysis. Specifically, I think the authors should include equations that describe how parameters are computed from the signal.\nAn additional feature that would benefit the software is the option to do statistical comparisons between experimental conditions, which currently does not seem to be possible. Artifact identification and correction is important, but also just the first step of an analysis. The target audience for this toolbox includes, according to the manuscript, people with no programming experience. These will profit from additional analysis options that could be built into the GUI.\nThe code is made publicly available via GitHub, which is a very suitable platform. The chosen license (GNU GPL v3) allows for any kind of re-use. Unlike some open-source software, the documentation that the authors provide with their toolbox is extensive, and should (in my opinion) suffice to help even the most inexperienced of users.\nDISCLAIMER: This reviewer is not an expert in skin conductance signals, but has published open software, including toolboxes for Python and Matlab. This means I am not in a position to sufficiently review the actual software, and will thus limit my review to the manuscript.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Partly\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [] }, { "id": "33784", "date": "11 May 2018", "name": "Jens Foell", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nReview of manuscript “Breathe Easy EDA: A MATLAB toolbox for psychophysiology data management, cleaning, and analysis”\nThe presented manuscript describes a newly developed Matlab toolbox named BEEDA, which was mainly designed to improve the way that researchers deal with artifacts due to respiration when working with EDA data. This is relevant as EDA is a widely used physiological measure, often in conjunction with other methods such as neuroimaging or pain stimuli, and because respiration has been shown to alter EDA-derived data in a way that can lead to false interpretations.\nThe authors describe the relationship between respiratory artifacts and genuine EDA signal, including the fact the two have a non-linear interrelationship and that due to their nature, respiratory artifacts are harder to identify than, for example, artifacts caused by movement. They then describe the procedure and workflow of the toolbox that they have developed to address respiration-related issues with the analysis of EDA signals. Examples are given for how the toolbox will display respiration that is unusual in its frequency or intensity (irregular breathing and sudden deep breath, respectively) and how the system allows to judge the temporal instance of these events in comparison to EDA responses of interest. The user can then run trial-by-trial analysis and flag or delete suspicious events.\nThe described toolbox was clearly designed to integrate all necessary parts of an EDA analysis into one package: apart from artifact identification and correction, it includes all standard EDA analysis procedures, so that the normal user should be able to rely on this toolbox alone. Further, it includes algorithms to calculate inter-rater reliability.\nWhile these additions are not strictly required for artifact cleaning software (as there are other software packages available that can be used to perform these tasks), I would expect them to increase convenience for the user, as this integrated system reduces the clutter caused by combining different toolboxes to work on the same data set. Further, the fact that the software package is free and includes a graphical user interface will facilitate its use by researchers within the scientific community.\nIn summary, the described toolbox provides a convenient way to run a data quality assessment and subsequent data analysis for EDA studies, with a special focus on correcting for respiratory effects. The manuscript is written clearly and succinctly and provides all necessary information for a user to understand and use the toolbox. After a thorough review of the text, I have no recommended edits.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-216
https://f1000research.com/articles/7-610/v1	18 May 18	{ "type": "Method Article", "title": "Estimating the contribution of studies in network meta-analysis: paths, flows and streams", "authors": [ "Theodoros Papakonstantinou", "Adriani Nikolakopoulou", "Gerta Rücker", "Anna Chaimani", "Guido Schwarzer", "Matthias Egger", "Georgia Salanti", "Theodoros Papakonstantinou", "Gerta Rücker", "Anna Chaimani", "Guido Schwarzer", "Matthias Egger", "Georgia Salanti" ], "abstract": "In network meta-analysis, it is important to assess the influence of the limitations or other characteristics of individual studies on the estimates obtained from the network. The percentage contribution matrix, which shows how much each direct treatment effect contributes to each treatment effect estimate from network meta-analysis, is crucial in this context. We use ideas from graph theory to derive the percentage that is contributed by each direct treatment effect. We start with the ‘projection’ matrix in a two-step network meta-analysis model, called the H matrix, which is analogous to the hat matrix in a linear regression model. We develop a method to translate H entries to percentage contributions based on the observation that the rows of H can be interpreted as flow networks, where a stream is defined as the composition of a path and its associated flow. We present an algorithm that identifies the flow of evidence in each path and decomposes it into direct comparisons. To illustrate the methodology, we use two published networks of interventions. The first compares no treatment, quinolone antibiotics, non-quinolone antibiotics and antiseptics for underlying eardrum perforations and the second compares 14 antimanic drugs. We believe that this approach is a useful and novel addition to network meta-analysis methodology, which allows the consistent derivation of the percentage contributions of direct evidence from individual studies to network treatment effects.", "keywords": [ "indirect evidence", "percentage contributions", "projection matrix", "flow networks" ], "content": "Introduction\n\nDecision making around multiple alternative healthcare interventions is increasingly based on meta-analyses of a network of relevant studies, which contribute direct and indirect evidence to different treatment comparisons1,2. Limitations in the design and flaws in the conduct of studies synthesized in network meta-analysis (NMA) reduce the confidence in the results: a treatment comparison in the network may be directly or indirectly informed by studies at high risk of bias. A relative treatment effect from NMA (hereafter the NMA effect estimate) is estimated as a linear combination of the available direct estimates of the treatment effect (i.e. the results from pairwise meta-analyses) and the indirect evidence on the treatment effect.\n\nSalanti et al. suggested that in order to assess the impact of study deficiencies on an NMA effect estimate, the limitations of studies contributing to direct estimates should be considered jointly, taking into account their relative contribution to the overall NMA effect estimate3. The percentage contribution matrix plays a key role in this approach: a matrix that shows how much each direct effect contributes to the estimation of the NMA effect.\n\nThe percentage contribution matrix is derived from the absolute contribution matrix. The absolute contributions of direct effects to an NMA effect is the projection matrix from a two-step NMA model4,5. In the first stage, all direct effects are derived from pairwise meta-analyses. In the second stage, the NMA effect estimates are produced as a linear combination of the derived direct effects. The respective projection matrix is called the H matrix and it is analogous to the hat matrix in a linear regression model. The elements in the H matrix can be viewed as generalized weights from pairwise meta-analysis, but they do not add up to 1 and depend on the precision of the available studies, the degree of between-study heterogeneity and the network structure.\n\nTo translate the entries of the H matrix into percentage contributions, Salanti et al. suggested normalizing the absolute entries of each row of H and interpret them as percentages3. However, H represents the flow of evidence in different paths; the weight of each path is assigned to each direct effect involved. Thus, ignoring the multiple occurrence of the same values by taking standardized absolute values is incorrect. In particular, such a process overestimates the contribution of comparisons involved in long paths and underestimates the weights of the shortest paths. In this paper, we address this issue and present a method that properly translates the entries of the H matrix into percentages. The methodology is based on the observation that the rows of the H matrix can be interpreted as flow networks4,6.\n\nTo illustrate the ideas presented in this paper, we will use a network of topical antibiotics for the treatment of chronic otitis media with ear discharge in patients with eardrum perforations7. This network was used in 3 and compares no treatment (x), quinolone antibiotic treatment (y), non-quinolone antibiotic treatment (u) and antiseptic treatment (v)7. The study outcome was the proportion of patients with persistent discharge from the ear after 1 week, measured using the odds ratio (OR). The network plot shown in Figure 1a shows that direct evidence exists for all comparisons except u versus x (non-quinolone antibiotic versus no treatment).\n\nx, no treatment; y, quinolone antibiotic; u, non-quinolone antibiotic; v, antiseptic.\n\nIn order to assess the confidence that should be placed in an NMA effect estimate, Salanti et al. suggested considering the quality of all pieces of evidence that contributed to it. For example, the studies directly comparing ‘u versus v’ were judged to be at high risk of bias; however, in order to judge the quality of the NMA effect estimate of ‘u versus v’, we need to take into account the amount of data that these studies contributed to its estimation.\n\n\nMethods\n\nWe first present the random-effects two-stage NMA model first described by Lu et al.5. We will employ a simplified version of the H matrix described by König et al.4 that does not take into account the correlation induced by multi-arm trials. We ignore this correlation for the sake of ease of interpretation of the entries in the H matrix; we discuss implications of multi-arm trials in the Discussion section. Taking advantage of previous findings on how the flow of evidence can be considered in NMA4,6, we present an algorithm to decompose the flow in a network and subsequently approximate the percentage contributions of direct effect estimates for each NMA effect estimate.\n\nConsider a network of T competing treatments. The set of treatments is denoted by V = {x, y, u, v, ...} and let x denote the reference treatment. The number of NMA effects to be estimated is (T2) but the estimation of T – 1 effects allows the derivation of the remaining effects via linear combination. We collect the T – 1 effects against the reference treatment x in a vector of basic parameters θ = (θxy,θxu,θxv, ...)′. In the case of a dichotomous outcome, θ is the parameter vector of all log-ORs compared to the common reference treatment x.\n\nWe assume that the distribution of effect modifiers is similar across comparisons and thus the transitivity assumption is plausible. The consistency assumption refers to the statistical manifestation of transitivity and implies that all sources of evidence are in agreement; this is expressed via the consistency equations\n\nθuv = θxv – θxu, for all u, v ∈ V\n\nLet us denote the number of comparisons with direct data (that is, at least one direct study) with D. For simplification, consider that there are no multi-arm studies. At the first stage of the NMA model, direct effects are estimated, using random-effects pairwise meta-analyses. The estimates of the direct effects are collected in a column vector θ^D of length D; their estimated variances are collected in a diagonal D × D matrix VD. At the second stage, the NMA effects are estimated as\n\n\n\nwhere H is\n\nH = Y(X′(VD)–1X)–1X′(VD)–1 Equation 2\n\nMatrix X is a D × (T – 1) design matrix expressing the linear relationships between the available direct effects and the basic parameters and Y is a (T2) × (T – 1) design matrix that links the NMA estimates with the basic parameters. Note that X is identical to Y only when there are direct studies for all treatment comparisons in the network.\n\nMatrix H is of dimensions (T2) × D and describes the influence of each direct effect (specified in the column) to an NMA effect (specified in the row). As Equation 2 implies, H is derived as a function of the variances of the direct effects θ^D and the network structure; therefore, the exact (absolute) contribution of each direct comparison depends on the precision of the available direct data and the comparison’s connectivity to the rest of the network. Note that it resembles the hat matrix in a linear regression model.\n\nLet us focus on a single row of the H matrix which, say, corresponds to the NMA effect estimate of ‘x versus y’ and is denoted by hxy. Elements of hxy are denoted by huvxy and show the absolute contribution of the direct effect θ^uvD indicated in the subscript (‘u versus v’) to the ‘x versus y’ NMA effect θ^xyN. Consider our motivating example which examines the set of treatments V = {x, y, u, v}. Equation 1 implies that the NMA treatment effect for the ‘x versus y’ comparison is derived as a linear combination of the direct meta-analyses\n\n\n\nThe element hxyxy represents the absolute but also the percentage contribution pxyxy of the direct evidence for the particular NMA effect. Assuming that the comparison ‘x versus y’ is not part of any multi-arm study, the evidence to derive the NMA effect estimate can be portioned into direct and indirect estimates\n\n\n\nwith θ^xyI denoting the indirect effect for the ‘x versus y’ comparison; hence pxyxy=hxyxy. While the percentage contribution of each direct effect to its NMA effect can be obtained as the diagonal of the H matrix, the percentage contributions of other direct relative effects via indirect evidence (e.g. the puvxy percentage contribution of θ^uvD to θ^xyN) cannot be easily derived from the absolute contributions (that is, from huvxy). In the next section we will present how the absolute contributions huvxy could be translated to percentage contributions puvxy.\n\nTo explain the method, we will continue focusing on one row of the matrix, say hxy, corresponding to the ‘x versus y’ comparison. Box 1 includes the definitions, along with the notation, of some of the notions used in this paper.\n\nComparison graph\n\nA comparison graph is defined as a graph Gxy = (V, E, F) constructed from a row of the H matrix, hxy; its definition derives from a set of vertices V, a set of edges E, and a set of flows, F.\n\nSet of vertices\n\nThe set of vertices is defined as the set of treatments examined in the network, V = {x, y, u, v ...}.\n\nSet of flows\n\nThe set of flows is defined as F = {fuv, ∀ uv ∈ E} where fuv is equal to \|huv\|.\n\nSet of directed edges\n\nThe set of directed edges E is defined as the set of direct comparisons respecting the signs of the entries of hxy. Edges are given a direction upon the definition of flows; the network itself corresponds to an undirected graph.\n\nSource\n\nSource is defined as a vertex with no incoming edges.\n\nSink\n\nSource is defined as a vertex with no outgoing edges.\n\nPath\n\nA path πi is defined as a sequence of connected directed edges belonging to E.\n\nStream\n\nA stream Si is defined as the composition of a path and its associated flow, Si = (φi, πi). with i = 1, ... , I where I is the total number of streams.\n\nKönig et al. showed that every row of the H matrix, hxy, can be interpreted as a flow network with source x and sink y, and visualised in a directed acyclic graph (DAG)4. Thus, we create a graph Gxy = (V, E, F) from hxy; its definition derives from a set of vertices V, a set of edges E, and a set of flows, F. The set of vertices is defined as the set of treatments examined in the network, V = {x, y, u, v ...}. Set E is defined as a set of directed edges E that correspond to observed direct comparisons respecting the signs of the entries of hxy. To simplify the notation, we drop from now on all superscripts assuming they all refer to xy. Then, the set E contains uv if huv > 0 or contains vu if huv < 0. The set of flows is defined as F = {fuv, ∀ uv ∈ E} where fuv is equal to \|huv\|.\n\nThe following conditions hold for the elements of set F (see Supplementary File 1 for proof):\n\na. The sum of outflows of node x (source) is 1\n\n\n\nb. The sum of inflows of node y (sink) is 1\n\n\n\nc. The flow passing through each internal node (any node except x or y) is conserved\n\n\n\nd. Gxy is acyclic; there is no path (sequence of edges) that visits the same vertex twice.\n\nConsider, for example, the graph Gxy in Figure 1b, which corresponds to the xy comparison of the network of four treatments of Figure 1a. The set of vertices is V = {x, y, u, v} and the set of directed edges is E = {xy, xv, vy, vu, uy}. Flows fuv are given along the edges; their numerical values are equal to the respective absolute entries of hxy and the direction of their corresponding edge is indicated in the subscript. As properties (a) to (d) imply, the arrows in Figure 1b indicate that the outflows of x, as well as the inflows of y, equal 1, and that the inflows equal the outflows in the intermediate nodes u and v.\n\nIn Figure 1b there are three different paths from x to y, one based on direct evidence, {xy}, and two based on indirect evidence, {xv, vy} and {xv, vu, uy}. A path is a sequence of connected directed edges belonging to E, and we denote it as πi. As property (d) implies, each node occurs at most once in πi. Then, given the above properties of fuv, we can assign a flow φi to each path πi. Flow φi is equal to the smallest fuv in the path πi. Figure 1c shows the three paths from x to y; π1, π2 and π3, and their corresponding flows. Path π1 corresponds to xy and its flow, φ1, equals the flow of the single edge in path, fxy = 0.635. Path π2 is constituted from two edges, xv and vy; thus, flow φ2 = min(fxv, fvy) = 0.251. The flow corresponding to the third path π3 is φ3 = min(fxv, fvu, fuy) = 0.114.\n\nWe define a stream, Si, as the composition of a path and its associated flow, Si = (φi, πi) with i = 1, ..., I where I is the total number of streams; here I = 3. Note that it holds ∑i=1Iφi=1.\n\nIn order to assign percentage contributions to each direct comparison, we need to split each stream’s flow to the involved edges in the stream’s path. It can be shown that an NMA effect is a linear combination of the direct effects combined within paths. More specifically, Equation 3 can be re-written as\n\n\n\nwith ∑i=13φi=1.\n\nTo approximate the percentage contributions per comparison, we suggest dividing φi by the length of the respective path πi, #πi. This will leave the percentage contribution of the direct evidence of the same treatment comparison equal to the diagonal of the H matrix and assign to each comparison involved in an indirect route a portion of the respective stream’s flow. Note that directed edges might be involved in more than one path; we thus define the percentage contribution of an edge uv as\n\n\n\nFigure 1d shows the derivation of the percentage contributions of each direct comparison in the network of topical antibiotics. Hence, from the row of the H matrix (0.635, 0.365, –0.114, –0.251, –0.114), which shows the absolute contributions of the direct effects θ^xyD,θ^xvD,θ^yvD,θ^yuD,θ^uvD to θ^xyN, we approximated their percentage contribution as 63.5%, 16.4%, 3.8%, 12.6% and 3.8%, respectively.\n\nIn this section, we present an iterative algorithm that generalizes the process outlined above to derive percentage contributions of each direct effect to the estimation of a ‘x versus y’ NMA effect. We start by defining a graph Gxy from hxy.\n\nThe algorithm is described as follows:\n\n0. Set initial graph G0 = (V, E0, F0) = Gxy. E0 contains uv if huv > 0 or contains vu if huv < 0. The set of flows is F0 = {f0,uv \| uv ∈ E0}; numerical values of f0,uv are equal to huv.\n\nThen, repeat the process below I times, equal to the number of streams in Gi, until Ei = {Ø}.\n\n1. In Gi–1, find the shortest path from x to y, πi, and define its flow as φi = min{fi–1,uv, uv ∈ πi}. Then, use πi and φi to define the stream Si = (φi, πi) .\n\n2. Recalculate the flow of edges uv ∈ πi by subtracting φi from the flow of the edges of the stream found: fi,uv = fi–1,uv – φi ∀ uv ∈ πi. The flow of the rest of the edges that do not belong to πi remain unchanged: fi,uv = fi–1,uv ∀ uv ∉ πi.\n\n3. Define Ei as the set of edges uv for which fi,uv > 0; this is Ei–1 after removing the edges with zero flow, Ei = Ei–1\\{uv \| fi,uv = 0}. Collect fi,uv to form the set Fi = {fi,uv \| uv ∈ Ei}.\n\n4. If Ei ≠ {Ø} define Gi = (V, Ei, Fi) and go to step 1.\n\nWhen the algorithm terminates, all streams Si = (φi, πi) have been identified and Equation 6 is used to derive the percentage contributions puv.\n\nRepeating the same process for all NMA effects, we derive all puvxy and collect them in a matrix P of the same dimensions as H. The presented algorithm could be described as a reverse maximum flow Edmonds Karp algorithm8, but instead of adding we remove augmenting paths.\n\nIt is possible that multiple shortest paths exist; in this case, the order in which one chooses such a path could in principle result in different percentage contributions per comparison. We can, thus, use the following modification in the algorithm to impose consistency. Instead of selecting the shortest path, we assign cost values ci,uv to each edge uv as follows: ci,uv = 2 – fi–1,uv. Then, we select the path from x to y with the minimum cost across comparisons included in πi. The definition of the cost values ci,uv assures that paths are selected from shortest to longest and removes any ambiguity regarding the selection of paths.\n\nCalculations in this paper were performed using R9.\n\n\nApplication\n\nWe apply the algorithm described above to the network of topical antibiotics7.\n\nDirect effects are obtained using the random effects model and the H matrix of dimension 6×5 is calculated using Equation 2. The H matrix, along with NMA effects, is given in Table 1.\n\nColumns correspond to direct comparisons and rows correspond to network meta-analysis (NMA) effects. Direct effects along with their variances and NMA effects with 95% confidence intervals (CIs) are given in the last column. Direct and NMA effects are measured as log odds ratios. Positive values favour the first treatment.\n\nx, no treatment; y, quinolone antibiotic; u, non-quinolone antibiotic; v, antiseptic.\n\nWe begin by applying step 0 of the algorithm. We construct the network G0 = Gxy = (V, E0, F0) with source x and sink y corresponding to row hxy. The set of vertices is V = {x, y, u, v} and the set of directed edges, taking into account the signs of the elements of hxy, is E0 = {xy, xv, vy, vu, uy}. The set of flows is F0 = {f0,uv \| uv ∈ E0}, where f0,uv equal the respective absolute values of Table 1 and are given along the edges of Figure 1b.\n\nThen, we apply the developed iterative algorithm until Ei = {Ø}. The iterations of the algorithm equal the number of existing streams from x to y and are illustrated in Figure 2.\n\nTreatment labels: x, no treatment; y, quinolone antibiotic; u, non-quinolone antibiotic; v, antiseptic.\n\n1. In G0, find the shortest path from x to y, π1 = {xy}. Define its flow as φ1 = min{f0,uv, uv ∈ π1} = f0,xy = 0.635. Define stream S1 = (φ1, π1) (Figure 2a).\n\n2. Recalculate the flow of edge xy ∈ π1 as f1,xy = f0,xy – φ1 = 0.635 – 0.635 = 0. The flow of the rest of the comparisons remains unchanged: f1,uv = f0,uv ∀ uv ∉ π1 (Figure 2b).\n\n3. Define E1 as the set of edges uv for which f1,uv > 0; edge xy is removed since its flow is zero, E1 = E0\\{xy}. Collect f1,uv to form the set F1 = {f1,uv \| uv ∈ E1} (Figure 2c).\n\n4. Define G1 = (V, E1, F1). As E1 = {xv, vy, vu, uy} ≠ {Ø}, go to step 1 (Figure 2c).\n\n1. In G1, find the shortest path from x to y, π2 = {xv,vy}. Define its flow as φ2 = min{f1,uv, uv ∈ π2} = f1,vy = 0.251. Define stream S2 = (φ2, π2) (Figure 2d).\n\n2. Recalculate the flow of edges xv and vy as f2,xv = f1,xv – φ2 = 0.365 – 0.251 = 0.114 and f2,vy = f1,vy – φ2 = 0.251 – 0.251 = 0. The flow of the rest of the comparisons remains unchanged: f2,uv = f1,uv ∀ uv ∉ π2 (Figure 2e).\n\n3. Define E2 as the set of edges uv for which f2,uv > 0; edge vy is removed since its flow is zero and thus E2 = E1\\{vy}. Collect f2,uv to form the set F2 = {f2,uv \| uv ∈ E2} (Figure 2f).\n\n4. Define G2 = (V, E2, F2). As E2 = {xv, vu, uy} ≠ {Ø}, go to step 1 (Figure 2f).\n\n1. In G2, find the shortest path from x to y, π3 = {xv, vu, uy}. Define its flow as φ3 = min{f2,uv, uv ∈ π3} = 0.114. Define stream S3 = (φ3, π3) (Figure 2g).\n\n2. Recalculate the flow of edges xv, vu and uy as f3,xv = f3,vu = f3,uy = f2,xv – φ3 = 0.114 – 0.114 = 0 (Figure 2h).\n\n3. Define E3 as the set of edges uv for which f3,uv > 0; edges xv, vu and uy are removed since their flow is zero, E3 = E2\\{xv, vu, uy}. Collect f3,uv to form the set F3 = {f3,uv \| uv ∈ E3} (Figure 2i).\n\n4. Define G3 = (V, E3, F3). The set of direct edges is E3 = {Ø} and the algorithm is terminated at this point (Figure 2i).\n\nFigure 1c shows the flows of the three streams identified when applying the above algorithm. We then calculate the percentage contributions of each comparison to the ‘x versus y’ NMA treatment effect estimate using Equation 6 (Figure 1d). For instance, to calculate pxv we first have to identify the relevant paths; these were π2 and π3. Consequently,\n\n\n\nThe calculations for deriving the percentage contributions of the other comparisons are shown in Table 2. Applying the algorithm to all NMA treatment effect estimates we get the entire percentage contribution matrix P (Table 3).\n\nx, no treatment; y, quinolone antibiotic; u, non-quinolone antibiotic; v, antiseptic.\n\nCells show the percentage contribution of direct comparisons indicated in the column to the network meta-analysis treatment effects indicated in the rows.\n\nx, no treatment; y, quinolone antibiotic; u, non-quinolone antibiotic; v, antiseptic.\n\nThe algorithm translating the H matrix into percentage contributions of direct effects can be applied to quantify the influence that a comparison-level characteristic has in the estimation of the NMA effects. For instance, if risk of bias judgements for direct meta-analyses are available, we can obtain an approximation of the percentage of each NMA treatment effect estimate that is coming from direct comparisons with a ‘high’, ‘moderate’ or ‘low’ risk of bias. Salanti et al. suggested the visualisation of this information using a bar plot, in which direct comparisons of the same risk of bias level have been grouped3. Figure 3 shows such a bar plot using the algorithm described in this paper; inspecting Figure 3 can support judgements regarding the importance of study limitations for different NMA treatment effect estimates. For instance, direct comparisons with high risk of bias contribute more than 50% in the estimation of the ‘u versus v’ comparison, potentially reducing the confidence that we can place in this particular NMA treatment effect estimate.\n\nRisk of bias per direct comparison has been assumed to be the majority of per trial risk of bias. The bar plot has been produced in CINeMA (Confidence In Network Meta-Analysis) software11. Studies are synthesized using the random effects model. x, no treatment; y, quinolone antibiotic; u, non-quinolone antibiotic; v, antiseptic.\n\nSo far we have illustrated how to derive percentage contributions for a network with four treatments. However, the algorithm can be straightforwardly applied to large networks of any structure, as soon as the involved treatments are connected. Consider for example a large network examining antimanic drugs (Figure 4)10. Let us concentrate on the comparison PLA versus OLA (‘placebo versus olanzapine’); the algorithm starts by applying step 0 and constructing network G0. Then, we continue by finding the shortest path in the first iteration, which corresponds to the direct comparison, and define its flow and stream S1 = (φ1, π1). The number of algorithm’s iterations is equal to the number of streams from placebo to olanzapine, which turns out to be 16. The resulting entire percentage matrix is given in Supplementary File 2.\n\nASE, asenapine; ARI, aripiprazole; PLA, placebo; HAL, haloperidol; QUE, quetiapine; LITH, lithium; ZIP, ziprasidone; OLA, olanzapine; DIV, divalproex; RIS, risperidone; CARB, carbamazepine; LAM, lamotrigine; PAL, paliperidone; TOP, topiramate; ASE, asenapine.\n\n\nDiscussion\n\nIn this paper, we present a new approach to derive percentage contributions of the direct comparisons to the treatment effect estimates in NMA. We made use of the fact that the composition of network treatment effect estimates can be interpreted as a flow of evidence. The calculation of the percentages uses the estimated variances of direct effects and thus incorporate associated uncertainty in their estimation. Applying the algorithm to networks of interventions can be used to quantify the contribution of potential study limitations to the NMA treatment effect estimates.\n\nAn assumption that underlies our algorithm is the equal split of the stream flow to the involved comparisons. Although indirect effects are not weighted averages, we find this approximation to be a pragmatic approach that reasonably reflects the amount that each comparison contributes to network effects. A limitation of using the H matrix described in this paper lies in the ignorance of the correlation induced by multi-arm trials. Alternatively, one can incorporate multi-arm trials by reducing their weights accordingly, as proposed by Rücker et al.12.\n\nAlternative methods to derive the relative contribution of all sources of evidence have been developed4,9,11,13. An alternative approach has been proposed to derive percentage study weights in a variety of meta-analysis models, including meta-regression, network meta-analysis and individual patient data meta-analysis14. This approach is based on the decomposition of Fisher’s information matrix and thus the derived weights are not influenced by the network structure. Further investigation of the degree of agreement between our algorithm and that of Riley et al.14 would be of interest.\n\nIn the example implemented in the Application, there is no other possible set of paths, and associated streams, that could be selected from x to y in order to partition the inflow of x: π1, π2 and π3 is the only possible set of streams (Figure 1c). Thus, even if we were taking paths using different criteria, i.e. from longest to shortest, according to values from the H matrix or even randomly, the percentage contributions given in Table 2 would be identical. However, cases exist where the selection of paths does influence the derivation of the P matrix. In Supplementary File 3, we elaborate on the selection of direct paths in the algorithm and discuss some alternative modifications of the algorithm. We are planning to examine the properties of the different approaches in greater detail in a follow up project.\n\nWe offer an R package10, which we also use in the software application CINeMA (Confidence In Network Meta-Analysis)11, that aims to simplify the evaluation of confidence in the findings from NMA. While CINeMA largely follows the framework previously developed by Salanti et al.3, the refinement of several methodological aspects is currently under development. Core aspects of the approach include the consideration of the relative contributions of each direct comparison to each NMA treatment effect estimate. To this end, CINeMA uses the percentage contribution matrix as described in this paper. The command netweight in Stata has also been updated to use the described approach.\n\nWe believe that the approach described in this paper is a useful and novel addition to network meta-analysis methodology, which allows the consistent derivation of the percentage contributions of direct evidence from individual studies to network treatment effects.\n\n\nData availability\n\nDataset 1: Outcome data from the example network of topical antibiotics for the treatment of chronic otitis media with ear discharge in patients with eardrum perforations7. Data labels: study, name of individual studies; id, id of the individual studies; t; treatment; r, number of events; n, sample size; rob, risk of bias per study. DOI: 10.5256/f1000research.14770.d20317413.", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nTP received funding from a Campbell Collaboration Method Grant (grant No CMG2.04).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary material\n\nSupplementary File 1. Proof that the sum of all outflows from the source is equal to the sum of all inflows to the sink and both are 1.\n\nClick here to access the data.\n\nSupplementary File 2. Percentage contribution matrix P for the network of antimanic drugs. Cells show the percentage contribution of direct comparisons indicated in the column to the network meta-analysis treatment effects indicated in the rows. ASE, asenapine; ARI, aripiprazole; PLA, placebo; HAL, haloperidol; QUE, quetiapine; LITH, lithium; ZIP, ziprasidone; OLA, olanzapine; DIV, divalproex; RIS, risperidone; CARB, carbamazepine; LAM, lamotrigine; PAL, paliperidone; TOP, topiramate; ASE, asenapine.\n\nClick here to access the data.\n\nSupplementary File 3. Considerations on the selection of streams.\n\nClick here to access the data.\n\n\nReferences\n\nEichler HG, Thomson A, Eichler I, et al.: Assessing the relative efficacy of new drugs: an emerging opportunity. Nat Rev Drug Discov. 2015; 14(7): 443–4. PubMed Abstract \| Publisher Full Text\n\nJansen JP, Trikalinos T, Cappelleri JC, et al.: Indirect treatment comparison/network meta-analysis study questionnaire to assess relevance and credibility to inform health care decision making: an ISPOR-AMCP-NPC Good Practice Task Force report. Value Health. 2014; 17(2): 157–73. PubMed Abstract \| Publisher Full Text\n\nSalanti G, Del Giovane C, Chaimani A, et al.: Evaluating the quality of evidence from a network meta-analysis. PLoS One. 2014; 9(7): e99682. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKönig J, Krahn U, Binder H: Visualizing the flow of evidence in network meta-analysis and characterizing mixed treatment comparisons. Stat Med. 2013; 32(30): 5414–29. PubMed Abstract \| Publisher Full Text\n\nLu G, Welton NJ, Higgins JP, et al.: Linear inference for mixed treatment comparison meta-analysis: A two-stage approach. Res Synth Methods. 2011; 2(1): 43–60. PubMed Abstract \| Publisher Full Text\n\nRücker G: Network meta-analysis, electrical networks and graph theory. Res Synth Methods. 2012; 3(4): 312–24. PubMed Abstract \| Publisher Full Text\n\nMacfadyen CA, Acuin JM, Gamble C: Topical antibiotics without steroids for chronically discharging ears with underlying eardrum perforations. Cochrane Database Syst Rev. 2005; (4): CD004618. PubMed Abstract \| Publisher Full Text\n\nCormen TH editor.: Introduction to algorithms. 2. ed., 8. printing, Cambridge, Mass.: MIT Press [u.a.]; 2007; 1180. Reference Source\n\nR Core Team: R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. 2018.\n\nFlow contribution R package [Internet]. Reference Source\n\nConfidence In Network Meta-Analysis [Internet]. Reference Source\n\nRücker G, Schwarzer G: Reduce dimension or reduce weights? Comparing two approaches to multi-arm studies in network meta-analysis. Stat Med. 2014; 33(25): 4353–69. PubMed Abstract \| Publisher Full Text\n\nPapakonstantinou T, Nikolakopoulou A, Rücker G, et al.: Dataset 1 in: Estimating the contribution of studies in network meta-analysis: paths, flows and streams. F1000Research. 2018. Data Source\n\nRiley RD, Ensor J, Jackson D, et al.: Deriving percentage study weights in multi-parameter meta-analysis models: with application to meta-regression, network meta-analysis and one-stage individual participant data models. Stat Methods Med Res. 2017; 962280216688033. PubMed Abstract \| Publisher Full Text" }	[ { "id": "34177", "date": "13 Jun 2018", "name": "Jochem König", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors undertake a redefinition of the term ‘percentage contribution’ in network meta-analysis. In ordinary meta-analysis of randomized clinical trials, pooled estimates of a treatment effect can be represented as a weighted mean of treatment effects from all contributing trials. Weights are positive and sum to one and naturally constitute the ‘percentage contribution’ of each study. In network meta-analysis treatment effect estimates are defined for all pairs of treatments. Each can be represented as a two-stage estimate, firstly pooling all trials comparing the same set of treatments, respectively, which gives direct effect estimates, and then pooling direct estimates into a network based mixed treatment comparison. The linear coefficients, used in the second stage constitute the ‘contribution of direct evidence to a network based comparison’. They can be read off, from the H matrix, as described by the authors. They share some but not all aspects with weights of a two-armed meta-analysis. Aspects shared in some sense are the flow properties described by the authors and in citation [4]. The aspect not shared, is that summation to one.\nIndeed, in a purely linear graph connecting two extreme treatments through a chain of, say 10, pairwise intermediary comparisons, the network estimate is the sum of all direct effects. Any bias present in one direct effect estimate translates 1:1 into the network based estimate. Accordingly, the entries of the H matrix are all equal to 1 in the relevant row.\nBoth, Salanti [3] and the present paper aim to rescale or transform the rows of the H matrix in order to achieve a ‘sum to one’ condition. In the case of the linear graph, the percentage contribution is 10% for each direct comparison. I am not convinced, that this number adequately captures the role of each direct effect estimate in this example.\nConsider now, an example where two treatments are connected through three paths with one, two, and three edges, respectively, each with flow 1/3. Then each direct comparison has the same influence on the network based treatment effect estimate: A bias of 1 unit in one direct estimate translates into a bias of 1/3 unit in the network based estimate. These proportions are conserved in Salanti’s notion of percentage contribution, but not so in the newly introduced concept. Salanti [3] attaches a percentage contribution of 1/6 to each comparison just as if we had pooled six equally precise trials comparing the same pair of treatments. But the influence of direct comparisons is different in this network and should be characterized by the number 1/3.\nThat is why I have objections against Salanti’s [3] concept of percentage contribution and even more so against the newly introduced concept, which attaches unequal contribution quantities of 1/9, 1/6 and 1/3, respectively.\nThe new method should not be introduced into routine of network meta-analysis. If this paper is going to be indexed, its draw backs and caveats should be very clearly set out.\n\nThe concept of streams is a nice one. It allows to represent a network based treatment effect as a weighted sum of a direct and indirect effect estimates each corresponding to one path from source to sink. Note however that these effects are stochastically interdependent and hence, the aggregation of streams is different from the usual process of pooling evidence.\n\nTo be somewhat less destructive, I propose to arrange the network graph with the source to the left and the sink to the right and all other vertices placed, such that all directed edges point from left to right. Then, due to the flow properties, any vertical cross-section gives a set of flows that sum to one. In that sense, the untransformed rows of the H matrix already contain percentage contributions. Displays used to present voter hiking can be used to represent these flows (see e.g. www.zeit.de/politik/deutschland/2017-05/waehlerwanderung-nrw-landtagswahl-cdu-spd-fdp).\n\nI propose to modify Table 3 and Figure 3 in a way that the direction of influence becomes clear. Does a positive bias in one contributing direct effect translate into a positive or negative bias of the network based effect estimate? The colored fields in Figure 3 could be labelled with meaningfully ordered pairs of treatment letters.\n\nDiscussion “The calculation of the percentages uses the estimated variances of direct effects and thus incorporate associated uncertainty in their estimation.” Note however, that in a linear network each direct effect estimate contributes equally, irrespective of the size and precision of the study.\n\nDiscussion “Applying the algorithm to networks of interventions can be used to quantify the contribution of potential study limitations to the NMA treatment effect estimates.” The two stage approach discussed in this paper is not essential. Both the flows and the percentage contribution defined in the paper can be distributed to single trials according to the weights used when pooling them into direct effect estimates. Then the risk of bias of single trials can be analyzed. A display similar to Figure 3 is possible that contains separately colored fields for individual studies.\n\nDiscussion, 2nd paragraph: Concerning multi-armed designs, König et al.[4] have defined the flow into and out of the vertices in a way that does not depend on the choice of reference treatment.\nSupplementary File 3: 1st paragraph: What is the vector space of Gxy over R? 2nd paragraph: ‘The number of directed paths’ obviously refers to the number of directed paths „which suffice to ‘spend’ the flow from source to sink”.\nThe authors state that, if no bridges exist, this number equals to df+1. However, it depends on the flows, too. Consider a network composed of two squares connected at one vertex. It has df=2 inconsistency degrees of freedom and generally 3 paths are needed to spend the flow between the two most distant vertices. However, if all edges have equal flows, only two paths are needed. Moreover, the connecting vertex has the same effect as the bridging edges. Hence the argument has to be extended to the case of graphs that can be split by cutting through a vertex.\nNote also, that a network estimate that compares treatments from different subgroups of treatments connected through a bridge or a ‘breaking’ vertex can be written as the sum of three respective two subnetwork effect estimates: one comparing the source treatment to the bridge post, one for the bridge, if present, and one comparing the second bridge post to the sink. Because it is an unweighted sum, the contribution of the three (or two) summands should be equal.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Partly", "responses": [ { "c_id": "3910", "date": "03 Sep 2018", "name": "Adriani Nikolakopoulou", "role": "Author Response", "response": "We are grateful to Dr. Jochem König for the time and effort he spent to review our paper. We believe that his valuable comments and suggestions have substantially improved the manuscript. We have addressed all his comments and revised the paper accordingly. Below you can find detailed replies to the reviewer’s suggestions and comments. Referee Report 13 Jun 2018 Jochem König, Division of Paediatric Epidemiology, Institute of Medical Biostatistics, Epidemiology, and Informatics (IMBEI), Johannes Gutenberg University Medical Center, Mainz, Germany Approved with Reservations 1. The authors undertake a redefinition of the term ‘percentage contribution’ in network meta-analysis. In ordinary meta-analysis of randomized clinical trials, pooled estimates of a treatment effect can be represented as a weighted mean of treatment effects from all contributing trials. Weights are positive and sum to one and naturally constitute the ‘percentage contribution’ of each study. In network meta-analysis treatment effect estimates are defined for all pairs of treatments. Each can be represented as a two-stage estimate, firstly pooling all trials comparing the same set of treatments, respectively, which gives direct effect estimates, and then pooling direct estimates into a network based mixed treatment comparison. The linear coefficients, used in the second stage constitute the ‘contribution of direct evidence to a network based comparison’. They can be read off, from the H matrix, as described by the authors. They share some but not all aspects with weights of a two-armed meta-analysis. Aspects shared in some sense are the flow properties described by the authors and in citation [4]. The aspect not shared, is that summation to one. Indeed, in a purely linear graph connecting two extreme treatments through a chain of, say 10, pairwise intermediary comparisons, the network estimate is the sum of all direct effects. Any bias present in one direct effect estimate translates 1:1 into the network based estimate. Accordingly, the entries of the H matrix are all equal to 1 in the relevant row. Both, Salanti [3] and the present paper aim to rescale or transform the rows of the H matrix in order to achieve a ‘sum to one’ condition. In the case of the linear graph, the percentage contribution is 10% for each direct comparison. I am not convinced, that this number adequately captures the role of each direct effect estimate in this example. Consider now, an example where two treatments are connected through three paths with one, two, and three edges, respectively, each with flow 1/3. Then each direct comparison has the same influence on the network based treatment effect estimate: A bias of 1 unit in one direct estimate translates into a bias of 1/3 unit in the network based estimate. These proportions are conserved in Salanti’s notion of percentage contribution, but not so in the newly introduced concept. Salanti [3] attaches a percentage contribution of 1/6 to each comparison just as if we had pooled six equally precise trials comparing the same pair of treatments. But the influence of direct comparisons is different in this network and should be characterized by the number 1/3. That is why I have objections against Salanti’s [3] concept of percentage contribution and even more so against the newly introduced concept, which attaches unequal contribution quantities of 1/9, 1/6 and 1/3, respectively. The new method should not be introduced into routine of network meta-analysis. If this paper is going to be indexed, its draw backs and caveats should be very clearly set out. Thank you for this insightful comment. We agree with the point you make which we find that enlightens potential applications or extensions of the presented method to account for bias in network meta-analysis. We would like to make a distinction between 'representation of the contribution of pieces of evidence in a 0 to 1 scale' and 'using this percentage contributions to represent and measure the impact and directionality of potential bias'. In this paper, we focus on the first aspect of translating the matrix into percentages and we subsequently apply it to depict the contribution of pieces of evidence according to a study characteristic. Our aim is to give a representation of the evidence that each comparison –or study, see our response to comment 6 below- contributes to the estimation of each NMA treatment effect without considering how this evidence could bias the results. Note that the bar plot of figure 3 gives a picture of the contribution of high, moderate or low risk of bias evidence without giving insight on the amount or direction of bias. Quantification of bias in the direct evidence and its impact on the NMA estimates is subject of further research but not addressed within this paper. We have now added a paragraph in the Discussion to highlight the limitations of the presented method as you recommend. The paragraph reads: “Study limitations may lead to biased NMA treatment effect; however, the amount and direction of bias in the NMA treatment effect as a result of the within-study bias is not straightforward to define and is not currently accommodated within the percentage contribution matrix. First, a single biased trial may affect an entire indirect route; thus, even if its percentage contribution is small, its consequences in the estimation of the NMA treatment effect may be important. Second, the direction of bias across studies involved in a stream may vary. For example, bias in two comparisons in the same stream may either cancel out or add-up in favor of one of the two treatments. We aim to extend the methods presented in this paper to develop a network meta-regression model that will use the direction and the amount of bias to determine whether and how much NMA treatment effect estimates will be biased as the result of within-study bias.” 2. The concept of streams is a nice one. It allows to represent a network based treatment effect as a weighted sum of a direct and indirect effect estimates each corresponding to one path from source to sink. Note however that these effects are stochastically interdependent and hence, the aggregation of streams is different from the usual process of pooling evidence. Thank you for this comment; indeed the aggregation of streams is not equivalent to the aggregation of studies in a pairwise meta-analysis. We write in ‘Percentage contributions of direct comparisons’: “It can be shown that an NMA effect is a linear combination of the direct effects combined within paths.” And we now added in the same section: “Equation 5 represents θxyN as a weighted sum of direct and two indirect effect estimates; the effects are stochastically interdependent and, hence, their aggregation is different from the aggregation of studies in a pairwise meta-analysis.” 3. To be somewhat less destructive, I propose to arrange the network graph with the source to the left and the sink to the right and all other vertices placed, such that all directed edges point from left to right. Then, due to the flow properties, any vertical cross-section gives a set of flows that sum to one. In that sense, the untransformed rows of the H matrix already contain percentage contributions. Displays used to present voter hiking can be used to represent these flows (see e.g. www.zeit.de/politik/deutschland/2017-05/waehlerwanderung-nrw-landtagswahl-cdu-spd-fdp). Thank you, we agree that horizontal representation is more intuitive. We re-arranged figures 1 and 2 accordingly. 4. I propose to modify Table 3 and Figure 3 in a way that the direction of influence becomes clear. Does a positive bias in one contributing direct effect translate into a positive or negative bias of the network based effect estimate? The colored fields in Figure 3 could be labelled with meaningfully ordered pairs of treatment letters. Thank you very much for this comment which we believe is related to your comment 1. We agree that such a representation would be useful but we believe it could confuse readers with respect to the scope of this paper which is not to quantify the impact of within-study bias. As said in comment 1, the scope of this paper is to represent the percentage contributions of each piece of evidence; combining this information with information of study characteristics in a figure such as figure 3 may be a useful representation of the contribution of these characteristics to the estimation of NMA treatment effects. The influence of these characteristics, though, is not straightforwardly derived and it is subject to further research. Thus, we would prefer to leave Table 3 and Figure 3 unmodified and construct such representations of the direction of bias in future work where we aim to develop models to account for study-level bias using the presented methodology. Please also see the reply to comment 1 and the addition to the Discussion. 5. Discussion “The calculation of the percentages uses the estimated variances of direct effects and thus incorporate associated uncertainty in their estimation.” Note however, that in a linear network each direct effect estimate contributes equally, irrespective of the size and precision of the study. We agree this statement is confusing and hence deleted this phrase from the Discussion. The paragraph now reads: “In this paper, we present a new approach to derive percentage contributions of individual studies to the treatment effect estimates in NMA. We made use of the fact that the composition of network treatment effect estimates can be interpreted as a flow of evidence. An assumption that underlies our algorithm is the equal split of the stream flow to the involved comparisons. Although indirect effects are not weighted averages, we find this approximation to be a pragmatic approach that reasonably reflects the amount that each comparison contributes to network effects.” 6. Discussion “Applying the algorithm to networks of interventions can be used to quantify the contribution of potential study limitations to the NMA treatment effect estimates.” The two stage approach discussed in this paper is not essential. Both the flows and the percentage contribution defined in the paper can be distributed to single trials according to the weights used when pooling them into direct effect estimates. Then the risk of bias of single trials can be analyzed. A display similar to Figure 3 is possible that contains separately colored fields for individual studies. Thank you for this comment; we agree that such a distribution of percentage contributions to single trials is possible and relevant and we have updated the paper accordingly. 1.We added an extra paragraph/subsection ‘Percentage study contributions’ where we elaborate on the process you describe. The new paragraph reads: “Matrix P (Table 3) shows the percentage contributions of each direct comparison to each NMA treatment effect estimate. These percentages can be distributed to individual studies within each comparison according to their weights from direct meta-analyses. For example, pxy=63.5% and there are two studies examining the xy comparison. The individual study weights for the two studies are 0.69 and 1.54 resulting to study percentage contributions of (0.69 /(0.69+1.54))63.5%=19.6% and (1.54/(0.69+1.54))63.5%=43.8% to the xy NMA treatment effect estimate. The application of this process to the entire matrix P leads to the matrix P* shown in Table 4. Adjusted weights as proposed by Rücker & Schwarzer (9) are used for multi-arm studies.” 2.Section ‘Using percentage contributions to quantify the impact of a characteristic in a direct comparison’ has been renamed to ‘Using percentage study contributions to quantify the impact of a characteristic in a direct comparison’ and follows the new ‘Percentage study contributions’ section. Text has been updated to refer to studies instead of direct comparisons and Figure 3 (and the new Table 4) now displays percentage study contributions to NMA treatment effect estimates. 3.The CINeMA (Confidence In Network Meta-Analysis) http://cinema.ispm.ch/ software has also been updated to display the bar chart using study contributions. 4. The first sentence of the Discussion now reads: “In this paper, we present a new approach to derive percentage contributions of individual studies to the treatment effect estimates in NMA.” 7. Discussion, 2nd paragraph: Concerning multi-armed designs, König et al.[4] have defined the flow into and out of the vertices in a way that does not depend on the choice of reference treatment. Thank you for this comment, we could indeed use the H matrix as described in section 3.3.1 of König et al.[4] (and implemented in netmeta as H.tilde) for multi-arm designs; some concerns about the interpretability of such a matrix have previously precluded us of doing so. In fact, the algorithm could be applied to any H matrix as long as its rows have the flow properties described in König et al.[4] and in our section ‘Comparison graph’. We have removed the discussion on multi-arm trials from the Discussion and added the following paragraph at the end of the Methods section: “The starting point for the developed algorithm was a simplified version of the H matrix that does not consider the correlation induced by multi-arm trials. Alternatively, one could use the matrix as described by König et al. (4) that properly accommodates multi-arm designs. Note that any matrix whose rows can be interpreted as flow networks can be used as the starting point of the algorithm.” Supplementary File 3: 1st paragraph: What is the vector space of Gxy over R? Thank you for this comment; as it was not clear and in order to simplify this part, we eliminated the notions of ‘vector space’ and ‘basis’ in the document. It now reads: “However, in the general case, the number of paths used in the algorithm is smaller than the number of elements of Π. The number of directed paths which suffice to ‘spend’ the flow from source to sink is less or equal to df+1 where df is the inconsistency degrees of freedom in the Lu and Ades inconsistency model (1), df=D-T+1.” 2nd paragraph: ‘The number of directed paths’ obviously refers to the number of directed paths „which suffice to ‘spend’ the flow from source to sink”. Thank you, we corrected it. The authors state that, if no bridges exist, this number equals to df+1. However, it depends on the flows, too. Consider a network composed of two squares connected at one vertex. It has df=2 inconsistency degrees of freedom and generally 3 paths are needed to spend the flow between the two most distant vertices. However, if all edges have equal flows, only two paths are needed. Moreover, the connecting vertex has the same effect as the bridging edges. Hence the argument has to be extended to the case of graphs that can be split by cutting through a vertex. Note also, that a network estimate that compares treatments from different subgroups of treatments connected through a bridge or a ‘breaking’ vertex can be written as the sum of three respective two subnetwork effect estimates: one comparing the source treatment to the bridge post, one for the bridge, if present, and one comparing the second bridge post to the sink. Because it is an unweighted sum, the contribution of the three (or two) summands should be equal. Thank you very much for these interesting notes with which we totally agree. We extended the ‘Number of directed paths’ section in Supplementary File 3 to include discussion on the ‘breaking’ vertex situation and the consequences you describe. In particular we wrote: “Another situation where the number of directed paths which suffice to ‘spend’ the flow from source to sink is less than df+1 elements occurs when a vertex has at least two inflows and two outflows, all equal between them. In such a situation the flow from source to sink may be spent in less than df+1 directed paths. Such a vertex will be called breaking vertex. Note that an NMA treatment effect estimate for treatments separated by a bridge or a breaking vertex (say treatments x and y) can be seen as an “indirect” comparison through either the intermediate bridge (constituting by treatments b1 and b2) or the intermediate breaking vertex (b0). In particular, in the case of a bridge, the NMA treatment effect estimate is derived as θxyN=θxb1N+θb1b0N+θb2yN and in the case of a breaking vertex the NMA treatment effect estimate is θxyN=θxb0N+θb0yN It is, thus, implied that in such a case the separate subnetworks contribute equally to the estimation of θxyN.” Is the rationale for developing the new method (or application) clearly explained? Yes Is the description of the method technically sound? Yes Are sufficient details provided to allow replication of the method development and its use by others? Yes If any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes Are the conclusions about the method and its performance adequately supported by the findings presented in the article? Partly Competing Interests: No competing interests were disclosed. I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above." } ] }, { "id": "36470", "date": "14 Aug 2018", "name": "Annette M. O'Connor", "expertise": [ "Reviewer Expertise Epidemiology", "infectious diseases", "research synthesis" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper described an updated approach to deriving the percentage contributions of the direct comparison used in treatment estimates. In the interests of full disclosure, I am not a statistician and was reviewing from the view point of an applied user of NMA. There are several papers on this topic in recent years, and it is an active body of work in this area. The article is straightforward to read, although made considerably more readable if one is aware of the content in references, 3,4 and 5.\n\nI have to admit that this is an example of why I appreciate the open review system. I enjoyed the paper and found it easy to read and follow, but the others reviewer's comments were a significant contribution also. I don't have any additional critique of the proposed approach, but I look forward to seeing how the investigators address those or provided some discussion, as this would help the less statistically inclined reader like me. Again, the reason I wanted this, is that I am an applied user and such discussions are very helpful. In particular, I would like to have seen some discussion of the 1st comment. I would also like to see if the comment about Figure 3 could be incorporated - perhaps not feasible.\n\nConcerning the approach to calculating the % contribution and how to weight the flows, I came away with the impression that the decision of equal split of the stream flow as arbitrary (perhaps not precisely the correct term) or perhaps pragmatic is better. Therefore, it is not surprising this is a debatable approach. Again, this was another reason why I am looking forward to see the investigators responses to the 1st reviewer's comments. I agree that the comparison of the methods of deriving the percentage contribution would be of interest but was not expecting that to be included.\n\nI look forward to this comparison being published.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4280", "date": "13 Dec 2018", "name": "Adriani Nikolakopoulou", "role": "Author Response", "response": "We are grateful to Dr. Annette M. O’Connor for her time and valuable comments and for recommending approval of our paper. We believe that we have been able to address all the comments raised and we have prepared a next version (Version 3) of the paper. Below are the detailed replies to reviewer’s comments; our responses to the first reviewer, Dr. Jochem König, can be found in https://f1000research.com/articles/7-610/v2. Referee Report: Annette M. O'Connor, Department of Veterinary Diagnostic and Production Animal Medicine, College of Veterinary Medicine, Iowa State University, Ames, IA, USA Approved This paper described an updated approach to deriving the percentage contributions of the direct comparison used in treatment estimates. In the interests of full disclosure, I am not a statistician and was reviewing from the view point of an applied user of NMA. There are several papers on this topic in recent years, and it is an active body of work in this area. The article is straightforward to read, although made considerably more readable if one is aware of the content in references, 3,4 and 5. I have to admit that this is an example of why I appreciate the open review system. I enjoyed the paper and found it easy to read and follow, but the others reviewer's comments were a significant contribution also. I don't have any additional critique of the proposed approach, but I look forward to seeing how the investigators address those or provided some discussion, as this would help the less statistically inclined reader like me. Again, the reason I wanted this, is that I am an applied user and such discussions are very helpful. In particular, I would like to have seen some discussion of the 1st comment. I would also like to see if the comment about Figure 3 could be incorporated - perhaps not feasible.Thank you very much for your positive feedback. Our answers to the first reviewer, Dr. Jochem König, can be found in https://f1000research.com/articles/7-610/v2. In summary, regarding his first comment, we distinguish between representation of contribution of pieces of evidence according to a study characteristic (addressed in this paper) and quantification of the impact of within-study bias in NMA treatment effects (subject of future research). A new paragraph in the Discussion highlighting these concerns reads: “Study limitations may lead to biased NMA treatment effect; however, the amount and direction of bias in the NMA treatment effect as a result of the within-study bias is not straightforward to define and is not currently accommodated within the proportion contribution matrix. First, a single biased trial may affect an entire indirect route; thus, even if its proportion contribution is small, its consequences in the estimation of the NMA treatment effect may be important. Second, the direction of bias across studies involved in a stream may vary. For example, bias in two comparisons in the same stream may either cancel out or add-up in favor of one of the two treatments. We aim to extend the methods presented in this paper to develop a network meta-regression model that will use the direction and the amount of bias to determine whether and how much NMA treatment effect estimates will be biased as the result of within-study bias.”As we explain in detail in our response to Dr. Jochem König (comment 4), we did not amend Figure 3 to incorporate directionality.Concerning the approach to calculating the % contribution and how to weight the flows, I came away with the impression that the decision of equal split of the stream flow as arbitrary (perhaps not precisely the correct term) or perhaps pragmatic is better. Therefore, it is not surprising this is a debatable approach. Again, this was another reason why I am looking forward to see the investigators responses to the 1st reviewer's comments. I agree that the comparison of the methods of deriving the percentage contribution would be of interest but was not expecting that to be included. I look forward to this comparison being published. We agree that equal splitting of stream flow is to some extent arbitrary but also pragmatic. We say in the Discussion: “An assumption that underlies our algorithm is the equal split of the stream flow to the involved comparisons. Although indirect effects are not weighted averages, we find this approximation to be a pragmatic approach that reasonably reflects the amount that each comparison contributes to network effects.”" } ] }, { "id": "37273", "date": "16 Aug 2018", "name": "John R. Thompson", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis report is also available as a separate PDF.\nI have no special knowledge of the measurement of information flow in a network meta-analysis and so I read this paper as a biostatistician with a general interest in meta-analysis.\nI like the idea of associating a descriptive measure of importance to each edge in a network and the authors of this paper present an interesting step in that direction, though my feeling is that this proposal will not turn out to be the final solution. I have two areas of concern, one is to do with terminology and the other with the scope of applicability of the proposal solution.\nIn criticising the terminology, I have to accept that the authors are largely following common practice and so my criticisms are partly aimed at the meta-analysis community. None the less there were three things that irritated me. Firstly, throughout the paper, proportions are referred to as percentages. Next, the title of the paper says it is about \"the contribution of studies in network meta-analysis\" while actually it is about the contribution of different effect estimates. Finally, the measure adopted is the weight given to each effect estimate when calculating the pooled estimate. In this paper and elsewhere in the meta-analysis literature, this weight is called a contribution. Perhaps I am being too pedantic but the weight and contribution are different ideas and it does not help with the terminology confuses them.\n\nNow the proposed method. Taking the example from the paper, the authors note that the pooled estimate of the xy effect can be calculated using equation 1, where θ̂Ixvy is the indirect estimate of xy along the path xy. Further it is true, at least for this example, that ΣLiφi = 1 where Li is the length of the path. So we can attach the weights φ to the edges of the network, sum them when an edge contributes to more than one estimate and the resulting weights will sum to one over the whole network.\n\nThis argument works for the example presented in the paper but it is not clear to me what conditions have to hold for it to work generally. The authors note in the paper that including a multi-arm trial in the meta-analysis would cause a problem, presumably because some of the direct or indirect estimates would not be independent. Are there any other conditions that have to hold? For example, can we have any structure of random effects in the meta-analysis model? What about the Bayesian models that are often used for network meta-analysis?\nThe authors take the matrix, H, which projects individual estimates such as xy, xy, uy, etc. into their predicted values under the meta-analysis model and they present an algorithm for converting those values into weights that are equivalent to the φ's.The algorithm is sensible and works for the simple example in the paper but one is again left wondering whether or not it works under all circumstances. After all, the algorithm is presented without any proof that it works.\nMy own feeling is that a contribution is best measured by the sledgehammer approach of analysing the network with and without a particular edge, but the authors' suggested approach is much less computationally demanding and I think that it would be appreciated by many applied researchers provided they were certain that it could be safely used with their particular network and their particular meta-analysis model.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Partly", "responses": [ { "c_id": "4279", "date": "13 Dec 2018", "name": "Adriani Nikolakopoulou", "role": "Author Response", "response": "We are grateful to Dr. John R. Thompson for his time and valuable comments and for recommending approval of our paper. We believe that we have been able to address all the comments raised and we have prepared a next version (Version 3) of the paper. Below are the detailed replies to reviewer’s comments; our responses to the first reviewer, Dr. Jochem König, can be found in https://f1000research.com/articles/7-610/v2.Referee Report 16 Aug 2018John R. Thompson, Department of Health Sciences, University of Leicester, Leicester, UK Approved I have no special knowledge of the measurement of information flow in a network meta-analysis and so I read this paper as a biostatistician with a general interest in meta-analysis.I like the idea of associating a descriptive measure of importance to each edge in a network and the authors of this paper present an interesting step in that direction, though my feeling is that this proposal will not turn out to be the final solution. I have two areas of concern, one is to do with terminology and the other with the scope of applicability of the proposal solution.In criticising the terminology, I have to accept that the authors are largely following common practice and so my criticisms are partly aimed at the meta-analysis community. None the less there were three things that irritated me. Firstly, throughout the paper, proportions are referred to as percentages. Next, the title of the paper says it is about \"the contribution of studies in network meta-analysis\" while actually it is about the contribution of different effect estimates. Finally, the measure adopted is the weight given to each effect estimate when calculating the pooled estimate. In this paper and elsewhere in the meta-analysis literature, this weight is called a contribution. Perhaps I am being too pedantic but the weight and contribution are different ideas and it does not help with the terminology confuses them.Thank you for these interesting notes on meta-analysis terminology. At first, we agree that referring to “percentage contributions” might not be fitting and calling them “proportions” is more appropriate. We, thus, replaced throughout percentages with proportions.Regarding the reporting of contribution of studies in the title of our paper, we have extended the method to quantify the contribution of studies as suggested by the first reviewer, Dr. Jochem König (comment 6). Thus, we left the title of the paper unmodified.Weights and contributions are indeed sometimes used interchangeably. In this paper, however, we believe that the definition of matrix and the set of flows render the distinction clear. We only talk about “weights” of paths in the Introduction, where we have not yet explained the terminology to be used and thus it is convenient describing the elements of matrix as generalized weights from pairwise meta-analysis. We apologise in advance if we have misunderstood your comment.Now the proposed method. Taking the example from the paper, the authors note that the pooled estimate of the xy effect can be calculated using equation 1, where θ̂Ixvy is the indirect estimate of xy along the path xy. Further it is true, at least for this example, that ΣLiφi = 1 where Li is the length of the path. So we can attach the weights φ to the edges of the network, sum them when an edge contributes to more than one estimate and the resulting weights will sum to one over the whole network. This argument works for the example presented in the paper but it is not clear to me what conditions have to hold for it to work generally. The authors note in the paper that including a multi-arm trial in the meta-analysis would cause a problem, presumably because some of the direct or indirect estimates would not be independent. Are there any other conditions that have to hold? For example, can we have any structure of random effects in the meta-analysis model? What about the Bayesian models that are often used for network meta-analysis?Thank you for this insightful comment. We have clarified at the beginning of the Methods section that we consider “a simplified version of the matrix described by König et al. [1] that does not take into account the correlation induced by multi-arm trials.”In this case, and considering the properties of the elements of set it always holds that Equation 3 can be re-written as Equation 5. We now clarify in the subsection ‘Proportion contributions of direct comparisons’:“It follows from the properties of the elements of set that each NMA effect can be written as a linear combination of direct and indirect effects, in the form of Equation 5”.We clarify at the end of ‘Algorithm to decompose flows into percentage contributions’ section that multi-arm studies can be handled if a modified matrix is used. We also clarify that the structure of random effects does not play a role in the calculations. In particular, we write:“The starting point for the developed algorithm was a simplified version of the matrix that does not consider the correlation induced by multi-arm trials. Alternatively, one could use the matrix as described by König et al. [1] that extends the definition of the matrix for multi-arm designs. Note that any matrix whose rows can be interpreted as flow networks can be used as the starting point of the algorithm. The estimator of heterogeneity as well as the assumption of a different or common heterogeneity across comparisons in the network does not modify any aspect of the method.”Within a Bayesian framework, the matrix is not defined. We added in the Discussion:“The derivation of the contribution of sources of evidence in a Bayesian NMA is not straightforward, as no analogue to the matrix exists. The application of the method described in this paper, however, can be used to derive useful approximations of the contributions of studies.”The authors take the matrix, H, which projects individual estimates such as xy, xy, uy, etc. into their predicted values under the meta-analysis model and they present an algorithm for converting those values into weights that are equivalent to the φ's.The algorithm is sensible and works for the simple example in the paper but one is again left wondering whether or not it works under all circumstances. After all, the algorithm is presented without any proof that it works.Thank you for pointing this out. The algorithm we present in the ‘Algorithm to decompose flows into proportion contributions’ is described for the general case of a network of interventions containing any number of studies, treatments and loops. Indeed, we do not present a proof that it works under all circumstances. We have, however, applied the algorithm in numerous NMA applications apart from the simple example used in the results of this paper. Moreover, as mentioned in the Discussion, the web application CINeMA (Confidence In Network Meta-Analysis) [2], that aims to simplify the evaluation of confidence in the findings from NMA, uses the proportion contribution matrix as described in this paper. Users of CINeMA have not reported any case that the algorithm does not work. We are planning to examine the properties of the different approaches on the selection of direct paths in the algorithm in a follow up project (preliminary work is described in Supplementary File 3), which could be combined with a proof that the algorithm (and its modifications) works.My own feeling is that a contribution is best measured by the sledgehammer approach of analysing the network with and without a particular edge, but the authors' suggested approach is much less computationally demanding and I think that it would be appreciated by many applied researchers provided they were certain that it could be safely used with their particular network and their particular meta-analysis model.We totally agree that such ‘leave one out’ approaches would be also of interest. Relevant methods have been proposed by Krahn et al. [3] for contribution to the effect and by Riley et al. [4] for contribution to the variance. The former is useful in measuring how much a direct effect could bias the NMA effect and thus is not relevant in this case where we opted for a representation of contribution of pieces of evidence according to a study characteristic. The latter is based on the decomposition of Fisher’s information matrix and is associated with some issues that need further investigation.We had already mentioned and discussed the Riley et al. approach in the Discussion. We now added:“Krahn et al. define influence functions to describe the extent to which changes in study effects would be translated into NMA treatment effects [3].”" } ] } ]	1	https://f1000research.com/articles/7-610
https://f1000research.com/articles/7-359/v1	22 Mar 18	{ "type": "Opinion Article", "title": "Factors targeting MED12 to drive tumorigenesis?", "authors": [ "Jörn Bullerdiek", "Birgit Rommel", "Birgit Rommel" ], "abstract": "Mediator Subcomplex 12 (MED12) is part of the transcriptional preinitiation machinery. Mutations of its gene predominantly occur in two types of highly frequent benign tumors, uterine leiomyomas and fibroadenomas of the breast, where they apparently act as driver mutations. Nevertheless, their presence is not restricted to benign tumors having been found at considerable frequencies in uterine leiomyosarcomas, malignant phyllodes tumors, and chronic lymphocytic leukemia also. Most of the mutations are located within exon 2 of the gene but in rare cases the intron 1/exon 2 boundary or exon 1 are affected. As to their type, predominantly single nucleotide exchanges with a hotspot in one codon are found, but small deletions clustering around that hotspot also are not uncommon. According to their presumed classification as gain-of-function mutations, these latter deletions are leaving the open reading frame intact. As to the types of mutations, so far no apparent differences between the tumor entities affected have emerged. Interestingly, this pattern with small deletions clustered around the hotspot of single nucleotide exchanges resembles that seen as a result of targeted gene editing. In contrast to other driver mutations the percentage of MED12-mutation positive tumors of independent clonal origin increases with the number of tumors per patient suggesting unknown etiological factors supporting site specific mutagenesis. These factors may act by inducing simultaneous site-specific double strand breaks the erroneous repair of which may lead to corresponding mutations. As inducers of DNA damage and its repair such as foreign nucleic acids of the microbiome displaying sequence homology to the putative target site might play a role. Interestingly, a 16 base pair homology of the hotspot to a putative terminator base-paired hairpin sequence of a Staphylococcus aureus tRNA gene cluster has been noted which might form R-loop like structures with its target sequence thus inducing said changes.", "keywords": [ "Mediator subcomplex 12 (MED12)", "mutations", "uterine leiomyomas", "fibroadenomas of the breast", "chronic lymphocytic leukemia (CLL)", "multiple tumors", "DNA-RNA hybrids", "Staphylococcus aureus" ], "content": "Introduction\n\nSurprisingly, the most common mutation in human tumors does not affect one of the famous suspects in the field (for review see (Vogelstein et al., 2013)) but the much less well-known gene encoding Mediator Subcomplex 12 (MED12). MED12 is part of the transcriptional preinitiation complex CDK8 (Elmlund et al., 2006) and encoded by a gene that maps to the X-chromosome at Xq13.1. One obvious reason why its mutations so far have gained much less interest than those of other genes frequently mutated in human tumors is that they affect, to a large extent, benign tumors. Moreover, within malignant tumors, they are virtually absent from most of the predominant epithelial neoplasms like cancers of colon, breast, and lung (Kandoth et al., 2013) whereas they only have been found at considerable frequencies in some malignant tumors of non-epithelial origin.\n\nAs in the benign tumors, however, mutations of that gene occur as apparent driver mutations in a predominant subset of uterine leiomyomas (Mäkinen et al., 2011; Markowski et al., 2012; McGuire et al., 2012), constituting the by far most frequent human symptomatic tumors of all. Likewise, these mutations are also found in a large subset of fibroadenomas of the breast (Lim et al., 2014; Piscuoglio et al., 2015; Yoshida et al., 2015), another frequent benign tumor which occurs predominantly in young and middle-aged women. Interestingly, they are not restricted to benign tumors but also frequently seen in their malignant counterparts, i.e. uterine leiomyosarcomas (Kämpjärvi et al., 2012; Markowski et al., 2013a; Pérot et al., 2012) and malignant phyllodes tumors (Nagasawa et al., 2015; Piscuoglio et al., 2015; Yoshida et al., 2015). Also, their presence in malignant tumors suggests that, albeit as a very rare event, certain additional mutations can trigger malignant transformation within formerly benign tumors harboring MED12 mutations. Apart from solid tumors, MED12 mutations recently were also detected in a significant percentage of roughly 5–9% of chronic lymphocytic leukemias (CLL) (Guièze et al., 2015; Kämpjärvi et al., 2015; Wu et al., 2017a). Furthermore, the same type of MED12 mutations was found in two canine vaginal leiomyomas (Markowski et al., 2013a).\n\nA closer look at the MED12 mutations does not reveal apparent differences between the type of mutations when comparing the different tumor entities. Predominantly, the mutations are clustered in the 5´ region of exon 2 of the gene with only a few mutations affecting the intron 1-exon 2 boundary or, much rarer, exon 1 or the exon 1-intron 1 boundary. Most of them are single base exchanges clearly clustered at two nucleotides of codon 44 where, albeit with different frequencies, guanins are found to be replaced by either A, C, or T. Besides these single base replacements, deletions and, more rarely, indels, usually affecting exon 2 or the intron 1/exon 2 boundary are found which always leave the reading frame intact indicating gain-of-function mutations. MED12 maps to the X-chromosome and, as revealed by cDNA sequencing, the mutations are apparently restricted to the active X-chromosome (Mäkinen et al., 2011; Markowski et al., 2012; McGuire et al., 2012).\n\nWe feel that for several reasons this highly frequent type of mutation might point to an unusual mechanism of mutagenesis underlying the development of the corresponding tumors. These reasons will be discussed herein and a hypothesis based on target-specific mutagenesis will be presented. Starting with a short introduction of the main tumors affected by MED12 mutations we will then address the molecular pathogenesis of uterine leiomyomas along with its clinical correlations followed by an in depth analysis of the pattern of MED12 mutations. Finally, we will present a hypothesis why these data indicate a so far unknown etiological mechanism favoring these particular highly frequent somatic alterations of the genome.\n\n\nIntroducing three tumor entities displaying a unique type of MED12 mutations\n\nBesides a few very rare tumors, three main tumor entities are often affected by mutations of MED12. First, these three entities, i.e. uterine leiomyomas (fibroids), fibroadenomas of the breast, and chronic lymphocytic leukemias, will be introduced.\n\nUterine leiomyomas (UL) are benign smooth muscle tumors of myometrial origin with an apparently very low tendency to undergo malignant transformation. Depending on their location it can be distinguished between submucosal, intramural, and subserosal UL. Roughly 40–70% of women in their reproductive age will develop UL with a well-documented higher prevalence among women of African and African-American origin. In this group, the UL develop on average at younger ages (Laughlin et al., 2010). Leiomyomas are often of large size and, as a matter of debate for more than 100 years, (Figure 1), multiple nodules occur at almost the same frequency or even more frequently than single nodules.\n\nIllustration of a uterus carrying multiple leiomyomas of various sizes and location (4, left) and uterus with a large pedunculated subserous leiomyoma (right, 6). Copperplate engraving from Virchow´s series of lectures entitled \"Die krankhaften Geschwülste\" (\"Morbid neoplasms\")(Hirschwald, Berlin, 1864/65).\n\nNevertheless, the majority of patients with UL are without symptom. The remaining 20–30% of the patients suffer from symptoms like in particular heavy menstrual bleeding, increased menstrual periods, reduced fertility, and pelvic pressure and pain. During menopause, UL cease growing and even shrink. Nevertheless, despite their benign nature UL are the most symptomatic human tumors of all and repeatedly have been reported even in mummies with the oldest of them dating back to the Middle Neolithic age (3,200–2,500 BC) (Fornaciari & Giuffra, 2012). Surgery (hysterectomy or myomectomy) is a common type of treatment, but a variety of other non-surgical methodologies to treat UL are available (for review see: (Williams, 2017)).\n\nWhile the etiology of these frequent tumors remains unclear (McWilliams & Chennathukuzhi, 2017), a bit more is known about pathogenetic factors. As to basic principles of their molecular pathogenesis, UL behave like the vast majority of other benign and malignant tumors. This includes a monoclonal origin which is triggered by so-called driver mutations. In the case of multiple UL almost every single nodule has been found to be of independent clonal origin (Mashal et al., 1994). Accordingly, different nodules are usually characterized by different driver mutations in these cases.\n\nMoreover, genetic subtypes, apparently belonging to different groups of driver mutations, exist. While MED12 mutations, as a group, constitute the most predominant genetic subtype (Mäkinen et al., 2011), another frequent subgroup of uterine leiomyomas carries rearrangements of the gene encoding the architectural transcription factor High Mobility AT-hook 2 (HMGA2), as a rule reflected by cytogenetically visible chromosomal translocations (Schoenmakers et al., 1995). Of note, the driver mutations of both subgroups occur in a mutually exclusive manner (Markowski et al., 2012). Besides these subgroups, other, more rare but also independent genetic subgroups of UL such as one characterized by either germline (hereditary leiomyomatosis and renal cell cancer (HLRCC), OMIM 605839) (Bayley et al., 2008) or somatic loss-of-function mutations of Fumarate Dehydrogenase (FH) seem to exist (Kämpjärvi et al., 2016).\n\nFibroadenomas of the breast are common benign tumors histologically composed of both stromal and epithelial components preferentially occurring in adolescent and young women. Their general incidence may be in the range of 10% in the corresponding age groups. Multiple tumors are not rare with some 10–15% of the patients having more than one FA.\n\nTheir name fibroadenomas (FA) and their classification as fibroepithelial tumors suggest a biphasic nature of these neoplasms. Nevertheless, as to their pathogenesis this classification seems to be misleading at least in the majority of cases because mutations are restricted to the stromal component of the tumors (Mishima et al., 2015). Between 50% and 60% of the FA harbor MED12 mutations (Lim et al., 2014; Mishima et al., 2015; Pfarr et al., 2015) Histologically, the occurrence of MED12 mutations correlates highly significant with the so-called intracanalicular growth pattern (Mishima et al., 2015; Pfarr et al., 2015). Interestingly, MED12 mutations have also been found in considerable percentages of other breast tumors of presumed stromal origin Phyllodes tumors and malignant Phyllodes tumors. In these tumors, the types of MED12 mutations are not obviously discernible from those observed in UL and FA.\n\nIn Western countries, chronic lymphocytic leukemia (CLL) is the most common type of leukemia in adults. The American Cancer Society expects an estimated number of about 20,940 new cases of CLL in the United States in 2018 with about 4,510 CLL-related deaths. Overall, CLL accounts for about one-quarter of the new cases of leukemia. The average person's lifetime risk of getting CLL is about one in 175 (0.57%). The average age at the time of diagnosis is around 70 years, with men slightly more often affected than women. Frequently, the disease is detected in patients not yet showing any severe symptoms and in its early stages patients often undergo a ‘watch and wait’ period prior to starting therapy.\n\nAs to its pathogenesis CLL is a monoclonal leukemia of B-cell origin with a number of subsets that can be distinguished based on their genetic characterization, apparently pointing to different driver mutations. The genetic alterations also allow the stratification of groups which may require different therapeutic approaches. As a valid predictive parameter, deletions and mutations of the TP53 gene are associated with a worse prognosis than other types of genetic changes and influence therapeutic decisions. Recently, mutations of MED12 have been added to the list of potential driver mutations in CLL. Across the sub-types, they constitute a relatively rare genetic group affecting roughly 5–9% of CLL patients (Guièze et al., 2015; Kämpjärvi et al., 2015; Wu et al., 2017a). Kämpjärvi et al. (Kämpjärvi et al., 2015) have presented evidence that MED12 mutations may represent a marker of worse prognosis.\n\n\nA closer look at the molecular pathogenesis of uterine leiomyomas\n\nAccording to the high prevalence of uterine leiomyomas MED12 mutations are by far best investigated in this tumor type. Thus, we have now characterized the subset UL affected and described how they can be distinguished from other types of UL.\n\nThere is ample evidence that HMGA2 rearrangements and MED12 mutations occur mutually exclusively in UL and thus constitute independent driver mutations (Markowski et al., 2012). Similarly, somatic MED12 mutations and biallelic Fumarate Hydratase (FH) inactivation occur in mutually exclusive manner in both HLRCC syndrome-associated and sporadic uterine leiomyomas suggesting that the latter constitutes a third small group with an independent molecular pathogenesis (Kämpjärvi et al., 2016). Accordingly, each of these genetic alterations alone as a driver mutation seems to be sufficient to induce the development of an UL without requiring any further mutations.\n\nAs to rearrangements of HMGA2 and mutations of MED12, the transcriptome of tumors of both groups clearly differs with MED12 mutation–positive and HMGA2-overexpressing samples clustering in distinct branches (Mehine et al., 2013). Accordingly, both mutations allow the two major genetic subtypes of UL to be distinguished, and the question arises whether or not the genetic subtypes are also reflected by a different clinical behavior and histopathology.\n\nTumors carrying HMGA2 rearrangements are usually solitary and, on average, of larger size than those with MED12 mutations, also usually presenting as single tumors (Markowski et al., 2014a) whereas the latter are smaller and often co-occur with other clonally independent nodules of the same genetic type (Mäkinen et al., 2011; Markowski et al., 2012; Markowski et al., 2014a). Among the women affected by MED12-mutation positive UL, more than two-thirds had more than one nodule (Heinonen et al., 2017) (Figure 2A) of this type. Even more impressive, in the same study 52 (8.7%) MED12-mutation positive tumors made their appearance as the sole tumor with this mutation, while 547 (91.3%) tumors were associated with at least one other tumor of this genetic subgroup (Figure 2B).\n\nProportion of women carrying a single MED12-mutated UL (52/176) vs. those with more than one such tumor (124/176) (A) and proportion of MED12-mutated UL appearing as single tumor (52/599) vs. those accompanied by at least one other MED12-mutated UL (547/599) (B). Data according to Heinonen et al. (2017).\n\nTo explain the high frequency of MED12 mutated tumors among multiple leiomyomas, Heinonen et al. speculated that \"the multiplicity of MED12-mutation-positive leiomyomas may derive from genetic predisposition and/or environmental factors rendering the myometrium susceptible to selection for MED12 mutations\" (Heinonen et al., 2017). However, as outlined later herein, the association of multiple tumors with MED12 mutations may be a key to the etiology of this type of UL.\n\nFurthermore, MED12-mutated UL are also significantly associated with a subserous location compared to UL lacking this mutation (Heinonen et al., 2017). As to histopathological features, a recent study by Wu et al. revealed that approximately 90% of the cells in HMGA2-rearranged UL were smooth muscle cells showing an overexpression of the protein, while in MED12-mutated UL a similar number of smooth muscle cells and other cells, i.e. mostly tumor-associated fibroblasts, were detected. These latter fibroblasts were lacking MED12-mutations (Wu et al., 2017b) and thus apparently can be classified as by-stander cells. This fits with an earlier observation that in cell cultures of leiomyomas with MED12-mutations a rapid disappearance of mutated cells was seen that became replaced by UL wild-type cells, thus challenging the results of a variety of in vitro experiments on the biology of UL (Bloch et al., 2017; Markowski et al., 2014b).\n\nBesides UL, the occurrence of MED12 mutations has been well-documented in malignant uterine smooth muscle tumors (leiomyosarcomas) and smooth muscle tumors of uncertain malignant potential (STUMP), too (Holzmann et al., 2015; Pérot et al., 2012). Hence, an origin of these tumors from pre-existing UL has been suggested. In contrast, similar cases with HMGA2 rearrangements have not been reported yet.\n\nTo gain further insight into the biology of MED12-mutated UL, Heinonen et al. have undertaken a systematic attempt to check all feasible distinct tumors with a size of 1 cm or larger in diameter from hysterectomy uteri for MED12 mutations (Heinonen et al., 2017). In their study, 599 out of 763 leiomyomas carried MED12 mutations (79%). Next, the data provided by the study of Heinonen et al. (Heinonen et al., 2017) have been used to analyze the number of MED12-mutation positive UL per patient. While it was shown before that in the majority of patients having surgery MED12-mutated tumors do not make their appearance as single nodules but instead are accompanied by other yet clonally independent tumors of this same genetic type (cf. Figure 2), we were interested to see how the number of these tumors per patient is distributed in this series. A slow decrease of the number of MED12-mutation positive tumors is noted (Figure 3). Nevertheless, from these Figures it is not possible to draw conclusions on the overall frequency and distribution of MED12-positive UL in the population because the results are biased by their restriction to symptomatic patients who had undergone surgical treatment. Nevertheless, they correspond more or less to the tumor numbers in general as seen from numerous other studies and thus the question arises if the MED12-positive tumors can be distinguished from the remaining UL as suggested by previous estimations (cf. Figure 2). Thus, we have next investigated if, in the case of multiple tumors, the percentage of MED12-mutation positive UL remains constant independent of the total number of tumors per patient. Surprisingly, it was noted that with a growing number of tumors, the percentage of MED12-mutated tumors clearly increased. Among solitary leiomyomas, only less than 40% of the tumors carried MED12 mutations but their frequency was approaching nearly 100% if twelve or more tumors were present per patient (Figure 4). Thus, in contrast to other mutations, those of MED12 seem to become more likely with an increase of the number of tumors.\n\nAbscissa: number of UL/patient, ordinate: number of patients in the corresponding category. For this diagram data on MED12 alterations published by Heinonen et al. (Heinonen et al., 2017) have been used.\n\nIncreasing percentage of MED12-mutation positive uterine leiomyomas (ordinate) with the number of tumors per patient (abscissa). Open rhombus indicates an interpolated value because no patients with 11 UL were present. For this diagram data on MED12 alterations published by Heinonen et al. (2017) have been used.\n\nAlong with previous data this distribution confirms that, as for its pathogenesis, the occurrence of multiple leiomyomas near exclusively can be attributed to just one genetic mechanism, i.e. MED12 mutations. This contradicts a statement by Mehine et al. that a shared clonal origin as a common feature of leiomyomas not carrying MED12 mutations offers one explanation for the common occurrence of multiple concurrent lesions (Mehine et al., 2015). Instead, a multitude of fibroids mainly appears to be a problem almost exclusively restricted to MED12-mutation positive tumors and we thus decided to analyze and compare the different MED12 mutations in more detail.\n\n\nA closer look at the patterns of MED12 mutations seen in various benign and malignant tumors\n\nMED12 mutations occur in uterine smooth muscle tumors, fibroepithelial tumors of the breast and in chronic lymphocytic leukemia. As a rule, only subsets display these genetic abnormalities which range from the predominant alterations in UL to those detected only in a small percentage of cases in CLL. So far, there is no evidence that the tumors affected by MED12 mutations in the three tumor entities described above differ with respect to their types of mutations. It has been speculated that unknown factors favor the occurrence of these particular mutations. To get further ideas on factors favoring them, the patterns of mutations have now been analyzed in more detail and compared between the different tumor entities paying particular attention to the small deletions occurring in exon 2 or at the intron1-exon 2 boundary, respectively.\n\nIn most cases single nucleotide exchanges are found with a clear predominance of those affecting nucleotides 130 and 131 belonging to codon 44. Less frequently other codons are mutated. Besides single nucleotide exchanges, deletions of small segments of the gene with varying sizes as well as indels affecting exon 2 or the intron1-exon 2 boundary are seen in some cases. As a rule, however, the transcript though affected by the deletions remains in frame. As to these latter genomic alterations accounting for roughly 15% of MED12 mutations, we have analyzed the positions of the deleted bases from a variety of papers analyzing UL, fibroepithelial tumors, and CLL. Adding the number of deleted bases per each position reveals an almost symmetric distribution that is clustered around the hotspot of single nucleotide exchanges (Figure 5). In a previous study by our group, the beginning of the MED12 deletions observed in uterine smooth muscle tumors was mostly located within exon 2 but in rare cases also upstream of the splice site within intron 1. Their size mainly ranged between 3 and 36 bp with a clear predominance of 15 and 18 bp (Markowski et al., 2013b). Of note, an analysis of the data provided by Heinonen et al. for UL revealed that as very rare exceptions even larger deletions as well as those residing in exon 1 can occur.\n\nLeft to right: Ideogram of the X-chromosome (commons.wikimedia.org), exon-intron structure of MED12 (NCBI map viewer), and plot depicting frequency of deletions at each position around the preferred site of single nucleotide exchanges (red solid arrows) seen in uterine smooth muscle tumors (blue), fibroepithelial tumors of the breast (green), and chronic lymphocytic leukemia (yellow). Deletions are plotted across all deleted base positions. Minor preferred sites of single nucleotid exchanges within exon 2 are indicated by dashed red arrows. For this diagram data on MED12 deletions from the following articles have been used: (Guièze et al., 2015; Kämpjärvi et al., 2015; Lim et al., 2014; Markowski et al., 2013b; Mishima et al., 2015; Nagasawa et al., 2015; Ng et al., 2015; Pfarr et al., 2015; Yoshida et al., 2015) only those deletions beginning and ending in the displayed region are listed have been considered.\n\nAs to size and position of these deletions, there is also no obvious difference between UL, fibroepithelial tumors of the breast, and CLL (Figure 5). Overall, this pattern of small deletions of various sizes clustered around the hotspot of single nucleotide exchanges resembles the results of genome editing based on targeted double-stranded breaks as for example those resulting from the usage of the CRISPR/Cas9 system (e.g. cf. Paquet et al., 2016).\n\nIf these mutations indeed arise by certain types of repair of site-specific DNA changes, one might expect that many other mutations occur in the target region of MED12. Of these, only the \"active ones\", i.e. those leaving the open reading frame intact and driving tumorigenesis, will lead to a clonal proliferation of their target cell giving rise to an UL whereas cells with other mutations of the hotspot region will remain quiescent or even become apoptotic (Figure 6). Therefore, future studies aimed at the detection of these \"non-driving\" mutations in single cells, especially from patients suffering from a multitude of UL, may be a reasonable attempt. However, from the pattern of nucleotide exchanges and deletions, a commonly affected sequence can be depicted that may be related to the etiology of UL.\n\nModel illustrating the occurrence and selection of MED12 mutations during the course of leiomyoma development. The scheme suggests that of a larger number of MED12-mutations only those associated with a gain of function act as driver mutations giving rise to UL.\n\n\nHypothesis and opinion\n\nMED12 mutations constitute highly frequent driver mutations in uterine leiomyomas and fibroadenomas, i.e. two tumor entities that occur almost exclusively in middle-aged and young women, respectively. In uterine leiomyomas, they even represent the by far most frequent genetic subtype with a clearly preferential occurrence in the case of multiple tumors. This is in sharp contrast to the other main genetic subtype of UL characterized by rearrangements of HMGA2 usually making its appearance in solitary nodules not accompanied by other tumors of the same genetic subtype.\n\nIt seems difficult to explain these findings just by independent random mutations followed by their selection. Nevertheless, additional factors favoring this multitude of tumors carrying the same type of mutations despite their independent clonal origin have remained enigmatic. After myomectomy, such factors may also account for the risk of recurrences that clearly increases with the number of UL that had been removed (Doridot et al., 2001; Fauconnier et al., 2000). Heinonen et al. (2017) have speculated that either genetic predisposition or environmental factors rendering the myometrium susceptible to selection for MED12 mutations may contribute to the multiplicity of MED12-mutation positive tumors. To describe the development of multiple tumors, those two explanations are well-compatible with a model of clonally unrelated nodules that occur successively and are endowed with a different growth rate as depicted in Figure 7A. Another alternative explaining the multiplicity of UL is the occurrence of clonally related nodules with marked genetic evolution as shown by Mehine et al. (Mehine et al., 2015) for UL lacking MED12 mutations and depicted here as Figure 7B. Nevertheless, a variety of studies indicate that most tumors with MED12 mutations are not clonally related and the multiplicity of these lesions thus needs other explanations. In addition to these both models possibly accounting for other genetic subtypes, as a third alternative factors as in particular infectious agents warrant consideration. The infection may lead to a synchronous initiation of multiple clonally independent lesions endowed with a different growth potential (Figure 7C).\n\nThree alternative models explaining the development of multiple uterine leiomyomas are depicted: (A) corresponds to a prima facie model of myomagenesis with multiple independently developing nodules having a different growth rate. (B) Illustrates the development of multiple clonally related tumors from one common predecessor as described by Mehine et al. (Mehine et al., 2015) for MED12-mutation negative UL. (C) offers an alternative explanation for the growth of multiple clonally unrelated nodules with simultaneous initiation but a different growth rate. Open circles indicate single mutated cells of origin affected by driver mutation.\n\nAs possible factors certain types of bacteria e.g. those involved in reproductive tract infections (RTIs) have long been suggested to be a cause of UL (e.g. (Witherspoon & Butler, 1934)). In the United States both reportable RTIs (i.e. chlamydia and gonorrhea) and fibroids disproportionately burden African American women which lead to the conclusion that the growth of fibroids might be triggered by inflammatory infections associated with the RTIs. Nevertheless, when exploring the relationship between self-reported RTIs and fibroid size, number, and total volume Moore et al. did not find strong associations (Moore et al., 2015). In a recent contribution by the same group women seropositive for genital Chlamydia trachomatis were even found to be less likely to have fibroids (Moore et al., 2018). In line with these findings, in the study by Heinonen et al. neither a history of pelvic inflammatory disease (PID) nor of Chlamydia infection was found to be significantly associated with the MED12-type UL while PID turned out to be significantly associated with the occurrence of MED12-wild type UL (p 0.0024) (Heinonen et al., 2017).\n\nAkin to bacteria, viruses have also been suggested to be involved in the development of UL. For example, EBV is known as a factor associated with the development of extra uterine smooth muscle tumors in HIV and post-transplant patients (see e.g. (Miettinen, 2014; Purgina et al., 2011; Ramdial et al., 2011)). However, so far no association between EBV and uterine leiomyomas has been demonstrated. As to another virus of the Herpes group, a recent study failed to reveal a significant association between HSV-2 seropositivity and the presence of fibroids (Moore et al., 2016) and in general not convincing evidence for viruses involved in the pathogenesis of UL has been presented.\n\nWhile infectious diseases as etiological agents of UL repeatedly have been assumed as such the question arises as to how they could act by site-specific targeting the hotspot region of mutations residing within exon 2 of MED12 and which infectious agents are possible candidates. We will herein present the opinion that the interaction between the human DNA and foreign nucleic acids derived from the infection plays a causal role. As an unorthodox hypothesis stimulating further discussions, we would like to advance the hypothesis that the MED12 mutations result from cleavage of R-loop structures. By definition, R-loops are derived from double stranded DNA where one strand forms a stable DNA-RNA hybrid helix whereas the former associated DNA strand remains single-stranded. R-loops with an \"exposed\" stretch of single-stranded DNA can give rise to instability and DNA double-strand breaks (Aguilera & Gómez-González, 2017; Freudenreich, 2018; Su & Freudenreich, 2017). While the hybrid helix is usually composed of DNA with endogenous RNAs it seems possible that such helices can be formed with foreign RNA as well. To this end, it has been hypothesized that circulating exogenous RNA sequences after their uptake may influence the function of cells through miRNA-like mechanisms (Wang et al., 2012) suggesting direct influences of these sequences of the microbiome on its host´s cells.\n\nTo search for possible sequence homologies we have depicted a target region 5´TGTAAAACAAGGTTTCAATAAC3´ covering 10 nucleotides upstream and downstream each of the two c. 130 and c. 131 (GG), respectively (cf. Figure 6). Of the resulting list with at least 15 identical nucleotides, a variety of human pathogens have been identified as e.g. Bacteroides fragilis (nt 2-20), Klebsiella pneumoniae (nt 2-20), Escherichia coli (nt 3-20), Vibrio vulnificus (nt 6-22), Staphylococcus aureus (nt 6-22, and nt 1-16, respectively), Staphylococcus argenteus (nt 6-22), and Clostridium botulinum (nt 3-19). When searching for abundantly expressed sequences an interesting candidate emerged. A 16-base pair sequence identical to the sense strand of the sequence of a Staphylococcus aureus 27-tRNA gene cluster immediately 3’ to an rRNA operon (Green & Vold, 1993) was noted. The homology covers a palindromic sequence which may act as a terminator of transcription (Green & Vold, 1993) and may also lead to the formation of a hairpin structure stabilizing the RNA molecule (Figure 8). Are these molecules likely to exist as circulating RNAs? Staphylococcus aureus belongs to the phylum of Firmicutes in general constituting the third most abundant sequence population in human plasma with a significant number of the reads mapping to various bacterial ribosomal RNAs and tRNAs (Wang et al., 2012). More specifically, evidence has been presented that RNA species from Staphylococcus are commonly present in blood (Leung & Wu, 2015).\n\nUpper part: Homology of the human MED12 hotspot region (red arrow) with the sense strand of the sequence of the Staphylococcus aureus 27-tRNA gene cluster immediately 3’ to a rRNA operon (Green & Vold, 1993). Blue dashed arrows indicate a palindromic sequence. Sequence from Staphylococcus aureus strain CFSAN007847 chromosome, complete genome; GenBank: CP017684.1; GenBank: FASTA; NCBI Blast. Lower part: bacterial tRNA and rRNA genes of the operon adjacent to the site of homology are shown in blue.\n\nIn summary, deduced from the types and patterns of MED12 mutations in human tumors we have presented evidence supporting the idea that nucleic acid sequences of the human microbiome may interact with the common hotspot of MED12 mutations. To stimulate further discussion a possible interaction of a sequence of Staphylococcus aureus with this hotspot has been considered in more detail as depicted in Figure 8 and Figure 9. As to the initial stage of leiomyoma development, the clearance of R-loops resulting from a hybrid helix between a human target cell and bacterial RNA may simultaneously give rise to multiple clonally independent cells of origin. Additional factors such as the site of origin, angiogenetic support, and the type of MED12 mutation, may endow the resulting monoclonal lesions with a different growth potential.\n\nAs an example for the putative sequence of the human microbiome inducing site specific mutations this scheme refers to Figure 8.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author declared that no grants were involved in supporting this work.\n\n\nReferences\n\nAguilera A, Gómez-González B: DNA-RNA hybrids: the risks of DNA breakage during transcription. Nat Struct Mol Biol. 2017; 24(5): 439–443. PubMed Abstract \| Publisher Full Text\n\nBayley JP, Launonen V, Tomlinson IP: The FH mutation database: an online database of fumarate hydratase mutations involved in the MCUL (HLRCC) tumor syndrome and congenital fumarase deficiency. BMC Med Genet. 2008; 9: 20. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBloch J, Holzmann C, Koczan D, et al.: Factors affecting the loss of MED12-mutated leiomyoma cells during in vitro growth. Oncotarget. 2017; 8(21): 34762–34772. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDoridot V, Dubuisson JB, Chapron C, et al.: Recurrence of leiomyomata after laparoscopic myomectomy. J Am Assoc Gynecol Laparosc. 2001; 8(4): 495–500. PubMed Abstract \| Publisher Full Text\n\nElmlund H, Baraznenok V, Lindahl M, et al.: The cyclin-dependent kinase 8 module sterically blocks Mediator interactions with RNA polymerase II. Proc Natl Acad Sci U S A. 2006; 103(43): 15788–15793. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFauconnier A, Chapron C, Babaki-Fard K, et al.: Recurrence of leiomyomata after myomectomy. Hum Reprod Update. 2000; 6(6): 595–602. PubMed Abstract \| Publisher Full Text\n\nFornaciari G, Giuffra V: Soft tissue tumors in palaeopathology: a review. Pathobiology. 2012; 79(5): 257–267. PubMed Abstract \| Publisher Full Text\n\nFreudenreich CH: R-loops: targets for nuclease cleavage and repeat instability. Curr Genet. 2018; 1–6. PubMed Abstract \| Publisher Full Text\n\nGreen CJ, Vold BS: Staphylococcus aureus has clustered tRNA genes. J Bacteriol. 1993; 175(16): 5091–5096. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGuièze R, Robbe P, Clifford R, et al.: Presence of multiple recurrent mutations confers poor trial outcome of relapsed/refractory CLL. Blood. 2015; 126(18): 2110–2117. PubMed Abstract \| Publisher Full Text\n\nHeinonen HR, Pasanen A, Heikinheimo O, et al.: Multiple clinical characteristics separate MED12-mutation-positive and -negative uterine leiomyomas. Sci Rep. 2017; 7(1): 1015. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHolzmann C, Markowski DN, von Leffern I, et al.: Patterns of Chromosomal Abnormalities that Can Improve Diagnosis of Uterine Smooth Muscle Tumors. Anticancer Res. 2015; 35(12): 6445–6456. PubMed Abstract\n\nKämpjärvi K, Mäkinen N, Kilpivaara O, et al.: Somatic MED12 mutations in uterine leiomyosarcoma and colorectal cancer. Br J Cancer. 2012; 107(10): 1761–1765. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKämpjärvi K, Järvinen TM, Heikkinen T, et al.: Somatic MED12 mutations are associated with poor prognosis markers in chronic lymphocytic leukemia. Oncotarget. 2015; 6(3): 1884–1888. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKämpjärvi K, Mäkinen N, Mehine M, et al.: MED12 mutations and FH inactivation are mutually exclusive in uterine leiomyomas. Br J Cancer. 2016; 114(12): 1405–1411. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKandoth C, McLellan MD, Vandin F, et al.: Mutational landscape and significance across 12 major cancer types. Nature. 2013; 502(7471): 333–339. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLaughlin SK, Schroeder JC, Baird DD: New directions in the epidemiology of uterine fibroids. Semin Reprod Med. 2010; 28(3): 204–217. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLeung RK, Wu YK: Circulating microbial RNA and health. Sci Rep. 2015; 5: 16814. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLim WK, Ong CK, Tan J, et al.: Exome sequencing identifies highly recurrent MED12 somatic mutations in breast fibroadenoma. Nat Genet. 2014; 46(8): 877–880. PubMed Abstract \| Publisher Full Text\n\nMäkinen N, Mehine M, Tolvanen J, et al.: MED12, the mediator complex subunit 12 gene, is mutated at high frequency in uterine leiomyomas. Science. 2011; 334(6053): 252–255. PubMed Abstract \| Publisher Full Text\n\nMarkowski DN, Bartnitzke S, Löning T, et al.: MED12 mutations in uterine fibroids--their relationship to cytogenetic subgroups. Int J Cancer. 2012; 131(7): 1528–1536. PubMed Abstract \| Publisher Full Text\n\nMarkowski DN, Huhle S, Nimzyk R, et al.: MED12 mutations occurring in benign and malignant mammalian smooth muscle tumors. Genes Chromosomes Cancer. 2013a; 52(3): 297–304. PubMed Abstract \| Publisher Full Text\n\nMarkowski DN, Nimzyk R, Belge G, et al.: Molecular topography of the MED12-deleted region in smooth muscle tumors: a possible link between non-B DNA structures and hypermutability. Mol Cytogenet. 2013b; 6(1): 23. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMarkowski DN, Helmke BM, Bartnitzke S, et al.: Uterine fibroids: do we deal with more than one disease? Int J Gyn Path. in press. 2014a; 33(6): 568–572. PubMed Abstract \| Publisher Full Text\n\nMarkowski DN, Tadayyon M, Bartnitzke S, et al.: Cell cultures in uterine leiomyomas: rapid disappearance of cells carrying MED12 mutations. Genes Chromosomes Cancer. 2014b; 53(4): 317–323. PubMed Abstract \| Publisher Full Text\n\nMashal RD, Fejzo ML, Friedman AJ, et al.: Analysis of androgen receptor DNA reveals the independent clonal origins of uterine leiomyomata and the secondary nature of cytogenetic aberrations in the development of leiomyomata. Genes Chromosomes Cancer. 1994; 11(1): 1–6. PubMed Abstract \| Publisher Full Text\n\nMcGuire MM, Yatsenko A, Hoffner L, et al.: Whole exome sequencing in a random sample of North American women with leiomyomas identifies MED12 mutations in majority of uterine leiomyomas. PLoS One. 2012; 7(3): e33251. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMcWilliams MM, Chennathukuzhi VM: Recent Advances in Uterine Fibroid Etiology. Semin Reprod Med. 2017; 35(2): 181–189. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMehine M, Kaasinen E, Mäkinen N, et al.: Characterization of uterine leiomyomas by whole-genome sequencing. N Engl J Med. 2013; 369(1): 43–53. PubMed Abstract \| Publisher Full Text\n\nMehine M, Heinonen HR, Sarvilinna N, et al.: Clonally related uterine leiomyomas are common and display branched tumor evolution. Hum Mol Genet. 2015; 24(15): 4407–4416. PubMed Abstract \| Publisher Full Text\n\nMiettinen M: Smooth muscle tumors of soft tissue and non-uterine viscera: biology and prognosis. Mod Pathol. 2014; 27(Suppl 1): S17–29. PubMed Abstract \| Publisher Full Text\n\nMishima C, Kagara N, Tanei T, et al.: Mutational analysis of MED12 in fibroadenomas and phyllodes tumors of the breast by means of targeted next-generation sequencing. Breast Cancer Res Treat. 2015; 152(2): 305–312. PubMed Abstract \| Publisher Full Text\n\nMoore KR, Cole SR, Dittmer DP, et al.: Self-Reported Reproductive Tract Infections and Ultrasound Diagnosed Uterine Fibroids in African-American Women. J Womens Health (Larchmt). 2015; 24(6): 489–495. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMoore KR, Smith JS, Cole SR, et al.: Herpes Simplex Virus Type 2 Seroprevalence and Ultrasound-Diagnosed Uterine Fibroids in a Large Population of Young African-American Women. Am J Epidemiol. 2016; 183(11): 961–968. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMoore KR, Smith JS, Cole SR, et al.: Chlamydia trachomatis Seroprevalence and Ultrasound-Diagnosed Uterine Fibroids in a Large Population of Young African-American Women. Am J Epidemiol. 2018; 187(2): 278–286. PubMed Abstract \| Publisher Full Text\n\nNagasawa S, Maeda I, Fukuda T, et al.: MED12 exon 2 mutations in phyllodes tumors of the breast. Cancer Med. 2015; 4(7): 1117–1121. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNg CC, Tan J, Ong CK, et al.: MED12 is frequently mutated in breast phyllodes tumours: a study of 112 cases. J Clin Pathol. 2015; 68(9): 685–691. PubMed Abstract \| Publisher Full Text\n\nPaquet D, Kwart D, Chen A, et al.: Efficient introduction of specific homozygous and heterozygous mutations using CRISPR/Cas9. Nature. 2016; 533(7601): 125–129. PubMed Abstract \| Publisher Full Text\n\nPérot G, Croce S, Ribeiro A, et al.: MED12 alterations in both human benign and malignant uterine soft tissue tumors. PLoS One. 2012; 7(6): e40015. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPfarr N, Kriegsmann M, Sinn P, et al.: Distribution of MED12 mutations in fibroadenomas and phyllodes tumors of the breast--implications for tumor biology and pathological diagnosis. Genes Chromosomes Cancer. 2015; 54(7): 444–52. PubMed Abstract \| Publisher Full Text\n\nPiscuoglio S, Murray M, Fusco N, et al.: MED12 somatic mutations in fibroadenomas and phyllodes tumours of the breast. Histopathology. 2015; 67(5): 719–729. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPurgina B, Rao UN, Miettinen M, et al.: AIDS-Related EBV-Associated Smooth Muscle Tumors: A Review of 64 Published Cases. Patholog Res Int. 2011; 2011: 561548. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRamdial PK, Sing Y, Deonarain J, et al.: Extra-uterine myoid tumours in patients with acquired immunodeficiency syndrome: A clinicopathological reappraisal. Histopathology. 2011; 59(6): 1122–1134. PubMed Abstract \| Publisher Full Text\n\nSchoenmakers EF, Wanschura S, Mols R, et al.: Recurrent rearrangements in the high mobility group protein gene, HMGI-C, in benign mesenchymal tumours. Nat Genet. 1995; 10(4): 436–444. PubMed Abstract \| Publisher Full Text\n\nSu XA, Freudenreich CH: Cytosine deamination and base excision repair cause R-loop-induced CAG repeat fragility and instability in Saccharomyces cerevisiae. Proc Natl Acad Sci U S A. 2017; 114(40): E8392–E8401. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVogelstein B, Papadopoulos N, Velculescu VE, et al.: Cancer genome landscapes. Science. 2013; 339(6127): 1546–1558. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWang K, Li H, Yuan Y, et al.: The complex exogenous RNA spectra in human plasma: an interface with human gut biota? PLoS One. 2012; 7(12): e51009. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWilliams ARW: Uterine fibroids – what’s new? [version 1; referees: 3 approved]. F1000Res. 2017; 6(F1000 Faculty Rev): 2109. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWitherspoon JT, Butler VW: The etiology of uterine fibroids with special reference to the frequency of their occurrence in the Negro: an hypothesis. Surg Gynecol Obstet. 1934; 58: 57–61.\n\nWu B, Słabicki M, Sellner L, et al.: MED12 mutations and NOTCH signalling in chronic lymphocytic leukaemia. Br J Haematol. 2017a; 179(3): 421–429. PubMed Abstract \| Publisher Full Text\n\nWu X, Serna VA, Thomas J, et al.: Subtype-Specific Tumor-Associated Fibroblasts Contribute to the Pathogenesis of Uterine Leiomyoma. Cancer Res. 2017b; 77(24): 6891–6901. PubMed Abstract \| Publisher Full Text\n\nYoshida M, Sekine S, Ogawa R, et al.: Frequent MED12 mutations in phyllodes tumours of the breast. Br J Cancer. 2015; 112(10): 1703–1708. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "33313", "date": "30 Apr 2018", "name": "Takeshi Kurita", "expertise": [ "Reviewer Expertise Biology of Uterine Lieomyoma" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis Opinion Article proposes a provocative idea that transcripts of Staphylococcus aureus contribute to an exceptionally high incidence of MED12 exon 2 mutations among human tumors. After providing an excellent summation of the current literature on demographics and characteristics of MED12 mutations in human neoplasms, the authors propose an intriguing concept: the involvement of mutagenic nucleic acids of bacterial origin in human tumors, based on the discovery of a Staphylococcus genome sequence of which the transcript may form an R-loop with human MED12 gene on the mutation hotspot. This novel concept merits further discussion among researchers of diverse biomedical fields.\n\nThere are several issues that I recommend the authors to address.\nI. The importance of the idea would increase if the following issues are resolved.\nThe methods of database search for homologous sequences and the statistical significance of the finding should be described. In other words, it is unclear how frequently such homologous sequences to the human genome appear in the genome of pathogenic microorganisms, and what proportion of such homologous sequences can form an R-loop with human genome if they are transcribed.\n\nThe palindromic sequence is located at 30 bp from the 3’ end of tRNA-Leu on the reverse strand in Staphylococcus aureus genome. Since it is not a typical template for transcription, inclusion of references that suggest transcriptional activity in such genomic regions of bacteria would help further discussion by increasing the feasibility of the proposed model. It would also be helpful to include a discussion on the possible mechanisms through which naked transcripts of low-copy number in the circulation could possibly reach the target locus in the genome of myometrial cells. For instance, internalization of shRNA for RNAi is achieved by viral transduction or chemical/physical transfection of shRNA expression vectors. The authors may speculate the mechanisms that unfold the minimum free energy structure of short hairpin transcripts and facilitate the hybridization to the genomic DNA. For example, shRNA is processed into siRNA, and then siRNA forms a complex with cellular proteins to elicit RNAi effect.\n\nII. There are issues in data presentation\nRegarding the title of Figure 2: “Single and multiple MED12-mutation positive uterine leiomyomas”, the adjectives “single and multiple” could modify “MED12-mutation”. Adding a hyphen (MED12-mutation-positive) and using of “solitary” instead of “single” would improve the readability. Error: The labels for single and multiple UL groups in Figure 2 are switched. The graphs in Figures 2, 3 and 4 are the same data in different presentation format. The Figures 3 and 4 seem to be redundant.\nIII. There are other factors that also likely contribute to the high prevalence of patients with multiple MED12 mutant ULs.\nA single hit on the active MED12 allele is sufficient for the pathogenesis of ULs. Human cancer cells usually carry multiple putative driver mutations even at the earliest stage. In contrast, most MED12 mutant ULs do not carry additional mutations, suggesting a MED12 mutation is sufficient to drive UL pathogenesis. Since the MED12 is on X-chromosome, a single hit on the active allele of MED12 has dominant effect. Hence, the development of UL through MED12 mutations should occur at a significantly higher rate compared to the neoplasms that require multiple genetic lesions for pathogenesis. Diversity of pathogenic MED12 mutations increases the prevalence of MED12-mutant ULs. Generally, driver missense mutations of human neoplasms are very specific. For instance, nearly all adult-type granulosa cell tumors carry FOXL2 c.402C>G (C134W) missense mutation. While single base replacements in the hotspot triplet bases can also result in conversion of C134 to F, S, Y, R and G, only C134W is pathogenic. In contrast, a variety of MED12 mutations, ≥ 10 missense and > 30 indel mutations, are associated with ULs. For instance, missense mutations that convert MED12 G44 to D, S, V, R, C and A are all pathogenic. Hence, even if the mutation rate per nucleotide is equal throughout the human genome, the incidence of pathogenic MED12 mutations should be many times higher than other pathogenic mutations. UL is a hormone dependent tumor. Another key factor that this Opinion Article does not address is the systemic hormonal environment. The pathogenesis of UL depends on estrogen and progesterone. Since ULs are counted only when they grow to a grossly recognizable size, patients with endocrine profiles favorable to the growth of ULs should have a higher number of tumors even if the incidence of pathogenic mutations in myometrial cells is equal among all women.\n\nDiscussion of these factors is not essential, as the model proposed in this article would work independently. Nevertheless, these additions would help balance the discussion.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Partly\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Partly", "responses": [] }, { "id": "34711", "date": "19 Jun 2018", "name": "Eric Glasgow", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis opinion article by Bullerdiek et al. reviews the types and etiologies of tumors associated with MED12, and put forward a hypothesis regarding the mechanism of MED12 mutation. MED12-associated tumors are understudied considering their high frequency, probably because they are most commonly associated with benign tumors. However, a better understanding of the etiology of these tumors would likely contribute substantially to our knowledge of the genetic mechanisms of tumorigenesis. The authors discuss the main types of tumors caused by MED12 mutations: uterine leiomyomas (ULs), fibroadenomas of the breast, and chronic lymphocytic leukemias. They point out that the characteristics of ULs associated with MED12 mutations differ markedly from those of ULs with other genetic causes. Within the same patient, multiple tumors with MED12 mutations of independent clonal origin are common. Bullerdiek et al. illustrate the presence of a mutational hotspot within the MED12 sequence with homology to tRNAs of multiple human pathogens. These qualities are suggestive of a genome editing mechanism warranting further investigation.\n\nImportant changes:\n\n-The gene name! The gene name is mediator complex subunit 12 (MED12), not Mediator Subcomplex 12 as stated in the article.\n\n-In the Abstract it is stated “(MED12)…presumed classification as gain-of-function mutations…”, and in the Introduction “…deletions and, more rarely, indels, usually affecting exon 2 or the intron 1/exon 2 boundary are found which always leave the reading frame intact indicating gain-of-function mutations”. The term gain-of-function should be removed. The function of MED12 is unknown and there is no evidence suggesting that these mutations are gain-of-function.\n\nMajor issues:\n\n-In Figure 2B, it is unclear how many patients in total were analyzed for tumor counts or whether the multiple MED12-mutation positive UL were concentrated in a small number of patients or spread evenly across patients analyzed.\n\n-Additionally, the labels appear to be switched in the key of Figure 2.\n\n-A major problem is that the authors claim that the occurrence of multiple ULs can be exclusively attributed to MED12 mutation. However, although the authors provide ample quantitative evidence of multiple ULs resulting from MED12 mutation within the same patient, the authors show no data regarding the occurrence of multiple ULs with other genetic mechanisms (i.e. HMGA2 arrangements). In order to make the claim that multiple UL is a feature unique to MED12-mutated tumors, the authors should provide data on non-MED12-mutated multiple UL for the sake of comparison. If the number of non-MED12-mutated multiple UL is zero, this should be made clear within the text of the article.\n\n-In reference to Figure 5, it is unclear whether the CLL mutations (yellow) cluster around the mutational hotspot as do the UL and fibroadenoma mutations, although the text of the article claims that the CLL mutations also cluster around the hotspot.\n\n-The authors point out that the evidence for a connection between infection and UL is dubious thus far, yet they go on to propose a mechanism for the role of infection in UL pathogenesis. To make a convincing argument, the authors must clarify why they still consider a pathogenic mechanism worth considering. -The authors should provide more detail about the pathogens with tRNA sequences matching the mutational hotspot. What is known about the involvement of these pathogens in pelvic infections? Does the prevalence of pelvic infections involving these pathogens approximately match the prevalence of UL? Is there any existing evidence for association between these pathogens and UL?\n\nMinor issues:\n\n-Second paragraph of the introduction: “As in the benign tumors, however, mutations of that gene occur as apparent driver mutations in a predominant subset of uterine leiomyomas (Mäkinen et al., 2011; Markowski et al., 2012; McGuire et al., 2012), constituting the by far most frequent human symptomatic tumors of all.” Do the authors mean “As in the malignant tumors...” Also, “that gene” is vague. Replace with MED12.\n\n-Second paragraph introduction: “Furthermore, the same type of MED12 mutations was found in two canine vaginal leiomyomas (Markowski et al., 2013a)” can be changed to “Furthermore, the same type of MED12 mutation was found in two canine vaginal leiomyomas (Markowski et al., 2013a)”\n\n-Third paragraph introduction: “Predominantly, the mutations are clustered in the 5 ́ region of exon 2 of the gene with only a few mutations affecting the intron 1-exon 2 boundary or, much rarer, exon 1 or the exon 1-intron 1 boundary” can be changed to: “Predominantly, the mutations are clustered in the 5 ́ region of exon 2 of the gene with only a few mutations affecting the intron 1-exon 2 boundary or, more rarely, exon 1 or the exon 1-intron 1 boundary”\n\n-The heading “Introducing three tumor entities displaying a unique type of MED12 mutations” should be changed to “Introducing three tumor entities displaying MED12 mutations”\n\n-Under subheading: Uterine Leiomyomas – the most frequent symptomatic human tumors, “Depending on their location it can be distinguished between submucosal, intramural, and subserosal UL” can be changed to “Depending on the location, the difference between submucosal, intramural, and subserosal UL can be distinguished”\n\n-The subheading “Fibroadenomas of the breast - Frequent benign tumors of adolescent and young women” can be changed to “Fibroadenomas of the breast - Frequent benign tumors in adolescent and young women” and the subheading “Chronic lymphocytic leukemias - most frequent leukemia of the adults” can be changed to “Chronic lymphocytic leukemias - most frequent leukemia in adults”\n\n-Under the heading “A closer look at the molecular pathogenesis of uterine leiomyomas,” the sentence, “According to the high prevalence of uterine leiomyomas MED12 mutations are by far best investigated in this tumor type” can be changed to “MED12 mutations are by far best investigated uterine leiomyoma tumor type.”\n\n-Second paragraph under the subheading: “Leiomyomas with MED12 mutation constitute their own genetic subtype which is also characterized by a distinct clinical and histopathological appearance,” the sentence “Accordingly, both mutations allow the two major genetic subtypes of UL to be distinguished, and the question arises whether or not the genetic subtypes are also reflected by a different clinical behavior and histopathology” can be changed to “Accordingly, these mutations allow the two major genetic subtypes of UL to be distinguished, and the question arises whether or not the genetic subtypes are also have different clinical behaviors and histopathologies”\n\n-The second paragraph under the subheading: “The percentage of MED12-mutated tumors is positively correlated with the total number of tumors per patient,” the sentence “Along with previous data this distribution confirms that, as for its pathogenesis, the occurrence of multiple leiomyomas near exclusively can be attributed to just one genetic mechanism, i.e. MED12 mutations” can be changed to: “Along with previous data this distribution confirms that the occurrence of multiple leiomyomas nearby can be exclusively attributed to just one genetic mechanism, i.e. MED12 mutations.”\n\n-At the end of the first paragraph under the heading: “A closer look at the patterns of MED12 mutations seen in various benign and malignant tumors” the word respectively is unnecessary.\n\n-Under “Hypothesis and Opinion” the sentence “Nevertheless, additional factors favoring this multitude of tumors carrying the same type of mutations despite their independent clonal origin have remained enigmatic” should be changed to “Nevertheless, additional factors favoring this multitude of tumors with independent clonal origin carrying the same type of mutation have remained enigmatic.” -Also, “In addition to these both models possibly accounting for other genetic subtypes, as a third alternative factors as in particular infectious agents warrant consideration” should be changed to “In addition to these models, the potential roles of infectious agents warrant consideration” -Also, “As to another virus of the Herpes group, a recent study failed to reveal a significant association between HSV-2 seropositivity and the presence of fibroids (Moore et al., 2016) and in general not convincing evidence for viruses involved in the pathogenesis of UL has been presented” should be changed to “As to another virus of the Herpes group, a recent study failed to reveal a significant association between HSV-2 seropositivity and the presence of fibroids (Moore et al., 2016) and in general no convincing evidence for involvement of viruses in the pathogenesis of UL has been presented”\n\nTo summarize, the authors build a fairly convincing argument for the need to more closely study the mechanism of MED12 mutation in tumorogenesis. However, as detailed above, there are multiple areas in which additional evidence and clarification is needed to support the authors’ claims. Additionally, correction of several grammatical errors and awkward phrasing throughout the manuscript would greatly improve readability.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? Partly\n\nAre arguments sufficiently supported by evidence from the published literature? Partly\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Partly", "responses": [] }, { "id": "36051", "date": "31 Jul 2018", "name": "Jose M. Teixeira", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe opinion article by Bullerdiek and Rommel provides an interesting hypothesis that is partly supported by the unusual role of MED12 in the origin of uterine leiomyoma. MED12 is mutated in these benign tumors at a hotspot location at the 5’ end of exon 2. How these mutations arise in the myometrium and how the mutant MED12 drives leiomyoma development are not known.\n\nI agree with many of the previous reviewers’ comments and will limit my comments to a few items so as not to be too repetitious. For example, I agree that there is no evidence that the MED12 mutations constitute a gain-of-function in the protein. Indeed, if MED12 mutation alone is sufficient for tumorigenesis, it would have been observed in a variety of other tumor types. In fact, no mechanisms for tumorigenesis unique to leiomyomas have been reported. Thus, how MED12 mutation leads to leiomyoma development is not known at this point. This should be changed in the text.\n\nI also agree that the argument for foreign RNA from bacterial infection causing MED12 mutation, although quite novel and certainly interesting, is weak. In addition to the reasons already discussed, ascending infections, such as the reproductive tract infections described by the authors, are likely to rarely involve the myometrium. In the case of S. aureus, it is even more unlikely because myometrial infection with that organism is likely associated with bacteremia, which is even more rare and cannot account for the prevalence of MED12 mutant fibroids in women. Also, although leiomyomas are normally found in reproductive age women, that doesn’t necessarily mean that sexually active women are more prone to the disease because of possible sexually transmission of pathogenic bacteria. Since parity is associated with a lower leiomyoma burden, the argument for reproductive tract infections being the culprit is not supported. Parity is also associated with increased risk of postpartum iatrogenic infection, but again parity is associated with decreased risk for leiomyomas. These caveats should be included in the text.\n\nThere is something special about the very common MED12 mutations in that hotspot and high prevalence of uterine leiomyomas (and the not-so-common fibroadenomas) in reproductive age women. It is possible that a hormonally-regulated factor expressed only in myometrium (and breast stroma) could be interacting with wt MED12 on the site where the hotspot mutations alter the protein structure, and mutation disrupts that interaction. How those MED12 mutations occur and why the tumors/fibroadenomas develop only in those tissues is a mystery that needs to be resolved in order to develop therapies targeting the mechanisms involved.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? Partly\n\nAre arguments sufficiently supported by evidence from the published literature? Partly\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? No", "responses": [] } ]	1	https://f1000research.com/articles/7-359
https://f1000research.com/articles/7-1930/v1	13 Dec 18	{ "type": "Systematic Review", "title": "A systematic review and critical evaluation of inflammatory cytokine associations in hidradenitis suppurativa", "authors": [ "John W. Frew", "Jason E. Hawkes", "James G. Krueger", "Jason E. Hawkes", "James G. Krueger" ], "abstract": "Background: The pathogenesis of hidradenitis suppurativa (HS) remains unclear. In order to develop effective treatment strategies, a deeper understanding of pathophysiology is needed. This is impaired by multiple small studies with inconsistent methodologies and the impact of co-occurring pro-inflammatory conditions such as smoking and obesity. Methods: This systematic review aimed to collate all published reports of cytokine studies in tissue, blood, serum and exudate. It was registered with PROSPERO (Registration number CRD42018104664) performed in line with the PRISMA checklist. Results: 19 studies were identified comprising 564 individual HS patients and 198 control patients examining 81 discrete cytokines. Methodology was highly varied and the quality of studies was generally low. There was a large degree of variance between the measured levels of cytokines. 78.2% of cytokines demonstrated heterogeneity by the chi-squared test for homogeneity and hence meta-analysis was not deemed appropriate. However, a strong and significant IL-17 signalling component was identified. Conclusions: Cytokines consistently elevated in lesional, peri-lesional and unaffected tissue are identified and discussed. Areas for further investigation include the role of dendritic cells in HS; the contribution of obesity, smoking, diabetes and the microbiome to cytokine profiles in HS; and examining the natural history of this disease through longitudinal measurements of cytokines over time.", "keywords": [ "Hidradenitis Suppurativa", "Cytokines", "Inflammation", "Pathogenesis", "IL-17", "TNF-alpha" ], "content": "Introduction\n\nHidradenitis Suppurativa (HS) is a chronic inflammatory disease, the exact pathophysiology of which remains poorly defined1. Dysregulation of the Th17: Treg axis2, IL-36 signalling pathways3 and keratinocyte-mediated inflammatory cytokines4 have been demonstrated in lesional skin, blood, serum, and exudate5–8 although contradictory results exist4,9. Given the variable and incomplete response of patients to treatment, including monoclonal antibodies1, some authors have proposed clinical10,11, and immunological5 subtypes of HS in an effort to better predict treatment outcome and response. Thus far, no current schema accurately predicts treatment efficacy.\n\nIn order to develop and implement effective treatment strategies in HS, a deeper understanding of the underlying inflammatory pathophysiology is needed. However, due to the heterogeneity of sampling methods, laboratory processing methods and data analysis, comparison across studies is problematic and potentially biased or inaccruate12. Heterogeneity of tissue sampling and laboratory techniques alone may explain the inconsistent and conflicting results regarding specific cytokines,4,9 however, no systematic analysis of cytokine studies has been undertaken to compare results, methodology, and analytical techniques.\n\nAn additional complicating factor is that clinical comorbidities, which are strongly associated with disease activity in HS, such as obesity13, diabetes14, inflammatory bowel disease15, and smoking16, also produce pro-inflammatory cytokines, which affect multiple organ systems including the skin15,17–19. Hence, it remains unclear whether the presence or absence of these conditions confound the findings of cytokine studies in HS, and whether clinical stratification of patients is necessary to identify significant pathogenic pathways, which may be amenable to pharmacological intervention. Critical evaluation and analysis of existing studies may also enable meta-analysis, which may identify cytokines, which, in smaller studies, do not have sufficient power to meet statistical significance when compared to controls.\n\n\nObjectives\n\nThe objectives of this systematic review are:\n\n1) To collate and describe all published reports of human cytokine studies in HS including those in skin, blood, serum and exudate.\n\n2) To critically evaluate the sampling, laboratory and analysis techniques used in each study to assess whether comparisons can be made across individual studies.\n\n3) To analyze the heterogeneity of published studies enable meta-analysis\n\n\nMethods\n\nThis systematic review was registered with PROSPERO20 (Registration number CRD42018104664) and was conducted in line with the PRISMA checklist21\n\nInformation sources for this review included PubMed (1946-July 1 2018), Scopus (2004- July 1 2018) and Web of Science (1990-July 1 2018) as shown in Figure 1. Search strategy is presented in Table 1\n\nEligibility criteria for this review included cohort studies, case-control studies and other observational studies with no restrictions of patient age, sex, ethnicity or language of publication. Eligible studies included:\n\n1) Studies reporting the results of cytokine investigations (in cutaneous tissue, serum, blood or exudate) in human subjects clinically diagnosed with hidradenitis suppurativa.\n\nStudies deemed not eligible included those which:\n\n1) Provide no new data but a review or summary of previously published data\n\n2) Provide no comparison with controls or non-lesional tissue\n\nData collection was performed independently by 2 authors (JWF & JEH), with any disagreements regarding inclusion of citations being referred to a third author (JGK) for mediation. Information was collected using a standardized data collection form (available as Extended data22) with the principal outcomes of interest being the cytokine of interest, measured level of cytokine in lesional HS skin or serum. Comparison data against either peri-lesional, unaffected or control skin or serum was also collated. If data from individual patients was not available then the aggregate data including average change and statistical analyses of the significance of change was collected.\n\nFor each individual cytokine, where more than one study reported results, heterogeneity was assessed using the chi-squared tests for homogeneity. Homogeneity was defined as a chi squared value >0.05. All statistical analysis was undertaken using R (version 3.5.1)\n\nPotential sources of bias in the identified studies are acknowledged including the small size of patient cohorts, the variability in sampling, laboratory techniques and the inclusion of patients being treated with a wide-variety of medications including immunosuppressants. Bias was also assessed using the NIH quality assessment tool for observational studies23.\n\n\nResults\n\nA total of 367 non-duplicated citations were identified in the literature review (Figure 1). 343 of these articles were removed upon review of titles and abstracts against the pre-defined eligibility criteria. Full text review of the remaining 24 articles excluded 5 review articles providing no new data. The remaining 19 studies2–9,24–33 included the results of 564 individual HS patients and 198 control patients, which were included in this systematic review.\n\nThe summarized demographic data of the patients and controls comprising this review are included in Table 2. The 564 reported cases comprised of 231 males (40.9% reported cases) and 333 females (59.0%). 24 cases were unreported (4.1%). The average age was 38.5 years (n=560, 18 cases unreported). 141 individuals were current smokers (82.4% reported cases), 8 ex-smokers (4.7% reported cases), 22 non-smokers (12.8% reported cases) and 407 unreported. Obesity (BMI>30) was reported in 85 individuals (42.5% reported cases), with 115 (57.5%) individuals non-obese (BMI<30) and unreported in 378 cases. 8 cases reported diabetes mellitus out of 24 reports (33% of reported cases). 12/38 cases reported a positive family history of HS (31.6% reported cases). Hurley Stage was reported as stage 1 in 68 individuals (17.4% reported), stage 2 in 199 individuals (51% reported cases) and stage 3 in 123 individuals (31.6% reported cases) with 188 cases going unreported. The average mHSS (modified hidradenitis suppurativa score) was 78.1 (n=247 cases). Biopsies were largely taken from the axillae (n=32, 43.8%) and groin (n=35, 48.0%), with a minority of samples being taken from the genital and perianal region (n=6, 8.2%). At the time of sampling patients were on treatment including Clindamycin+ Rifampicin (n=18); adalimumab (n=26); Metformin (n=2); levothyroxine (n=1); MABp1 (n=10); tetracyclines (n=12) Infliximab (n=2); other antibiotics (n=4). Treatment was not specified in 74 cases, with no treatment in 86 individuals and treatment withheld in 85 patients.\n\nBMI= Body Mass Index mHSS= modified Hidradenitis Suppurativa Score (Sartorius Score) NR= Not Reported SD= Standard Deviation Y= Yes N=No Ex= Ex Smoker\n\nOnly 5/19 (26.3%) studies analysed both lesional tissue and serum levels of cytokines, enabling direct comparison between these two compartments. 8/19 (42.1%) studies provided age and sex matched controls, 5/15 (33.3%) studies stratified by disease severity and no studies stratified by lesion site or comorbidities. 8/19 (42.1%) studies stratified or accounted for treatment or reported discontinuing treatment up to 3 weeks prior to sample collection (Table 3).\n\nTable 2: Critical Evaluation of Methodology of Studies Included in This Review Key:L= Lesional, PL= Perilesional, U= Uninvolved, C= Control S=Serum, Y=Yes, N=No, NR= Not Reported,\n\nA total of 81 discrete cytokines were analysed over the 19 studies (presented in Table 4). 6 studies provided a total of 78 outcomes from tissue of lesional or peri-lesional biopsies, 4 studies provided a total of 30 results from serum analysis and 1 study provided 15 results from exudate analysis. The remaining 8 studies did not provide quantification of cytokine levels but did provide analysis of the change and significance between lesion and control samples. The degree of change between lesional and control samples varied widely from 1.5 times the control level (IL-1RA p=0.0112) to 149 times the control level (IL-17 p<0.05). 33 cytokines were evaluated in more than one study. Only IL-1β, IL-6, IL-8, IL-17A and TNF-α had data from 5 or more separate studies.\n\nKey: L= Lesional ; PL= Perilesional; C= Control; NS= Not Significant ; HSs= HS Serum; Cs= Control Serum; HSe= HS Exudate; Ce= Control Exudate; I = Inflamed lesional skin, S= Scarred lesional skin, #= Vs CAMP, = NT (Non-Treated) Samples ,* = Stimulation by Pam2CSK4 Lipopeptide,*** Stimulation by Muramyl Dipeptide (MDP), + Heat Killed Candida Albicans; ++ Heat Killed Staph Aureus, +++ Lipopolysaccharide;\n\nCytokines and inflammatory proteins which were elevated in more than one study in lesional tissue included IL-1β, IL-6R, IL-10, IL-17A, IL-36α, IL-36β, IL-36γ, IL-36RA, TNF-α, sTNFR2, hBD1, hBD2, hBD3, s100A7, LL37/Cathelicidin, CCL3, CCL5, CCL27 and BLC. Cytokines and inflammatory proteins elevated in peri-lesional tissue included IL-1β, IL-17, IL-36β, IL-36RA, IL-37, IL-38 and TNF-α. IL-37 was the only cytokine identified which showed significant differences between lesional and peri-lesional tissue, with a 1.81 times elevation in lesional compared to peri-lesional tissue (p=0.0002)3. IL-17 was elevated in unaffected HS tissue compared to control patient tissue (p<0.05) in one study31. In HS tissue, S100A9, hBD1 and hBD2 were reduced but this data did not meet statistical significance. Two studies measuring IL-1β levels showed no statistically significant difference between lesional and control skin7,25. No significant elevation of IL-6 was seen in lesional tissue compared to control with the exception of 1 study25. IL-8 levels only just made significance in two studies5,7, with one study showing significant elevation of IL-8 in lesional compared to control tissue24. Two additional studies showed no significant difference4,8. TNF-α levels were significantly elevated compared to control tissue in two studies7,31 but not significantly in 2 additional studies4,24. sTNFR1 was significantly elevated in one study26 whilst showing a non-significant difference in a second study25. CCL5 was significant in 2 studies in lesional tissue compared with controls4,26. One methodology using muramyl dipeptide (MDP) did not reach statistical significance compared to stimulation with Pam2CSK4 Lipopeptide, and non-treated (NT) cells. IFN-γ was elevated in lesional tissue with no significance in one study28 and significance in another4.\n\nElevated cytokines and inflammatory proteins in HS serum included IL-1β, IL-6, IL-8, IL-10, IL-12p70, IL-17, TNF-α, sTNFR1, CRP, ESR, LC2, and MMP2. TNF-β, and IFN-γ were elevated in wound exudate from active HS lesions. IFN-γ was noted to be decreased in HS patient serum compared to healthy control serum, despite the elevation in wound exudate. Conflicting results were seen in serum findings in IL-10, IL-17 and IFN-γ. One study demonstrated elevated serum IL-10 levels compared to control5 whereas two other studies8,27 showed no significant difference. Whilst two studies4,5 illustrated elevated IL-17 Serum levels in HS patients, one study7 showed no significant difference between patients and controls. IFN-γ showed no statistically significant decrease in the serum of HS patients compared to control in one study9 but a significant difference in a larger, higher powered study4.\n\nBecause adalimumab improves HS through TNF antagonism1,2, this cytokine must be classified as pathogenic. TNF mediates inflammation in a classic “sepsis” cascade in tissues—in this pathway LPS from gram negative bacteria activates TNF release from cells, and then TNF stimulates production of IL-1b, IL-6, and IL-8, leading to neutrophil attraction into sites of infection2,4. Increases in IL-1β and IL-8 measured in HS, as well as neutrophil accumulation, could result from this pathway. Alternatively, in psoriasis, TNF is a major cytokine that acts on the IL-23/Type 17 T-cell pathway at two points. First TNF induces IL-23 synthesis in myeloid (CD11c+) dendritic cells in the skin34. Second, TNF (as well as other cytokines that also activate NF-kB) act synergistically with IL-17A or IL-17F to increase synthesis of many other cytokines, chemokines, and inflammatory molecules in keratinocytes and other cell types. There are several clues that an IL-23/Type17 T-cell pathway may be active in HS which include detection of Th17 T-cells in skin infiltrates, increased production of IL-17A, and increased production of LL-37/cathlecidin, S100A7, S100A8, S100A9, LCN2, IL-8, beta-defensins and IL-36; which are all molecules induced by IL-17 in keratinocytes, as also the presence of psoriasis-like epidermal hyperplasia in some reports. The increased production of CCL204, would be predicted to increase tissue infiltration of both Th17 T-cells and CD11c+ DCs, which have both been observed in HS, and increased production of TGF-β could increase differentiation of Th17 T-cells from precursors and/or influence scarring in skin lesions. If IL-17 is driving inflammation in HS, one would expect to see increased production of additional chemokines that regulate neutrophil chemoattraction (CXCL1, CXCL2, CXCL3). Epidermal hyperplasia is not presently explained in HS, but this could be related potentially to increased expression of IL-19, IL-20 or IL-22, which are associated with the IL-23/Type 17 T-cell axis. If IL-22 is produced in HS lesions, this would implicate Th22 T-cells as a T-cell type also associated with the IL-23/Type 17 T-cell axis. There is an uncertain role for other T-cell subsets in HS. Increased production of CXCL9 and IP-10 (CXCL10) are often linked to production of IFN-γ from Th1 T-cells in inflammatory sites, but IL-26 or IL-29, which are also cytokines produced by Th17 T-cells are alternative activators of STAT1 and CXCL9 production. IL-32 production in HS may also be linked to a T-cell subset that produces this cytokine. Low production of Th2 associated cytokines (IL-4, IL-5, or IL-13) has been measured in HS, suggesting an unlikely role of this T-cell subset. Likewise, the presence and function of T regulatory cells (Tregs) in HS lesions needs further study. IL-10 which is elevated in HS could be produced by either Tregs or the cDC1 (BDCA3+) DC subset, but levels may be inadequate to control tissue inflammation. At present, dendritic cell subsets are also incompletely characterized in HS. Potential sources of IL-12 or IL-23 are CD11c+ DCs, which includes the tissue resident BDCA-1+ (cDC2) subset and less mature inflammatory DCs, which are abundant cells in inflammatory lesions of psoriasis or atopic dermatitis but have not been investigated in HS. Cytokine contributions by other cell types such as innate lymphoid cells, macrophages, mast cells, and other leukocytes also remains to be determined.\n\nThe methodologies of cytokine analysis varied widely (Table 5). 92 results were produced using electrochemical luminescence (ECL) procedures from three separate systems and manufacturers. 62 results were produced using ELISA. 18 results4 were performed with either ELISA or ECL but not further specified. 15 results were produced using polymerase chain reaction (PCR) with three separate systems from three manufacturers. Four discrete cytokines (IL-10, IL-17, TNF-α and IFN-γ) were analysed using all three techniques (ECL, ELISA and PCR), whilst 15 discrete cytokines (IL-6, IL-8, IL12p40, IL-17A, IL-22, IL-23, S100A7, S100A8, S100A9, RNAse7, IP-10, CCL5, CCL20, CCL27) were analysed using ELISA and ECL only. We note IL-17 levels may well be below the lower limit of quantification with ELC and ELISA based approaches, with only the Singulex platform having the ability to quantify levels of IL-17 present in blood and serum of normal subjects.\n\nTable 4: Antibodies Used for Identification of Cytokines in Studies Included in this Systematic Review. ECL: Electrochemicoluminescence\n\nAssessment of bias is presented in Table 6. Two of the 14 questions regarding participation rate and loss to follow up were considered not applicable. All included studies identified clear objectives and a clearly defined study population. No clear inclusion or exclusion criteria were specified for 17 of the 19 studies. Power estimation was made for one study33, and recording of all exposures (disease activity, comorbidities etc) were made prior to assessment of the outcomes (cytokine levels). The timeframe of analysis was sufficient to identify an association, but only 10 of the 19 studies (52.6%) documented different levels of exposures (disease severity, metabolic comorbidities, family history etc). There were no serial measures of cytokine levels in the majority of studies. Only three studies5,25,33, examining cytokine levels after monoclonal antibody administration has measurements at two distinct time points. Outcomes of interest (cytokine levels) were measured consistently within studies, however there was great variance in the methods of measurement and analysis between studies (Table 5). No studies took into account known confounding variables into analysis of their results by stratification or regression analyses.\n\nKey: Y = Yes; N= No, NR= Not Reported N/A = Not Applicable\n\n36 of the 81 identified cytokines or inflammatory proteins were assessed by more than 1 study. 23 of those cytokines had raw data available. No studies had sufficient measures of spread in order to calculate I2measure of heterogeneity and so chi-squared statistic was used as an alternate marker of heterogeneity (Table 7) along with a funnel plot (Figure 3). In total, 18 individual cytokines (78.2%) were found to demonstrate heterogeneity. Only eight cytokines (Serum IL-10, Lesional IL-1α, IL-12p70, hBD1, hBD2, hBD3, S100A9 and GMCSF) illustrated homogeneity. Due to this high level of heterogeneity and concerns regarding the methodological quality of included studies, meta-analysis was not deemed appropriate to perform.\n\n\nDiscussion\n\nThe overall quality of reporting in the identified studies was low with little consistency between methodologies and cytokines examined. There was also great variability in the ages, genders, comorbidities, associated conditions and treatments of the patients included in these studies. This was again reflected in the high number of cytokines with statistical heterogeneity (Table 7). The studies presenting conflicting data are often those studies with lower numbers of patients as well as lack of matched controls and/or lack of stratification by treatment. Meta-analysis using individual patient data would be required in order to account for these factors and re-assess the relationship between lesional and control cytokine levels.\n\nIn assessing the relationship between lesional and peri-lesional tissue, it has been demonstrated by many authors that different cytokines are present in peri-lesional tissue as opposed to lesional tissue. The definition of peri-lesional tissue is fairly consistent in the studies examined being 2cm from an active HS nodule on unaffected skin. However, no studies reported ultrasound examination of the peri-lesional skin to ensure that subclinical extension of the adjacent nodule (either in the dermis or the subcutaneous tissue) was being inadvertently sampled. This is an important differentiation to make in terms of identifying the subclinical pathogenic processes that precipitate this disease.\n\nThe raw data collated illustrates a number of paradoxically elevated levels of control cytokines (IL-15, IL-16) (Table 4). Many of these control readings lie near the lower detection limit of specific assays in individual papers, and thus the possibility of erroneously elevated control readings cannot be excluded. The wide interquartile ranges of studies which did report individual patient data7, suggest that analyzing aggregate data is not optimal and is prone to misrepresentation of the relationship between clinical disease, comorbidities and cytokine levels. Furthermore, high levels of heterogeneity within the measurements of individual cytokines suggest that examination of and correction for other variables or confounders is required.\n\nRegarding methods of cytokine analysis, a number of authors have identified variability in cytokine levels measured with different forms of multiplex assays as well as traditional ELISA methods35–39. Different methods of cytokine analysis are known to be prone to variability, with some cytokines more sensitive than others. For example, IFN-γ and IL-1β were overestimated compared with ELISA methods37, whilst IL-6 levels were underestimated37. IL-6 levels when compared across four different multiplex assays showed significant variation in detectable range, accuracy and responsiveness36. The correlation of TNF-α between ELISA and Multiplex assays was also poor (r=0.31)36. Issues also exist with minimum detectable levels of cytokines with specific bead-based arrays36 As an example, minimal detectable dose readings reported for IL-12p70 using some multiplex arrays39 are higher than the levels reported in lesional HS samples6. Therefore, whilst the general trends in the level of consistently elevated or suppressed cytokines in HS are reliable, the quantification of individual cytokines as well as the relationship between comorbidities and cytokine levels requires further research with consistent, reliable and accurate methodologies in order to further dissect the inflammatory cascade in this disease.\n\nThe majority of elevated cytokines and inflammatory proteins identified in lesional skin of HS (TNF-α, IL-1β, IL-6, IL-8, IL-11, IL-23, IL-17A, IL-33, IL-36, LL-37, S100A7, S100A8, S100A9, GM-CSF, TGF-β, hBD2, hBD3, CCL3, CXCL9, CXCL11, PDGF, CCL5, CCL-20, MIF, GM-CSF and LCN2) are those known to be produced by keratinocytes, as well as perpetuating a self-amplification pathway34 (Figure 2). Additionally T-cells produce IL-17A, IL-17F, IL-26, IL-29, and IFN-γ; dendritic cells produce IL-12, IL-23 and possibly IL-39; neutrophils produce S100A8 and S100A9 (calgranulin); and innate lymphoid cells also contribute IFN-γ, IL-17A and IL-17F. This inflammatory model has been well documented and explored in both psoriasis and atopic dermatitis34,40. The psoriasiform epidermal hyperplasia seen in HS (mediated by IL-17 and maintained by IL-23-mediated Th17 stimulation)34 reflects this common inflammatory pathway.\n\nImmunological ‘priming’ occurs due to the contribution of adipose tissue, genetic susceptibility, smoking-related inflammatory mediators and obesity related pro-inflammatory signals and the composition of the microbiome. Increased activity of cDC1, cDC2 and T cells lead to both keratinocyte hyperplasia via the actions of IL-12 and IL-23, as well as a Th17 predominant immune response. Alterations of antimicrobial peptides (AMP’s) also occur throughout the epidermis. The dermal inflammation interacting with the hyperplastic epidermis result leads to a self-perpetuating inflammatory feed forward mechanism mediated by IL-36, Il-1B and TNF-a. The development of scarring and sinus tracts is associated with MMP2, ICAM-1 and TGF-Beta, with possible augmentation of ICAM-1 and TGF-B signaling via specific components of the microbiome. TNF-a, PGE2 and CXCL2 then lead to additional feed forward mechanisms perpetuating the inflammatory cycle.\n\nIL-1a = Red, IL-10 = Blue, IL-12p70 = Green, hBD1 = Purple, hBD2 = light purple, hBD3 = Black, S100A9 = White, GMCSF = Yellow.\n\nThe other elevated non-keratinocyte produced cytokines in HS (IL-4, IL-5, IL-10, IL-16, IL-17A, IL-22, IL-32, IL-36, hBD1), are produced by a combination of dendritic cells, monocytes, neutrophils and CD4+ T cells. IL-4 and IL-5 as key cytokines in the Th2 axis are consistent with the findings of Mast cells in HS41, as well as the pruritus, which is frequently reported by patients. IL-10 in HS is produced by Treg cells2 (although dendritic cells may also be a source), and whilst quantitatively the IL-10 signal appears paradoxically elevated, it can be explained by the up-regulation of T cells including Treg cells, which although significantly elevated from baseline, are not elevated enough in comparison to TH17/IL-17/IL-22 signal to counteract this strong pro-inflammatory cascade2. Further exploration of these cytokines may reveal the initial trigger(s) of the inflammatory cascade in HS, or correlations with known pro-inflammatory comorbidities.\n\nIn light of investigations in psoriasis and atopic dermatitis, the role of dendritic cells in HS needs to be clarified, as dendritic cell influx has been reported in histological studies41,42, and they may contribute to the high IL-10 and IL-15 levels reported. IL-32 is a second cytokine produced by dendritic cells, but has only been reported in one study29. Further research into the functional role of IL-32 in the activity of dendritic cells in HS would be of value. The role of IL-20, IL-22, IL-24 and IL-26 needs further clarification. IL-19, TSLP and CCL17 (TARC) have not yet been examined in HS and this is required in order to further explore the role of dendritic cell, monocyte and T cell activation and migration in this disease.\n\nIt is well established that smoking, obesity and diabetes are strongly associated with HS13–19,42,43. The immunological effects of smoking include increase in number and responsiveness of dendritic cells, altered function of Treg cells and activation of Th17 pathways44, whilst obesity and diabetes can result in production of IL-1β, IL-6 and TNF-α through activated macrophages in adipose tissue45,46. These potential mechanistic pathways (which may prime or contribute towards inflammation in HS) require validation in functional studies. However, if they are a significant contributor to inflammation, the presence or absence of these comorbidities need to be considered in future cytokine studies as confounding variables in order to identify significant biochemical markers independent of these other pro-inflammatory states that reflect the pathogenesis of HS.\n\nThe role of the microbiome42,43 in stimulating chronic inflammation has parallels in diabetes47 and colonic inflammation48 and the presence of Porphyromonas and Peptoniphilus species has been associated with a subpopulation of patients with HS42. Porphyromonas has been associated with systemic inflammation and atherosclerosis through aberrant toll-like-receptor 4 signalling48 and is not part of the natural cutaneous flora43. Altered cutaneous and gastrointestinal microbiome can also act via microbiome metabolites (including lipopolysaccharides, short chain fatty acids and bile salts)49 through stimulation of myeloid dendritic cells via G Protein Coupled Receptors (including GPR41, GPR43 and GPR109A)49,50. The microbiome may be implicated as a trigger factor for the initial inflammatory cascade in HS in a proportion of patients. Similarly, the presence of genetic polymorphisms as reported in HS51 have the potential to up-regulate inflammatory activity through shedding of IL-6R, IL-15R, TNF-α52 as well as up-regulating the response of dendritic cells to LPS stimulation via ADAM17 (which has been demonstrated to be elevated in a published gene expression study of HS)53. These pathways may be involved prior to the activation of keratinocyte-mediated inflammation, and hence, may reveal novel targets for new interventions to control the disease prior to the onset of destructive inflammation.\n\nThe limitations to this study include the high degree of methodological variability (Table 5) and high impact of bias (Table 6) within the included studies. The lack of individual patient data has also prevented any further analysis into the contribution of comorbidities such as smoking and obesity to variable levels of cytokines in lesional tissue and/or serum. This, along with the high level of heterogeneity in many cytokines (Table 7), has resulted in analyses of the collated data being limited to descriptive analyses only and limited the generalisability of results.\n\n\nConclusions\n\nThrough this review we have catalogued the various cytokines that have been reported as elevated in lesional, peri-lesional tissue, serum or exudate of HS patients. We have also identified those cytokines with inconsistent results and identified methodological factors that may explain variability in findings. We have identified a number of missing links in disease pathogenesis with respect to cytokine actions and pathways that must be addressed in future work. Areas for further investigation include the role of dendritic cells in HS, the contribution of obesity, smoking, diabetes and the microbiome to cytokine profiles in HS, and examining the natural history of the disease through longitudinal measurements of cytokines over time.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\nOSF: Extend data. Data Collection Sheet Cytokine. Review HS. https://doi.org/10.17605/OSF.IO/N2E7A22\n\nLicense: CC0 1.0 Universal\n\nOSF: PRISMA checklist for ‘A systematic review and critical evaluation of inflammatory cytokine associations in hidradenitis suppurativa’. https://doi.org/10.17605/OSF.IO/N2E7A22\n\nLicense: CC0 1.0 Universal", "appendix": "Grant information\n\nSupported in part by a grant from the National Center for Advancing Translational Sciences (NCATS) [UL1 TR001866], National Institutes of Health (NIH) Clinical and Translational Science Award (CTSA) program.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nHoffman LK, Ghias MH, Lowes MA: Pathophysiology of hidradenitis suppurativa. Semin Cutan Med Surg. 2017; 36(2): 47–54. PubMed Abstract \| Publisher Full Text\n\nMoran B, Sweeney CM, Hughes R, et al.: Hidradenitis Suppurativa is characterized by Dysregulation of the Th17:Treg Cell Axis, which is Corrected by Anti-TNF Therapy. J Invest Dermatol. 2017; 137(11): 2389–2395. PubMed Abstract \| Publisher Full Text\n\nHessam S, Sand M, Gambichler T, et al.: Interleukin-36 in hidradenitis suppurativa: evidence for a distinctive proinflammatory role and a key factor in the development of an inflammatory loop. Br J Dermatol. 2018; 178(3): 761–767. PubMed Abstract \| Publisher Full Text\n\nHotz C, Boniotto M, Guguin A, et al.: Intrinsic Defect in Keratinocyte Function Leads to Inflammation in Hidradenitis Suppurativa. J Invest Dermatol. 2016; 136(9): 1768–1780. PubMed Abstract \| Publisher Full Text\n\nKanni T, Tzanetakou V, Savva A, et al.: Compartmentalized Cytokine Responses in Hidradenitis Suppurativa. PLoS One. 2015; 10(6): e0130522. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBanerjee A, McNish S, Shanmugam VK: Interferon-gamma (IFN-γ) is Elevated in Wound Exudate from Hidradenitis Suppurativa. Immunol Invest. 2017; 46(2): 149–158. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nThomi R, Schlapbach C, Yawalkar N, et al.: Elevated levels of the antimicrobial peptide LL-37 in hidradenitis suppurativa are associated with a Th1/Th17 immune response. Exp Dermatol. 2018; 27(2): 172–177. PubMed Abstract \| Publisher Full Text\n\nJiménez-Gallo D, de la Varga-Martínez R, Ossorio-García L, et al.: The Clinical Significance of Increased Serum Proinflammatory Cytokines, C-Reactive Protein, and Erythrocyte Sedimentation Rate in Patients with Hidradenitis Suppurativa. Mediators Inflamm. 2017; 2017: 2450401. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTen Oever J, van de Veerdonk FL, Joosten LA, et al.: Cytokine Production Assays Reveal Discriminatory Immune Defects in Adults with Recurrent Infections and Noninfectious Inflammation. Clin Vacc Immunol. 2014: 21(8): 1061–1069. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCanoui-Poitrine F, Le Thuaut A, Revuz JE, et al.: Identification of three hidradenitis suppurativa phenotypes: latent class analysis of a cross-sectional study. J Invest Dermatol. 2013; 133(6): 1506–11. PubMed Abstract \| Publisher Full Text\n\nvan der Zee HH, Jemec GB: New Insights into the diagnosis of hidradenitis suppurativa: Clinical presentations and phenotypes. J Am Acad Dermatol. 2015; 73(5 Suppl 1): S23–26. PubMed Abstract \| Publisher Full Text\n\nZhou X, Fragala MS, McElhaney JE, et al.: Conceptual and Methodological Issues relevant to Cytokine and inflammatory marker measurements in clinical research. Curr Opin Clin Nutr Metab Care. 2010; 13(5) : 541–547. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nErgun T: Hidradenitis Suppurativa and the Metabolic Syndrome. Clin Dermatol. 2018; 36(1): 41–47. PubMed Abstract \| Publisher Full Text\n\nGarg A, Birabaharan M, Strunk A: Prevalence of type 2 diabetes mellitus among patients with hidradenitis suppurativa in the United States. J Am Acad Dermatol. 2018; 79(1): 71–76. PubMed Abstract \| Publisher Full Text\n\nLukach AJ, Saul MI, Ferris LK, et al.: Risk Factors for Hidradenitis Suppurativa in Patients with Inflammatory Bowel Disease. Dig Dis Sci. 2018; 63(3): 755–760. PubMed Abstract \| Publisher Full Text\n\nGarg A, Papagermanos V, MIdura M, et al.: Incidence of Hidradenitis Suppurativa among tobacco smokers: a population-based retrospective analysis in the U.S.A. Br J Dermatol. 2018; 178(3): 709–714. PubMed Abstract \| Publisher Full Text\n\nSgambato JA, Jones BA, Caraway JW, et al.: Inflammatory profile analysis reveals differences in cytokine expression between smokers, moist snuff users, and dual users compared to non-tobacco consumers. Cytoine. 2018; 107: 43–51. PubMed Abstract \| Publisher Full Text\n\nSpoto B, Di Betta E, Mattace-Raso F, et al.: Pro- and anti-inflammatory cytokine gene expression in subcutaneous and visceral fat in severe obesity. Nutr Metab Cardiovasc Dis. 2014; 24(10): 1137–43. PubMed Abstract \| Publisher Full Text\n\nRehman K Akash MS: Mechanisms of inflammatory responses and development of insulin resistance: how are they interlinked? J Biomed Sci. 2016; 23(1): 87. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFrew J, Hawkes J Krueger J: Cytokine Studies in Hidradenitis Suppurativa: A Systematic Review of Observational Studies. PROSPERO 2018:CRD42018104664, Accessed 30 July 2018. Reference Source\n\nMoher D, Liberati A, Tetzlaff J, et al.: Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009; 6(7): e1000097. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFrew J: A Systematic Review and Critical Evaluation of Inflammatory Cytokine Associations in Hidradenitis Suppurativa. 2018. http://www.doi.org/10.17605/OSF.IO/N2E7A\n\nNIH NIoH: Quality Assessment Tool for Observational Cohort and Cross Sectional Studies. 2018; Accessed 30 July 2018. Reference Source\n\nEmelianov VU, Bechara FG, Gläser R, et al.: Immunohistological pointers to a possible role for excessive cathelicidin (LL-37) expression by apocrine sweat glands in the pathogenesis of hidradenitis suppurativa/acne inversa. Br J Dermatol. 2012; 166(5): 1023–1034. PubMed Abstract \| Publisher Full Text\n\nJiménez-Gallo D, de la Varga-Martínez R, Ossorio-García L, et al.: Effects of adalimumab on T-helper-17 lymphocyte- and neutrophil-related inflammatory serum markers in patients with moderate-to-severe hidradenitis suppurativa. Cytokine. 2018; 103: 20–24. PubMed Abstract \| Publisher Full Text\n\nvan der Zee HH, Laman JD, de Ruiter L, et al.: Adalimumab (antitumour necrosis factor-α) treatment of hidradenitis suppurativa ameliorates skin inflammation: an in situ and ex vivo study. Br J Dermatol. 2012; 166(2): 298–305. PubMed Abstract \| Publisher Full Text\n\nWolk K, Wenzel J, Tsaousi A, et al.: Lipocalin-2 is expressed by activated granulocytes and keratinocytes in affected skin and reflects disease activity in acne inversa/hidradenitis suppurativa. Br J Dermatol. 2017; 177(5): 1385–1393. PubMed Abstract \| Publisher Full Text\n\nThomi R, Kakeda M, Yawalkar N, et al.: Increased expression of the interleukin-36 cytokines in lesions of hidradenitis suppurativa. J Eur Acad Dermatol Venereol. 2017; 31(12): 2091–2096. PubMed Abstract \| Publisher Full Text\n\nThomi R, Yerly D, Yawalkar N, et al.: Interleukin-32 is highly expressed in lesions of hidradenitis suppurativa. Br J Dermatol. 2017; 177(5): 1358–1366. PubMed Abstract \| Publisher Full Text\n\nLima AL, Karl I, Giner T, et al.: Keratinocytes and neutrophils are important sources of proinflammatory molecules in hidradenitis suppurativa. Br J Dermatol. 2016; 174(3): 514–521. PubMed Abstract \| Publisher Full Text\n\nKelly G, Hughes R, McGarry T, et al.: Dysregulated cytokine expression in lesional and nonlesional skin in hidradenitis suppurativa. Br J Dermatol. 2015; 173(6): 1431–1439. PubMed Abstract \| Publisher Full Text\n\nSchlapbach C, Hänn T, Yawalkar N, et al.: Expression of the IL-23/Th17 pathway in lesions of hidradenitis suppurativa. J Am Acad Dermatol. 2011; 65(4): 790–8. PubMed Abstract \| Publisher Full Text\n\nKanni T, Argyropoulou M, Spyridopoulos T, et al.: MABp1 Targeting IL-1α for Moderate to Severe Hidradenitis Suppurativa Not Eligible for Adalimumab: A Randomized Study. J Invest Dermatol. 2018; 138(4): 795–801. PubMed Abstract \| Publisher Full Text\n\nHawkes JE, Chan TC, Krueger JG: Psoriasis pathogenesis and the development of novel targeted immune therapies. J Allergy Clin Immunol. 2017; 140(3): 645–653. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBastarache JA, Koyama T, Wickersham NE, et al.: Validation of a multiplex electrochemiluminescent immunoassay platform in human and mouse samples. J Immunol Methods. 2014; 408: 13–23. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nThompson DK, Huffman KM, Kraus WE, et al.: Critical appraisal of four IL-6 immunoassays. PLoS One. 2012; 7(2): e30659. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nde Koning L, Liptak C, Shkreta A, et al.: A multiplex immunoassay gives different results than singleplex immunoassays which may bias epidemiologic associations. Clin Biochem. 2012; 45(10–11): 848–851. PubMed Abstract \| Publisher Full Text\n\nTighe PJ, Ryder RR, Todd I, et al.: ELISA in the multiplex Era: Potentials and Pitfalls. Proteomics Clin Appl. 2015; 9(3–4): 406–22. PubMed Abstract \| Publisher Full Text\n\ndupont NC, Wang K, Wadhwa PD, et al.: Validation and comparison of luminex multiplex cytokine analysis kits with ELISA: determinations of a panel of nine cytokines in clinical sample culture supernatants. J Reprod Immunol. 2005; 66(2): 175–191. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrunner PM, Guttman-Yassky E, Leung DY: The immunology of atopic dermatitis and its reversibility with broad-spectrum and targeted therapies. J Allergy Clin Immunol. 2017; 139(4S): S65–S76. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nvan der Zee HH, de Ruiter L, Boer J, et al.: Alterations in leucocyte subsets and histomorphology in normal-appearing perilesional skin and early and chronic hidradenitis suppurativa lesions. Br J Dermatol. 2012; 166(1): 98–106. PubMed Abstract \| Publisher Full Text\n\nRing HC, Thorsen J, Saunte DM, et al.: The Follicular Skin Microbiome in Patients With Hidradenitis Suppurativa and Healthy Controls. JAMA Dermatol. 2017; 153(9): 897–905. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGuen-Revillet H, Jais JP, Ungeheuer MN, et al.: The Microbiological Landscape of Anaerobic Infections in Hidradenitis Suppurativa: A Prospective Metagenomic Study. Clin Infec Dis. 2017; 65(2): 282–291. PubMed Abstract \| Publisher Full Text\n\nQiu F, Liang CL, Liu H, et al.: Impacts of cigarette smoking on immune responsiveness: Up and down or upside down? Oncotarget. 2017; 8(1): 268–284. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKothari V, Galdo JA, Mathews ST: Hypoglycemic agents and potential anti-inflammatory activity. J Inflamm Res. 2016; 9: 27–38. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRichards JL, Yap YA, McLeod KH, et al.: Dietary metabolites and the gut microbiota: an alternative approach to control inflammatory and autoimmune diseases. Clin Transl Immunology. 2016; 5(5): e82. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBasson A, Trotter A, Rodruiguez-Palacios A, et al.: Mucosal Interactions between Genetics, Diet, and Microbiome in Inflammatory Bowel Disease. Front Immunol. 2016; 7: 290. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKramer CD, Genco CA: Microbiota, Immune Subversion, and Chronic Inflammation. Front Immunol. 2017; 8: 255. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKim CH: Immune regulation by microbiome metabolites. Immunology. 2018; 154(2): 220–229. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTan JK, McKenzie C, Marino E, et al.: Metabolite-Sensing G Protein-Coupled Receptors-Facilitators of Diet-Related Immune Regulation. Annu Rev Immunol. 2017; 35: 371–402. PubMed Abstract \| Publisher Full Text\n\nFrew JW, Vekic DA, Woods J, et al.: A systematic review and critical evaluation of reported pathogenic sequence variants in hidradenitis suppurativa. Br J Dermatol. 2017; 177(4): 987–998. PubMed Abstract \| Publisher Full Text\n\nGooz M: ADAM-17: the enzyme that does it all. Crit Rev Biochem Mol Biol. 2010; 45(2): 146–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBlok JL, Li K, Brodmerkel C, et al.: Gene expression profiling of skin and blood in hidradenitis suppurativa. Br J Dermatol. 2016; 174(6): 1392–4. PubMed Abstract \| Publisher Full Text" }	[ { "id": "43493", "date": "07 Feb 2019", "name": "Barbara Horváth", "expertise": [ "Reviewer Expertise Hidradenitis suppurativa" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThank you for the opportunity to review this manuscript and congratulations to the authors for their great efforts in putting this systematic review together. Research on the role of cytokines in HS is important, as it may lead to new targets for therapy and a better understanding of the pathophysiology of HS.\n\nSummary\nThis systematic review focused on collecting all data published on cytokine studies in tissue, blood, serum and exudate in hidradenitis suppurativa. 81 discrete cytokines were examined in HS patients (n=564) and control patients (n=198) in 19 studies. Methodology varied greatly among studies, which were generally of low quality. When measuring levels of cytokines, substantial variance was found and the majority of cytokines showed heterogeneity. IL-17 signalling appeared to be a significant component. Suggestions for further research were discussed.\n\nQuestions\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Yes.\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Yes. However, I wonder why the term ‘hidradenitis suppurativa’ is not in the search strategy and ‘hidradenitidis suppurative’ is? ‘Hidradenitidis’ is not an existing word, as far as I know and will not provide any search results. Please adjust.\n\nIs the statistical analysis and its interpretation appropriate? Yes, as far as I can judge as a non-statistician. The analyses used are ones I have little experience with myself. I’ll refrain from commenting on this section.\n\nAre the conclusions drawn adequately supported by the results presented in the review? Partly. The last conclusion ‘examining the natural history of the disease through longitudinal measurements of cytokines over time’ is not discussed anywhere else in this article. First, I suggest changing ‘history’ to ‘course’. Moreover, I am wondering, how the authors propose to do this. Monitoring the natural course of the disease, would mean patients cannot receive any treatment for their HS, during this proposed study. Depending on how long the natural course is meant to be monitored, I don’t think it is ethical to withhold patients from treatment. Please elaborate on this conclusion with a specific proposal or otherwise rephrase or maybe leave out this conclusion.\n\nOther comments Page 9 last paragraph/Page 28 – 1st paragraph: You state that ‘psoriasiform epidermal hyperplasia is seen in HS’. Please provide a reference for this statement. The reference provided only references to the pathway likely responsible for this in psoriasis.\n\nPage 28 – 4th paragraph: ‘These potential mechanistic pathways (which may prime or contribute towards inflammation in HS) require validation in functional studies.’ Could you please provide an example on how such a functional study should be designed to produce reliable results?\n\nTable 4: the abbreviation ‘Lpa’ is not clarified in the key section of the table. Does ‘Le’ (page 11, IL-1a, first row) mean lesion exudate?\n\nTable 6: the number four of question four is missing in the top row of the table on both pages (24-25). Please insert.\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Yes\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Yes\n\nIs the statistical analysis and its interpretation appropriate? Yes\n\nAre the conclusions drawn adequately supported by the results presented in the review? Partly", "responses": [] }, { "id": "43995", "date": "25 Feb 2019", "name": "Aude Nassif", "expertise": [ "Reviewer Expertise microbiology", "genetics", "therapeutics", "clinical forms of HS and associated diseases" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis very instructive study aims at analyzing previous cytokine studies in HS patients, in skin tissue, blood, serum and exudates, to assess relevancy and reliability of these studies.\nThe authors have performed an extensive work, methods seem perfectly appropriate. The authors are very critical and rigorous in their approach, looking for confounding factors, which is highly desired.\nThe authors could also mention that genetic heterogeneity may play a role in the diversity of results and encourage using similar phenotypes for future studies.\nThis analysis brings up a very important and honest contribution to the current knowledge in cytokines involved in HS and therefore deserves indexing.\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Yes\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Yes\n\nIs the statistical analysis and its interpretation appropriate? I cannot comment. A qualified statistician is required.\n\nAre the conclusions drawn adequately supported by the results presented in the review? Yes", "responses": [] }, { "id": "43994", "date": "04 Mar 2019", "name": "Evangelos J. Giamarellos-Bourboulis", "expertise": [ "Reviewer Expertise Immunology", "genetics", "hidradenitis", "anti-cytokine therapies" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a long time needed review trying to shed light in the pathogenesis of hidradenitis suppurativa (HS). My concerns are coming from the biggest hurdle the authors had to overcome from the very beginning of their attempt i.e. the great heterogeneity of the existing evidence. Due to this, I find over-exaggerated the conducted approach to set-up a mechanistic interpretation for the disease. I believe that the heterogeneity is so vast that it is almost impossible to suggest the pathways implicated in the pathogenesis of HS. To this end, I suggest that the mechanistic parts are omitted and Figure 2 as well.\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Yes\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Yes\n\nIs the statistical analysis and its interpretation appropriate? Yes\n\nAre the conclusions drawn adequately supported by the results presented in the review? Partly", "responses": [] } ]	1	https://f1000research.com/articles/7-1930
https://f1000research.com/articles/6-1692/v1	15 Sep 17	{ "type": "Research Article", "title": "A survey of quality of life indicators in the Romanian Roma population following the ‘Decade of Roma Inclusion’", "authors": [ "Rebecca Powell Doherty", "Daniel Müller-Demary", "Alexandra Hosszu", "Ana Duminica", "Andrea Bertke", "Bryan Lewis", "Stephen Eubank", "Daniel Müller-Demary", "Alexandra Hosszu", "Ana Duminica", "Andrea Bertke", "Stephen Eubank" ], "abstract": "Background: This study explores how the Roma in Romania, the EU’s most concentrated population, are faring in terms of a number of quality of life indicators, including poverty levels, healthcare, education, water, sanitation, and hygiene. Methods: 135 surveys were conducted across five geographically diverse Romanian communities. Household participants were selected through a comprehensive random walk method. Analyses were conducted on all data using Pandas for Python. Results: These data indicate that the Roma in Romania face significant disparities in education, with Roma students less likely to progress beyond 8th grade. In addition, the Roma population remains significantly disadvantaged with regard to safe and secure housing, poverty, and healthcare status, particularly in connection to diarrheal disease. In contrast, however, both Roma and non-Roma in rural areas face difficulties regarding full-time employment, sanitation, and water, sanitation, and hygiene infrastructure. Conclusions: These data demonstrate the challenges that remain to the Roma population in Romania, and also point to the myriad of ways in which all rural Romanians, regardless of ethnicity, are encountering hardship. This study highlights the areas in which improvements can be made to ensure the Roma, and indeed all Romanian citizens, have access to and confidence in sanitation services, clean water, and adequate healthcare treatment.", "keywords": [ "Roma", "Romania", "rural populations", "water quality", "healthcare", "development", "global health", "decade of Roma inclusion" ], "content": "Introduction\n\nIn the years that followed independence from Soviet rule and the democratic election of 1990, the southeastern European country of Romania received significant aid from the International Monetary Fund (IMF), World Bank (WB), European Bank for Reconstruction and Development (EBRD), European Investment Bank (EIB), the US Agency for International Development (USAID), and other donors1. This influx of investment enabled Romania to make great strides in multiple areas of development and meet a number of the goals set forth in the United Nations Millennium Development Goals (UN MDGs)2. In particular, the issues of severe poverty and hunger have significantly improved for ethnic Romanians and affluent minorities, with severe poverty (as defined by the United Nations) decreasing from 10 per cent to 4.1 per cent as of 20062. In addition, maternal mortality has fallen by half to 17 deaths/100,000 births, infant mortality has decreased 25 per cent, and Romania has seen a significant decrease in adolescent pregnancy, concomitant with a significant increase in the use of modern contraceptives2. Vaccination rates, particularly for measles, hover around 98 per cent, up from less than 70 per cent at the time of independence; HIV/AIDS cases have decreased and life expectancy for those living with HIV has increased dramatically; and there has been a significant decrease in domestic violence2.\n\nFor the Roma, the second most numerous minority in the country (after Hungarians), however, such progress was not extended. Despite enjoying a reprieve from targeted discrimination during the Soviet era, Romanian independence brought on a renewal of oppressive policies and behaviours against the Roma. The Roma are Europe’s most marginalised group3, a minority population numbering between 10–12 million individuals across the continent and the UK4. Emerging from slavery in the late 19th century, they have historically faced discrimination in employment, education, and access to healthcare5. Numerous studies indicate Roma have a significantly reduced lifespan compared to non-Roma and suffer greater rates of communicable and waterborne diseases6–8. In multiple countries, they are less likely to have access to basic services, including a municipal water supply, waste water treatment, or trash disposal9, and they are routinely used as political scapegoats across the continent, from France to Moldova10. Romania boasts the largest concentration of Roma in the European Union (EU), at approximately 1.85 million individuals, representing 9.3 per cent of the overall population of 19.8 million, though official census numbers vary4.\n\nThe addition of eastern European countries (including Bulgaria, Romania, and Hungary) to the EU in the mid-2000s has renewed interest in the well-being of this population, as indicated by the EU’s targeted attempt to improve the circumstances of the Roma through the recently concluded Decade of Roma Inclusion (DRI), a ten year long initiative by twelve European countries to improve the socio-economic status and social standing of the Roma minority across the continent11. Numerous studies have explored the success of the DRI, both during its implementation and since its conclusion, and outcomes vary, depending on the sector and goal in question8,12–14. This study has been prompted, in part, to explore how the EU’s largest concentration of Roma are faring in terms of poverty levels, healthcare, education, and water, sanitation, and hygiene (WASH), as well as to fill the gap in available literature that focuses solely on Romania. In particular, we examine the connection between physical WASH infrastructure relative to incidence of disease and overall health status.\n\n\nMethods\n\nCombining questions from a validated WASH survey previously used for multiple use service strategy research (MUS) in Burkina Faso (personal communication to authors) and the WHO core questions on drinking-water and sanitation15 with questions related to demographics, socio-economic status, and healthcare access and history, we conducted 135 surveys each consisting of 56 total questions across five geographically diverse communities throughout Romania. Communities were chosen at random from a list of those that had previously participated with Agentia Impreuna in education and anti-discrimination capacity-building programs for communities with prominent Roma populations. Household participants were selected through a comprehensive random walk method, with survey teams accompanied by both Roma and non-Roma community leaders. Any household with an individual over the age of 18 present and willing to participate, regardless of ethnicity, was included until the desired 30 surveys per community were achieved or there were no further willing participants. Interviews were conducted by trained volunteers who either spoke the national language (Romanian) or were accompanied by a certified translator. The team interviewed only one member of each household, who provided information about all members of the household.\n\nSurveys (Supplementary material 1 and Supplementary material 2) and procedures were approved by the Virginia Tech Institutional Review Board (IRB) prior to study implementation (VT IRB #16-475), and all interviews and analysis were carried out according to IRB protocol.\n\nInformed consent was obtained from all individual participants included in this study. A brief explanation of the survey questions and the intended use of the data was provided to each participant, and the individual’s agreement to participate in the survey interview was considered consent, as indicated by the IRB protocol. Further, interviewers ensured each participant understood that he or she could refuse to answer any question and could withdraw their consent at any time. Survey participation was entirely anonymous, and no identifying information was obtained. In addition, the IRB stipulated that location data for the participating villages remain unavailable, due to the vulnerable population and minority status of some study participants. All demographic information was self-reported, and those who were considered part of the Roma sample self-identified as either Roma or Rudar (a sub-set of Roma people who do not speak Romani), in response to a question that explicitly asked for their ethnicity (Dataset 1).\n\nAll data analyses were conducted via Pandas with Python (version 2.7.11 & 0.18.0) notebook and the software package Epipy16,17 (Dataset 2–Dataset 3). Descriptive statistics were broken down by community, ethnicity, gender, age, household size, education level, marital status, employment, literacy, and geographical description (urban versus rural). WASH parameters were defined using the UN descriptions as provided in the DRI progress report through 2013, as well as the addition of a ‘safe water score’, which included the option of a private, protected well water source in addition to tap water in the home11. The overall WASH score for each participating household is an aggregate of the following UN parameters: indoor toilet (improved sanitation), indoor bathroom (improved sanitation II), piped water to tap (improved water source), and insecure housing (a 0–3 score reflecting the status of the floor, walls, and roof of a dwelling). The overall ‘WASH Safe’ score exchanged the improved water source parameter for the aforementioned safe water score. In addition, time to primary drinking water sources has been converted to a numerical scale, based on 15 minute intervals, up to one hour (0–4 scale). Distance to primary drinking water is indicated both by a percentage of those in each ethnic group who travel a kilometre or more and the average distance travelled by each group. Similar to the WASH score, the healthcare score is an aggregate of self-reported immunization, reported incidence of diarrheal event, access to primary care physician (PCP), and reported medical insurance status. Finally, the poverty score is an aggregate of available electricity in dwelling, available gas source in dwelling, and the UN indicator of severe poverty (surviving on 2USD/person/day or less). Univariate analyses compared the Roma sample to the non-Roma sample for each variable (using non-Roma as the reference population), as well as urban areas to rural ones (with urban areas as the reference population) for some parameters. Odds Ratios (ORs) with 95 per cent confidence intervals are reported, as are t-test results (95 per cent confidence interval) with accompanying p-value where appropriate.\n\nMultivariate linear regression analyses were conducted by using combinations of the four aggregate scores, as explained in primary analysis, and by including parameters that demonstrated significance in univariate modelling (Dataset 2–Dataset 3).\n\n\nResults\n\nAnalyses of demographic data and breakdown by percentage indicate our sample population is, overall, predominantly Roma (72.6 per cent vs. 27.4 per cent non-Roma), split evenly by sex (50.4 per cent Female, 49.6 per cent Male), and average approximately 47 years of age (Table 1). Three of the five sample communities are rural (more than 25km from a city centre), one is suburban (between 10–25km from a city centre), and one is urban (less than 10km from a city centre). There is no significant difference between Roma and non-Roma in the sample population on the basis of marital status, age, or sex. However, our data indicate notable disparities in level of education (secondary school completion for Roma vs. high school completion for non-Roma), household size (5.3 individuals for Roma vs. 4.2 individuals for non-Roma), and literacy rate (61 per cent literate Roma vs. 97.4 per cent literate non-Roma) (Table 1). Little difference is noted in full-time employment rates between the groups (26.6 per cent Roma vs. 32.4 per cent non-Roma), though some difference is observable between rural and urban communities (Table 1).\n\nRomania, 2016. M=male, F=female, FT=full-time, UE=unemployed, DL=day labour.\n\nUsing parameters utilized by the DRI in the 2011 progress report, univariate analysis indicates little difference between Roma and non-Roma with regard to specific WASH variables. The non-Roma are slightly more likely to have an indoor toilet (21.6 per cent non-Roma vs 17.3 per cent Roma) and bathroom (21.6 per cent non-Roma vs 20.4 per cent Roma), but the Roma are more likely than non-Roma to have tap (indoor or outdoor) water (20.4 per cent Roma vs 8.1 per cent non-Roma), whether piped in from a personal well or a municipal water source (Table 2). However, when considering all safe water options (including a protected well without a tap to the home or garden), non-Roma report greater accessibility (59.5 per cent non-Roma vs 50 per cent Roma). In addition, Roma are significantly more at risk to inhabit insecure housing, regardless of geographical region, than non-Roma (27.6 per cent Roma vs 5.4 per cent non-Roma) (Table 2). Interestingly, while the Roma population have greater access to tap water (indoor or outdoor), they are less likely to use it as their primary drinking water source, demonstrated by the increased time and distance Roma are likely to travel to secure safe drinking water (12.2km Roma vs. 10.8km non-Roma; Table 2). Of interest, however, is the increased time all individuals in suburban and urban areas must travel to secure drinking water compared to their rural counterparts (16–30 minutes (1.2 on 0–3 scale) urban vs. 0–15 minutes (1.0 on 0–3 scale) rural) (Table 3).\n\nRomania, 2016. Reference population for all variables is non-Roma. * indicates significance at 95% CI level. ** indicates significance at 90% CI level.\n\nRomania, 2016. Reference population for all variables is urban. * indicates significance at 95% CI level.\n\nIn addition to physical infrastructure, we analysed the differences between Roma and non-Roma with regard to key factors contributing to overall health status. Roma are more than twice as likely to report at least one household member suffering from moderate to severe diarrhoea (lasting more than 3 days) than non-Roma (58.1 per cent Roma vs 40.5 per cent non-Roma; OR 2.04) (Table 2). In addition, while there is little difference in access to a primary care physician between the groups, Roma are approximately 1.5 times less likely to report having received an immunization of any kind (87.8 per cent Roma vs 97.1 per cent non-Roma; OR 1.58) and fewer Roma possess medical insurance (81.6 per cent Roma vs 89.1 per cent non-Roma; OR 1.86) than non-Roma (Table 2).\n\nFinally, we used the UN definition of extreme poverty (2USD/person/day or less) in addition to two other variables as an overall indicator of impoverished conditions (Table 2). Roma report a slightly greater, though not significant, incidence of lacking working electricity in their homes or dwellings (13.2 per cent Roma vs 2.7 per cent non-Roma), as well as lacking piped gas and/or the ability to purchase gas tanks (32.7 per cent Roma vs. 18.9 per cent non-Roma, p=0.12) (Table 2). Moreover, Roma report greater incidences of severe poverty (2USD/day/person or less) than non-Roma (55.1% per cent vs. 43.2 per cent) (Table 2), although overall, those in rural areas are significantly more susceptible to extreme poverty than those in suburban or urban communities (61.8 per cent rural vs. 32.6 per cent urban) (Table 3).\n\nFollowing univariate analysis, we used general multivariate linear regression analysis for four distinct models, combining categories that indicated a specific score (WASH, WASH Safe, poverty, healthcare) or approached a level of significance in the univariate analysis (Table 4). These analyses further demonstrate the significant (α = 0.05) disparity between Roma and non-Roma.\n\nRomania, 2016. All models use non-Roma as reference. * indicates significance at 95% CI level. ** indicates significance at 90% CI level.\n\nA multivariate combination of demographic variables further highlights the difference in education level and household size between Roma and non-Roma. Roma households are significantly larger than non-Roma households, but whether this is a correlation with birth rate or the presence of multiple generations in a single dwelling is beyond the scope of this study. Furthermore, Roma individuals are far less likely to complete required education (10th grade) than non-Roma individuals (MOD1; Table 4). In our univariate analysis, we broke down the score categories to their individual components and identified significant factors to further explore. Multivariate analysis of these parameters points to insecure housing as having the strongest correlation with being Roma, followed by access to tap water (improved water source), and less significantly, the occurrence of moderate or severe diarrhoea (MOD2; Table 4).\n\nFinally, we analysed our four score categories, using two different approaches. We first analysed the WASH score, as defined by the DRI, together with the healthcare and poverty scores (MOD3; Table 4). Healthcare and poverty equally significantly correlate with being Roma. The WASH score, however, is negatively correlated to the Roma, indicating that Roma individuals actually have an advantage over non-Roma individuals. To further investigate this question, we ran an additional analysis with healthcare and poverty, but substituting our WASH Safe score (MOD4; Table 4). The significant difference observed in healthcare and poverty remains, but when protected well water is included alongside tap water in the definition of improved or safe water sources, the disparity associated with WASH is eliminated.\n\n\nDiscussion\n\nA number of studies have examined the various factors the Decade of Roma Inclusion sought to address in Roma communities across the EU, both during the implementation of the project and since its conclusion in 20155,11,13,18,19. Unfortunately, while some improvements did occur, a number of studies indicate the DRI did not achieve its stated goals in the areas of education, housing, employment, and health status of Roma in participating countries20,21. Our study supports these conclusions, particularly with regard to education, healthcare, and poverty. However, disparities that other studies have highlighted in multiple countries with regard to employment and sanitation do not necessarily occur in Romania19,22–24. Rather, both the Roma and non-Roma in rural Romania face similar challenges regarding access to full-time employment and water, which are exacerbated by a lack of municipal sanitation services in over 800 Romanian communities25. The lack of significant difference between Roma and non-Roma in our sample in relation to indoor toilets and bathrooms does not indicate that either ethnic group has an advantage, but rather all those who reside in rural communities face a disadvantage, regardless of ethnicity. Notably, our findings indicate that, in some instances, the Roma appear to have a slight advantage over non-Roma (Table 4). Using the DRI definition of piped water to an indoor or outdoor tap, our analyses indicate Romanian and other non-Roma individuals lag behind the Roma in ‘improved water sources’. However, when one accounts for the prevalence of private, protected wells (WASH Safe score), the disparity is minimized and no longer significant (Table 4). We postulate this distinction is indicative of how our survey collected this type of data, and future iterations will refine how we classify ‘safe’ and ‘improved’ water sources.\n\nOf additional interest is the key indicator that those in suburban and urban areas, Roma and non-Roma alike, take longer to reach their chosen primary drinking water sources than do their rural counterparts. However, this statistic is potentially ambiguous. The urban community included in this study reported overwhelmingly that it had recently been subject to a contamination of the municipal water supply with coliform bacteria and, thus, the majority of residents therein reported the need to purchase water rather than use the taps available in their homes. It was not possible to collect data regarding the behaviour of these residents prior to the contamination event. Furthermore, the suburban community included here recently experienced the loss of a bridge, connecting the far side of the river to the village centre on the other side. Those individuals stranded on the far side of the bridge (predominantly Roma) reported numerous problems with their wells, requiring them to travel 5km or more to the nearest crossing to reach a shop or market until the bridge is restored. Therefore, this statistic is potentially a reflection of the walking or driving time that would otherwise be unnecessary.\n\nDespite the evidence presented that Roma and non-Roma alike are subjected to ineffective sanitation and hygiene services throughout the country, one should note that the Roma population still reports a greater incidence of diarrheal disease and a reduced rate of immunization than the non-Roma population. There are potentially a number of reasons for this. Unlike in other countries5,24, the Romanian Roma report fairly equivalent rates of medical insurance and access to primary care, but the type of treatment received when care is sought was beyond the scope of this study and may be a contributing factor. Indeed, Roma individuals have elsewhere reported poor health related to both their unhygienic circumstances and the care they receive19,26,27. In addition, as has already been noted, both literacy rates and overall levels of education are significantly decreased in the Romanian Roma population. This is in contrast to education rates in Roma populations of other countries, as the educational component of the DRI has been lauded as the most successful portion of the initiative, albeit only for primary school attendance20,21. Rates of disease and healthcare status overall are inversely associated with education28, which may offer another possible explanation for the disparity in diarrheal disease rates. It is important to consider, however anecdotally, the Roma do report some knowledge of personal water treatment and safety (data not shown), through the use of salt or lime in personal wells and a commitment to boiling water before drinking or cooking if possible. However, the lack of infrastructure and services works against these individual and imperfect efforts. Furthermore, for those Roma who do have access to tap water (municipal or otherwise), many of them report using an alternative primary water source. While these same individuals indicate that they believe their tap water to be safe (data not shown), their daily activities are in direct contrast to this assertion.\n\nOverall, while these data demonstrate the ongoing challenges following the Decade of Roma Inclusion as applied to the Roma population in Romania, this study also points to the myriad of ways in which all Romanians, regardless of ethnicity, are encountering challenges. It highlights the areas in which improvements can be made to ensure all Romanian citizens have access to and confidence in basic sanitation services, clean water, and adequate healthcare treatment.\n\nThe primary limitation to this study is the sample size of 135 individuals. Time and funding constraints, as well as limited personnel, inhibited our ability to interact with more than 30 households per community and restricted the study to five communities. Future efforts will expand the population included in similar studies by increasing the number of communities engaged, and will seek to enroll equal numbers of Roma and non-Roma. Additionally, subsequent studies can use these and other data to generate detailed models that explore specific initiatives that could be implemented to address discrepancies in equality and access, and progress the literature around Roma health disparities beyond analysis and into intervention testing.\n\n\nData availability\n\nDataset 1: Coded survey data. Romania, 2016. Excel file of compiled responses to survey questions. Coded and de-identified. Numerical code corresponds to responses as indicated on the study surveys (Supplementary material 1 and Supplementary material 2).\n\nDOI, 10.5256/f1000research.12546.d17723329\n\nDataset 2: Python Notebook data analysis and statistics. Romania, 2016. Python Notebook analysis of survey data.\n\nDOI, 10.5256/f1000research.12546.d17723430\n\nDataset 3: Python Notebook data analysis and statistics. Romania, 2016. Python Notebook analysis of survey data, exported as a PDF file.\n\nDOI, 10.5256/f1000research.12546.d17723531", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work has been partially supported by the National Institutes of Health and National Institute of General Medical Sciences - Models of Infectious Disease Agent Study Grant 5U01GM070694-13, the Defense Threat Reduction Agency - Comprehensive National Incident Management System Contract HDTRA1-11-D-0016-0001, and the Virginia-Maryland College of Veterinary Medicine.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nWe would like to thank Dr. Ralph P. Hall for his assistance developing the surveying instrument, and all the staff members at Agentia Impreuna for helping to make this project possible. We thank our external collaborators and members of the Network Dynamics and Simulation Science Laboratory (NDSSL) for their suggestions and comments. In addition, we are grateful to the numerous friends and family who contributed financially to support our time in the field, with particular thanks to George K. Agardi, Sr. and Dr. Susan Evans.\n\n\nSupplementary material\n\nSupplementary material 1: Quality of Life Survey, English. Romania 2016. Survey questions provided in English.\n\nClick here to access the data.\n\nSupplementary material 2: Quality of Life Survey, Romanian. Romania, 2016. Survey questions provided in Romanian.\n\nClick here to access the data.\n\n\nReferences\n\nEuropean Investment Bank: EIB loan signatures in Romania amounted to EUR 211 million in 2015. 2015.\n\nUnited Nations Development Programme: Millennium Development Goals in Romania. 2007.\n\nEuropean Union: European parliament resolution of 31 January 2008 on a European strategy on the Roma. 2008. Reference Source\n\nWorld Bank: Roma. 2015. Reference Source\n\nCook B, Wayne GF, Valentine A, et al.: Revisiting the evidence on health and health care disparities among the Roma: a systematic review 2003–2012. Int J Public Health. 2013; 58(6): 885–911. PubMed Abstract \| Publisher Full Text\n\nHajioff S, McKee M: The health of the Roma people: a review of the published literature. J Epidemiol Community Health. 2000; 54(11): 864–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZeman CL, Depken DE, Senchina DS: Roma health issues: a review of the literature and discussion. Ethn Health. 2003; 8(3): 223–49. PubMed Abstract \| Publisher Full Text\n\nParekh N, Rose T: Health inequalities of the Roma in Europe: a literature review. Cent Eur J Public Health. 2011; 19(3): 139–42. PubMed Abstract\n\nGladdis K: Life inside the Romanian gypsy ghetto that is so grim the town mayor sealed it off behind a wall. Daily Mail. 2013. Reference Source\n\nNacu A: From silent marginality to spotlight scapegoating? A brief case study of France’s policy towards the Roma. J Ethn Migr Stud. 2012; 38(8): 1323–8. Publisher Full Text\n\nFriedman E: Decade of Roma Inclusion Progress Report 2005–2013. 2014. Reference Source\n\nFlecha A: Healthier lives for European minority groups: school and health care, lessons from the Roma. Int J Environ Res Public Health. 2013; 10(8): 3089–111. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFésüs G, Östlin P, McKee M, et al.: Policies to improve the health and well-being of Roma people: the European experience. Health Policy. 2012; 105(1): 25–32. PubMed Abstract \| Publisher Full Text\n\nGould R: Roma rights and Roma expulsions in France: Official discourse and EU responses. Critical Social Policy. 2015; 35(1): 24–44. Publisher Full Text\n\nWorld Health Organization: Core Questions on Drinking-Water and Sanitation for Household Surveys. 2006. Reference Source\n\nRivers C: Modeling Emerging Infectious Diseases for Public Health Decision Support. Blacksburg, VA: Virginia Polytechnic Institute and State University; 2015. Reference Source\n\nMcKinney W, editor: Data Structures for Statistical Computing in Python. 9th Python in Science Conference; SciPy. 2010. Reference Source\n\nCurcic S, Miskovic MA, Plaut S, et al.: Inclusion, Integration or Perpetual Exclusion? A Critical Examination of the Decade of Roma Inclusion, 2005–2015. European Educational Research Journal. 2014; 13(3): 257–67. Publisher Full Text\n\nMasseria C, Mladovsky P, Hernández-Quevedo C: The socio-economic determinants of the health status of Roma in comparison with non-Roma in Bulgaria, Hungary and Romania. Eur J Public Health. 2010; 20(5): 549–54. PubMed Abstract \| Publisher Full Text\n\nBojadjieva A: Roma Inclusion Index. Budapest, Hungary: Decade of Roma Inclusion Secretariat Foundation; 2015. Reference Source\n\nJovanovic Z: Why Europe’s “Roma Decade” Didn’t Lead to Inclusion. In: Foundations OS, editor. 2015. Reference Source\n\nMolnár A, Adány R, Adám B, et al.: Health impact assessment and evaluation of a Roma housing project in Hungary. Health Place. 2010; 16(6): 1240–7. PubMed Abstract \| Publisher Full Text\n\nJanevic T, Jankovic J, Bradley E: Socioeconomic position, gender, and inequalities in self-rated health between Roma and non-Roma in Serbia. Int J Public Health. 2012; 57(1): 49–55. PubMed Abstract \| Publisher Full Text\n\nKuhlbrandt C, Footman K, Rechel B, et al.: An examination of Roma health insurance status in Central and Eastern Europe. Eur J Public Health. 2014; 24(5): 707–12. PubMed Abstract \| Publisher Full Text\n\nDuminica G: Raport Anual, 2015. Bucharest, Romania: Agentia Împreunã; 2015.\n\nCasals M, Pila P, Langohr K, et al.: Incidence of infectious diseases and survival among the Roma population: a longitudinal cohort study. Eur J Public Health. 2012; 22(2): 262–6. PubMed Abstract \| Publisher Full Text\n\nTeira R, Suárez-Lozano I, Lozano F, et al.: Characteristics and outcome of HIV infection in gypsies in the Spanish VACH Cohort. Enferm Infecc Microbiol Clin. 2010; 28(5): 266–72. PubMed Abstract \| Publisher Full Text\n\nRegidor E, De Mateo S, Calle ME, et al.: Educational level and mortality from infectious diseases. J Epidemiol Community Health. 2002; 56(9): 682–3. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPowell Doherty R, Müller-Demary D, Hosszu A, et al.: Dataset 1 in: A survey of quality of life indicators in the Romanian Roma population following the ‘Decade of Roma Inclusion’. F1000Research. 2017. Data Source\n\nPowell Doherty R, Müller-Demary D, Hosszu A, et al.: Dataset 2 in: A survey of quality of life indicators in the Romanian Roma population following the ‘Decade of Roma Inclusion’. F1000Research. 2017. Data Source\n\nPowell Doherty R, Müller-Demary D, Hosszu A, et al.: Dataset 3 in: A survey of quality of life indicators in the Romanian Roma population following the ‘Decade of Roma Inclusion’. F1000Research. 2017. Data Source" }	[ { "id": "26651", "date": "06 Nov 2017", "name": "Ciprian Marius Ceobanu", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe main issue is related to the size of the sample and its geographical distribution. It is hard to generalize over the entire Roma minority the conclusions of the study even if the statistical approach is appropriate.\nThe geographical distribution of the Roma population is pretty different over the Romanian national territory. The lack of indication of the geographical area of the subjects of the sample is a flaw. Another issue regards the random walk method for sampling which, in our opinion, is not representative for the entire Roma population. Maybe a multilevel sampling would be more appropriate than a simple random walk.\nDespite the fact that the conclusions of the study are correct, these are pretty well known to all levels; also these are commonplaces that were specified in the documents of Decade of Roma Inclusion as directions for future action. There are a lot of significant reports that draw the same conclusions (see The World Bank documents for instance1). From this point of view, there is no original approach to the Roma problem in Romania.\nThe conclusion following the Decade of Roma Inclusion is that despite the efforts that there were made, there still remain lots of issues regarding the integration of the Roma population. Also,solving great structural problems of Romania will certainly improve the Roma population situation.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] }, { "id": "26652", "date": "12 Dec 2017", "name": "Martin McKee", "expertise": [ "Reviewer Expertise Public Health – and have written extensively on Roma health" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe introduction is well written and provides a good overview of the situation facing the Roma population. There are a few more recent references that could be included, such as an evaluation of the Decade of Roma inclusion in Hungary but, in general, the authors have found most of the relevant information.\nThe fundamental challenge facing anyone doing research among the Roma population in this region is how to develop a sampling frame. There are numerous methodological problems, in particular varying degrees of assimilation (see, for example, work by K Kosa). Previous studies, such as that by the UNDP or in Hungary, have used Roma communities, identifiable by their socio economic and physical characteristics, while recognising that this is imperfect. However, this paper would benefit from a more detailed description of the communities from which the samples were drawn, in particular, how they relate to Romania as a whole. Given that, in many parts of Romania, Roma live in distinct settlements, separate from the Romanian population, even within individual villages, could the authors comment on any implications that their sampling strategy had for generalisability?\nGiven the high levels of distrust that many Roma, justifiably, have, some studies have sought to ensure involvement of Roma fieldworkers, or at least, involvement of community leaders. Can the authors comment on what measures they took in this regard?\nThe greatest problem in this paper is the very small sample size. Overall, less than 100 Roma respondents were included and only 37 non-Roma. Given the numerous problems involved in sampling in a study such as this, this is really far too few from which to draw any meaningful conclusion. This is noted in the limitations but I’m not really convinced that a study of this size can be regarded as much more than a pilot. I would suggest that it is described in this way, with many more caveats than there are at present.\nMinor points:\nI’m not sure that it is appropriate to use the words of Soviet rule for the countries of south-eastern Europe. Arguably, Romania was one of the most independent of the Soviet bloc states.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] }, { "id": "28652", "date": "02 Jan 2018", "name": "Pilar Carrasco-Garrido", "expertise": [ "Reviewer Expertise Public health" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a relevant manuscript from a public health standpoint because one of the main contributions of the present work is to determine quality of life indicators in the Romanian Roma, but methodologically it has significant shortcomings:\nThe comparison between the population of Burkina Faso and the population of Romania is not adequate. They are very different populations. This aspect is a methodological problem.\n\nDiscussion of the results is limited. Some aspects are not adequately discussed.\n\nThere is no information about the survey's non-response rate.\n\nFew references and some of them unrelated to the purpose of the study. It does not seem correct to incorporate a press article as a reference.\nSorry but, I do not recommend the indexing of the article.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/6-1692
https://f1000research.com/articles/7-1926/v1	11 Dec 18	{ "type": "Opinion Article", "title": "The principles of tomorrow's university", "authors": [ "Daniel S. Katz", "Gabrielle Allen", "Lorena A. Barba", "Devin R. Berg", "Holly Bik", "Carl Boettiger", "Christine L. Borgman", "C. Titus Brown", "Stuart Buck", "Randy Burd", "Anita de Waard", "Martin Paul Eve", "Brian E. Granger", "Josh Greenberg", "Adina Howe", "Bill Howe", "May Khanna", "Timothy L. Killeen", "Matthew Mayernik", "Erin McKiernan", "Chris Mentzel", "Nirav Merchant", "Kyle E. Niemeyer", "Laura Noren", "Sarah M. Nusser", "Daniel A. Reed", "Edward Seidel", "MacKenzie Smith", "Jeffrey R. Spies", "Matt Turk", "John D. Van Horn", "Jay Walsh", "Gabrielle Allen", "Lorena A. Barba", "Devin R. Berg", "Holly Bik", "Carl Boettiger", "Christine L. Borgman", "C. Titus Brown", "Stuart Buck", "Randy Burd", "Anita de Waard", "Martin Paul Eve", "Brian E. Granger", "Josh Greenberg", "Adina Howe", "Bill Howe", "May Khanna", "Timothy L. Killeen", "Matthew Mayernik", "Erin McKiernan", "Chris Mentzel", "Nirav Merchant", "Kyle E. Niemeyer", "Laura Noren", "Sarah M. Nusser", "Daniel A. Reed", "Edward Seidel", "MacKenzie Smith", "Jeffrey R. Spies", "Matt Turk", "John D. Van Horn", "Jay Walsh" ], "abstract": "In the 21st Century, research is increasingly data- and computation-driven. Researchers, funders, and the larger community today emphasize the traits of openness and reproducibility. In March 2017, 13 mostly early-career research leaders who are building their careers around these traits came together with ten university leaders (presidents, vice presidents, and vice provosts), representatives from four funding agencies, and eleven organizers and other stakeholders in an NIH- and NSF-funded one-day, invitation-only workshop titled \"Imagining Tomorrow's University.\" Workshop attendees were charged with launching a new dialog around open research – the current status, opportunities for advancement, and challenges that limit sharing. The workshop examined how the internet-enabled research world has changed, and how universities need to change to adapt commensurately, aiming to understand how universities can and should make themselves competitive and attract the best students, staff, and faculty in this new world. During the workshop, the participants re-imagined scholarship, education, and institutions for an open, networked era, to uncover new opportunities for universities to create value and serve society. They expressed the results of these deliberations as a set of 22 principles of tomorrow's university across six areas: credit and attribution, communities, outreach and engagement, education, preservation and reproducibility, and technologies. Activities that follow on from workshop results take one of three forms. First, since the workshop, a number of workshop authors have further developed and published their white papers to make their reflections and recommendations more concrete. These authors are also conducting efforts to implement these ideas, and to make changes in the university system. Second, we plan to organise a follow-up workshop that focuses on how these principles could be implemented. Third, we believe that the outcomes of this workshop support and are connected with recent theoretical work on the position and future of open knowledge institutions.", "keywords": [ "open science", "open scholarship", "academia", "reproducibility", "data science", "credit", "education" ], "content": "Summary\n\nIn the 21st Century, research is increasingly data- and computation-driven. Researchers, funders, and the larger community today emphasize the traits of openness and reproducibility. In March 2017, 13 mostly early-career research leaders who are building their careers around these traits came together with ten university leaders (presidents, vice presidents, and vice provosts), representatives from four funding agencies, and eleven organizers and other stakeholders in an NIH- and NSF-funded one-day, invitation-only workshop titled “Imagining Tomorrow’s University.” Workshop attendees were charged with launching a new dialog around open research – the current status, opportunities for advancement, and challenges that limit sharing.\n\nThe workshop examined how the internet-enabled research world has changed, and how universities need to change to adapt commensurately, aiming to understand how universities can and should make themselves competitive and attract the best students, staff, and faculty in this new world. During the workshop, the participants reimagined scholarship, education, and institutions for an open, networked era, to uncover new opportunities for universities to create value and serve society. They expressed the results of these deliberations as a set of 22 principles of tomorrow's university across six areas: Credit and Attribution (A), Open Scholarship Communities (C), Outreach and Engagement (O), Education (E), Preservation and Reproducibility (P), and Technologies (T):\n\nCredit and Attribution (A)\n\nA1 Stakeholders (funders, universities, and researchers) should incentivize credit and attribution for a diverse range of research products and activities, such as research software, dissemination, infrastructure, data products and repositories.\n\nA2 Research should be assessed on its own merits, not based on its appearance in exclusive publication venues. This could be through the use of article-level metrics or narrativized impact measures. Journal impact factors should not be used to evaluate researchers.\n\nA3 Institutions that comparatively measure attention scores should evaluate attention measures with care, since such altmetrics may not correlate with measures of quality.\n\nA4 Institutions should provide appropriate career paths for staff working on open research and maintain the institutional infrastructure required for open research, including recognizing and valuing new, emergent forms of digital outcomes, such as software and data creation, curation and preservation, that are crucial to open research endeavors. These pathways may require rethinking existing classifications and assessments of tenure-track, non-tenure-track, and staff categories of university participants, and funder support for personnel in these categories.\n\nOpen Scholarship Communities (C)\n\nC1 Scholarly communities are both the target and the product of open scholarship and open research. Therefore, the fostering of communities is a key driver for open research.\n\nC2 These communities can take many forms. They may, for example, coalesce around tools, practices, shared interests, shared data or software, or shared hashtags. They can be short- or long-lived, explicitly funded, or emerge organically from shared interests among the participants. All of these variations are valuable components in the open research social ecosystem, and they must be supported as such by multiple resources, including travel funds, virtual networks, compensations for networking events, etc.\n\nC3 Community activities often serve to evaluate, encourage, and improve the use of tools, software, platforms, and data that are critical to open scholarship. All stakeholders must take steps to encourage these communities to develop, such as supporting common standards (and rewarding those who work on them), funding projects that form a “connective tissue” between different communities. They should also actively encourage sharing practices for tools, and people across communities.\n\nOutreach and Engagement with the Public (O)\n\nO1 Outreach and engagement, for public access to and understanding of research outcomes, depend on the audience and may encompass products, processes, and dissemination.\n\nO2 Universities and researchers share a mutual interest in interactions with diverse audiences.\n\nO3 Institutions should value, recognize, and support researchers who participate in outreach and engagement activities, crediting them as components of \"service\" to the institution.\n\nO4 Access to researchers (scholars) builds trust in the products and processes of research (scholarship).\n\nEducation (E)\n\nE1 Every student should be guided to understand and learn how to access and use data and software to be well prepared for diverse, modern career paths.\n\nE2 Through open research training, universities should play an active role in increasing research by enabling evidence-based decisions, accelerating discovery, and extending impact to broader communities.\n\nE3 Universities should encourage their faculty to engage in open educational practices, including creating and assigning open educational resources, and reflecting open culture in their courses.\n\nPreservation and Reproducibility (P)\n\nP1 The scholarly publication and communication ecosystem should support open and reproducible research, and enable credit for these efforts. Universities should encourage these initiatives by creating incentives (e.g., promotion and tenure categories, service recognition) for such activities.\n\nP2 Research funders should support open and reproducible research by making reproducibility part of their merit review criteria. They should also create new scholarly communication venues or support open scholarship efforts, and encourage, require, and reward reproducible research efforts.\n\nP3 Incentives that promote the public sharing and distribution of scholarly knowledge for open/transparent/reproducible research practices must be put in place. Publishers must require, when appropriate, submissions that provide open and reproducible workflows, and embed this requirement in their own workflows.\n\nP4 Universities should recognize the activities of faculty to educate and train researchers on open and reproducible research skills. Global and national bodies (e.g., National Academies) should promote this recognition across universities.\n\nTechnologies (T)\n\nT1 Open source technologies, tools and platforms provide intrinsic value to researchers and educators and are an effective way of accelerating open scholarship. Academic institutions should favor and encourage open source solutions as much as possible.\n\nT2 A diverse and interoperable set of tools for open research should be known, shared, and clearly documented.\n\nT3 Institutions should provide and support foundational open scholarship infrastructure (technological and human) for all members of campus.\n\nT4 Institutions should recognize the contributions made by all members of campus to open scholarship infrastructure.\n\nActivities that follow on from workshop results take one of three forms. First, since the workshop, a number of workshop authors have further developed and published their white papers to make their reflections and recommendations more concrete. These authors are also conducting efforts to implement these ideas, and to make changes in the university system. Second, we plan to organise a follow-up workshop that focuses on how these principles could be implemented. Third, we believe that the outcomes of this workshop support and are connected with recent theoretical work on the position and future of open knowledge institutions.\n\n\n1 Introduction\n\nThe culture of today’s research universities is built from elements of past universities, including the medieval roles and responsibilities of faculty, the 18–19th century structure of departments, and the mid-20th century research funding model. While for most of history education was reserved for elites, educational access in the United States has been repeatedly expanded, beginning with creation of state-funded universities in the 18th century, and the later creation of land-grant institutions in the 19th century. In the 20th century, this democratization of education was accelerated by the post-World War II GI Bill. Today, higher education is seen as necessary for the majority of young adults to succeed in the growing knowledge economy.\n\nAcademic research is also changing. Research practices have been evolving rapidly, based largely on advances in computing capability and capacity. Examples include the use of computational and data science as new research methods, and concomitant changes in research culture, such as vastly increased sharing via the Internet, and collaboration via open source software, open science, and more generally, open scholarship. Simply put, research and scholarship are increasingly data- and computation-driven. At the same time, researchers, funders, and the larger community increasingly emphasize the traits of openness and reproducibility. While the idea of the openness has long been a part of research institutions, new technologies amplify the expectations around openness because the general distribution of information, and the creation of virtual/distributed collaborations (e.g. open source communities), are much easier technically via the internet.\n\nIn March 2017, in response to this changing environment, 13 early-career research leaders who are building their careers around these traits came together with 10 university leaders (presidents, vice presidents, and vice provosts), representatives of four funding agencies, and 11 organizers and other stakeholders in an NIH- and NSF-funded one-day, invitation-only workshop called “Imagining Tomorrow’s University.” Workshop attendees were charged with launching a new dialog around open research: current status, opportunities for advancement, and challenges limiting sharing.\n\nThe workshop addressed the changing nature of research and the associated shifts in university needs. Some guiding questions at the workshop included:\n\nHow should the university change?\n\nHow should it adapt its structure, mission, infrastructure, education, and recruitment plans?\n\nDo we need new educational programs?\n\nDo we need new disciplines or new departments?\n\nHow can universities recognize the value in new types of research products such as software and data?\n\nDoes research staffing need to change?\n\nDo research data engineers or research software engineers have a place in modern scholarship?\n\nWhat are different measures of success for faculty active in open science/open research?\n\nThese issues can be summarized as seeking to understand how universities can and should maximize their competitiveness in attracting the best students, staff, and faculty, and better serve the public. In the workshop, the participants tried to reimagine scholarship, education, and institutions for an open, networked era, to discover new opportunities for universities to create value and serve society, expressed as a set of principles.\n\nThe workshop included 37 participants and two facilitators (list of attendees can be found in the Appendix). The participants worked in six areas:\n\ncredit and attribution,\n\ncommunities,\n\neducation,\n\noutreach and engagement,\n\npreservation and reproducibility, and\n\ntechnologies,\n\neach of which came to consensus on a set of principles. In addition, the participants discussed three cross-cutting topics: libraries, career paths, and campus IT organizations & data repositories.\n\n\n2 Workshop inputs\n\nIn total, 12 of the 14 invited university researchers submitted white papers as their inputs to the workshop, as summarized in Section 2.1. In addition, 12 invited university leaders completed a brief survey, with results summarized in Section 2.2. (One university researcher who submitted a white paper and two university leaders who completed the survey did not attend the workshop due to conflicts that arose at the last minute.)\n\nBerg1 presents personal reflections on the role that open research can play in defining the purpose and activities of the university. He includes specific recommendations on how the public university can recommit and push the boundaries of its role as the creator and promoter of public knowledge, serving a vital role to the continued economic, social, and technological development of society. The recommendations are: requiring that research products be made openly available, supported via institutional repositories and copying policies; converting technology commercialization offices into research impact offices; empowering and funding university libraries to support open knowledge dissemination; and infusing open knowledge dissemination and best practices into education. Berg also includes some thoughts on how this applies specifically to the field of engineering and how a culture of openness and sharing within the engineering community can help drive societal development.\n\nBik2 defines open science as “any type of scientific research effort that is freely available and publicly accessible.” This includes “both traditional research products (peer-reviewed publications, underlying datasets) as well as non-traditional initiatives and products (blog posts, slide decks, course syllabi and materials, scientific software, as well as analysis scripts and code).” She says that making science open is important because research is mostly taxpayer-funded; open science is more accountable and more reproducible, more democratic and more accessible, and helps to build a scientist’s reputation. Open science also impacts society by making scientists and their work more visible and more accessible to the public and to policymakers, and more visible to each other. She also discusses some challenges: how open scientists can find each other at a university; that open science has a cost in terms of increased work and reporting systems that don’t support it; merit and promotion policies that don’t recognize or reward open science; and a lack of guidelines and support from the university.\n\nBoettiger3 defines open science as “just science without the barriers created by other incentives.” He says open science has four main pillars: open access (including concerns about paywalls and licenses), open data, open code, and open context, all of which he supports in his own research practices via preprints, data publishing, software publishing and containers, and an open lab notebook. He also teaches about open science, and makes his teaching materials open. He believes that there are social challenges in changing the underlying incentive structure for research, technical challenges in developing solutions that help researchers realize the benefits from open science practices rather than just the costs, and educational challenges that if met could develop the next generation of researchers who implement the needed social and technical changes.\n\nBrown4 thinks Twitter and blogging are integral to the pursuit of open science, and he blogged answers to the workshop questions, including defining open science both as “the philosophical perspective that sharing is good and that barriers to sharing should be lowered as much as possible” and as the practice of lowering the barriers. He says that open science should drive science forward faster, increase its societal impact, open opportunities for serendipity, and “aid with reproducibility and replication, decrease the effects of economic inequality in the sciences by liberating ideas from subscription paywalls, and providing reusable materials for teaching and training.” He believes that “while most scientists are supportive of open science in theory, in fact most scientists are leery of actually sharing things widely before publication,” because of the incentive systems in place. He says that prominent senior researchers need to “visibly and loudly abandon the broken ‘journal prestige’ system, forcefully push back against university administration on matters of research evaluation and tenure, and be a loud presence on grant panels and editorial boards.”\n\nEve5 says that while for published research work to be “open,” it must be free to read online and free to re-use, open research also concerns the practices of academia opening itself to inspection and collaboration in new ways. Without research being open, research institutions such as universities are not “woven into the tapestry of modern citizenship,” researchers cannot fully pursue new knowledge, and replication and rigor are problematic. Eve is the youngest full professor of English Literature in the UK, and he believes this is due in part to his service and charitable activities, such as being a founder and co-CEO of the Open Library of Humanities, a charitable/not-for-profit open access publishing company that funds or publishes 27 fully open access, zero-fee-for-authors journals. It is supported on an ongoing basis by an international consortium of over 220 (and growing) academic libraries. He describes a range of social, technical, and economic challenges in the implementation of open research, and suggests a set of university changes that could address them, including: not using journal-level or press-level metrics, policies that promote open research practice, strong local green open access policies, and university presses moving to open access as dissemination vehicles.\n\nHowe et al.6 believe that “increased transparency in the scientific process can broaden and deepen scientific inquiry, understanding, and impact,” but that this is not quick, effortless, or easy. They propose “that open science can most effectively enable this evolution when it is conceptualized as a multifaceted pathway that includes: the provision of accessible and well-described data, along with information about its context, the methodology and mechanisms necessary to reproduce data analyses, and training products that provide transparent understanding of how the data can be applied to answer questions.” They suggest that doing this often requires investments by researchers, and that changes should be carefully planned across the entire university to avoid unintended consequences.\n\nHowe and Grechkin7 consider open science to be, “a movement to bring the incentives that drive science back in line with the stated values of science,” practices, norms, and tools that reward the sharing of knowledge in addition to the creation of knowledge. They believe that this movement is progressing, but also suggest considering an alternative, “a more transformative vision for wide open science.” Wide open science means 1) wide open experiments, where each experiment consists of a pre-registered hypothesis, a visualization of the result, enough text to interpret and understand it, and the code, data, and environment needed to recreate it; 2) wide open data, supported by tools that automatically curate, integrate, clean, and standardize available data for reuse and reproducibility; and 3) wide open publishing, where overlay journals superimpose a journal-like structure on open access materials to “reduce publication friction while enabling community-driven peer review and curation.”\n\nKhanna8 defines open science as “the dissemination of research in any open forum, publicly available for all to access.” She believes that this will improve research and learning, ensure transparency, and foster more collaborative research. She suggests that, “it is our duty and moral obligation to inform the public about the work being done in our research laboratories,” in part to generate interest and maintain funding. However, to do this, for example, to publish data in an open forum, takes time away from traditional research, teaching, and service. Tenure evaluation often relies mainly on publications and funding, and open science may not be seen as a contribution. If openness were rewarded within the tenure process, perhaps as an expected part of research, teaching, and service, it would increase.\n\nMayernik9 suggests that “the movements by national governments, funding agencies, universities, and research communities toward ‘open data’ face many difficult challenges.” He believes that this is because, “researchers” data and metadata practices are expected to be robust and structured,” but that they are not. This is in part because researchers are expected to be good at research, not good at depositing data or creating metadata, and because, “making data open in a transparent way can involve a significant investment of time and resources with no obvious benefits [to the researcher].” He relates the concepts of accountability and transparency with researcher actions, and suggests that achieving them is an ongoing process, not the results of one-time acts.\n\nMcKiernan10 discusses open scholarship, such as the sharing of articles, code, data, and educational resources, as having the potential to improve or even transform university research and education, and to increase the external impact of universities. She presents numerous case studies and her own personal experiences as a practicing open scholar. Tension is created by incompatibilities between institutional policies and personal practice in many forms of academic evaluation. She proposes actions universities could take to support open scholarship, and explains their benefits. She says, “I do not think most of these actions would require new funding, but rather a redistribution of existing funds and a rewriting of internal policies to better align with university missions of knowledge dissemination and societal impact.”\n\nNiemeyer11 defines open research as “the activity of performing scientific research in a manner that makes products and findings accessible to anyone. This includes sharing data openly (open data), publicly releasing the source code for research software (open source software), and making the written products of research openly accessible (open access).” He believes it is important because of six benefits it supports: accessibility, reproducibility, impact, establishing priority, encouraging trust, and being nice. He has created an open policy for his group’s research. Niemeyer says, “the challenges impeding greater adoption of open science practices are mainly institutional and cultural, rather than technical,” and he makes four recommendations for universities to overcome the challenges: 1) Tenure and promotion should consider the accessibility/openness of research products along with their quantity and “quality”; 2) recognize research products such as software and data as equal to traditional publications in scholarly impact; 3) recognize that publishing in traditional venues may hinder openness, so reduce their importance for promotion and tenure; and 4) support efforts to teach undergraduate and graduate students open science skills, and those necessary to work with software and data, with the same enthusiasm that traditional lab courses receive.\n\nSengupta and Shanahan12 talk about opening the practice and ongoing work of science, rather than its products. They suggest that we engage “the public in the dynamic, conceptual and representational work involved in creating scientific knowledge.” They propose “public computing spaces, a genre of open-ended, public learning environment where visitors interact with open source computing platforms to directly access, modify and create complex and authentic scientific work” as a possible model of open science in the university.\n\nSome common themes of all white papers are:\n\nOpen scholarship is perhaps the most broad term we can use; it includes open science, open humanities, and open research, and can be defined as opening products such as articles, data, software, educational resources, or more broadly, opening the process of scholarship.\n\nCosts and benefits for scholars:\n\n– Researchers respond to how they are evaluated, which today mostly does not reward open scholarship.\n\n– Sharing can lead to increased progress and knowledge, but can have a cost when it is not rewarded.\n\nBenefit to society:\n\n– Openness can reduce the negative image of the university as an ivory tower.\n\n– Most research funding comes from the public, and they should be given access to the research outputs that they have supported.\n\n– If we can involve the public in the whole process and not just the outputs, they may have more appreciation of scholarship and the scientific process.\n\nWhile parts of the university have the ostensible goal of disseminating research (e.g., university press, technology transfer office), they are often siloed into centers that are measured on financial return.\n\nThe results of the survey given in advance of the workshop to the university leaders in attendance are presented next.\n\n1. To what extent is research at your university becoming substantially more data- and computation-driven?\n\n2. How important are the open science themes of sharing and reproducibility to your university’s researchers?\n\n3. How much investment is your university making to integrate open science into the research environment and curriculum?\n\n4. What are the most important opportunities presented by open science for your university? (First number is how many of the 12 respondents chose this item)\n\n11 To gain access to the data resources necessary for research\n\n9 To gain access to the software and tools necessary for research\n\n9 To improve discovery processes\n\n7 To gain access to the computational and storage infrastructure necessary for research\n\n7 To increase industrial relationships and partnerships\n\n6 To improve educational outcomes\n\n5 To increase funding opportunities\n\n5 To recruit students and postdocs\n\n5 To increase recognition/rating of the university\n\n4 To recruit faculty\n\n3 Creating new curriculum to prepare students for careers in open science across different sectors\n\n2 To improve pedagogical material and apply best practices for developing curricula\n\n5. What are the most important challenges presented by open science for your university? (First number is how many of the 12 respondents chose this item)\n\n10 To change the work culture of existing faculty and researchers\n\n8 To reward open science work in tenure and promotion processes\n\n6 To enhance library services for data curation and sustainability\n\n5 To improve licensing, ownership, and other legal practices for open science\n\n5 To develop technology infrastructure and staffing for open science\n\n4 To create pedagogical material for open science curricula\n\n3 To recruit faculty with experience in open science\n\n3 Providing new teaching programs and training in open science\n\n3 To develop a workforce for sustaining access to data and software\n\n2 To recruit and retain non-tenure track faculty and staff\n\n6. What is an open science success story that you find compelling?\n\nThe respondents mentioned:\n\nCyberinfrastructure to support data sharing and collaboration, such as CyVerse (formerly iPlant) in the life sciences; the Southern California Earthquake Center (SCEC) initially funded as an NSF STC but with ongoing collaboration among 22 core institutions studying impacts of earthquakes on California; the INSPIRE platform in high energy physics which facilitates gatherings of scientists to review data, discuss new or expanded findings and then collaborate on publication; or at one institution a medical electronic data warehouse, with over six million de-identified patient records available for research across the university.\n\nInclusion of the public and schools in science around open data and platforms, leading for example to the discovery of supernovae by elementary school children13.\n\nThe combination of the private sector with federal investments leading to new landscapes for knowledge creation, as in genomics.\n\nA European Union-like model for funding open collaborative spaces integrating people, publications, data, tools in a seamless manner, through a small tax on grants that then allows open shared data to be hosted for free and restricted access data to be hosted with a fee.\n\n7. How can the outcomes of this workshop help your relationship with your stakeholders (such as your board of trustees or alumni)?\n\nThe respondents mentioned:\n\nDevelop an authoritative report on the value of open science for the university stakeholders (both internal and external), including key points on current trends showing that this is a critical shift in higher education; exemplar outcomes showing the importance for discovery, current adoption issues, and recommendations/narratives/specific implementation strategies for institutions wishing to embrace open science/research.\n\nDevelop principles for supporting open science to help universities attract and grow students and young faculty as well as industry and government partners.\n\nConnect open science to the topic of reproducibility and rigor/transparency in researcha\n\nWhile internal stakeholders (faculty, postdoctoral researchers, students) were generally seen as important, for universities dealing with classified and sensitive information, there is a responsibility and opportunity to educate their boards about data and open science.\n\n8. What other related issues are important that we should discuss at the workshop?\n\nThe respondents mentioned:\n\nThe federal government has a diversity of requirements for data sharing, and a diversity of financial models (not just the commonly encountered unfunded mandate.)\n\nThe international landscape and international collaboration threats.\n\nIs there a way to articulate both intellectual and monetary value of open science to a university?\n\nReward and recognition structures for those working primarily on data, algorithms and computational models. Tenure and promotion policies and their implementation.\n\nWho has responsibility to curate the data (not just to store it)? For example, who develops the metadata that allows one to best use the data?\n\n9. If you want to elaborate on any of the answers above, or provide any other inputs, please do so here\n\nThe respondents mentioned:\n\nAt the moment, there is stronger interest than there is investment or implementation at our university. This is primarily because much of the infrastructure and appropriate research practice needs to be developed, and this path is not completely understood.\n\nOpen science works best when communities of researchers develop standards for data sharing. Simply making data available is insufficient. There needs to be national leadership to ensure common data standards and shared libraries.\n\n\n3 Workshop agenda\n\nThe workshop began with an introductory dinner, where attendees had the opportunity to meet each other, learn a little about each other’s backgrounds, and talk about what they wanted to get out of the workshop. The next morning began with brief remarks intended to set the stage for the workshop: by Ed Seidel on behalf of the university leaders; by Dan Katz on behalf of the organizers; and by Stuart Buck and Rajiv Ramnath (remotely) on behalf of the funders. The group then discussed how to divide up the topics, deciding to organize around\n\nCredit and attribution\n\nCommunities\n\nEducation\n\nOutreach and engagement\n\nPreservation and reproducibility, and\n\nTechnologies\n\nwhile recognizing that there would be overlaps and cross-cutting issues.\n\n\n4 Discussion topics and principles\n\nMost of the remaining time at the workshop was spent with the participants divided into groups discussing the topics mentioned above. The groups generally discussed the assigned topic, and typically identified a small number of principles associated with the topic. Late in the day, each group presented its results to the full workshop and received feedback. During these discussions, at least three cross-cutting topics were identified: libraries, career paths, and campus information technology organizations & data repositories; see Section 5. The workshop attendees also discussed how to write up the workshop and future steps.\n\nResearchers respond to a variety of incentives in their daily activities. Many of these revolve around personal job security, hiring, promotion, and tenure. Credit and attribution form a core part of this, since the labor of evaluative job panels is often delegated to publication venues. This creates restrictions on both the types of practice that researchers will undertake and the forms of material that they are willing to publish. Without incentive structures that measure and value open scholarship, we are unlikely to see a large-scale transition.\n\nThe current set of assessment, credit, and attribution systems in the academy do not respect a diverse range of outputs, products, and activities. Instead, they coerce innovative work into known media forms in order to be congruent with assessors’ expectations. As a result, those working on digital and software outputs are often disadvantaged in the academic credit ecosystem. Those who collaborate on projects also fare badly by current standards, with poor recognition of non-authorial contributors. Those producing non-traditional research outputs would gain by changes to university assessment and credit/attribution procedures, since much contemporary scholarship and research now rests upon software and data outputs, which must be properly attributed. Proper credit for both traditional and non-traditional works also depends on avoidance of plagiarism.\n\nHowever, it is also apparent that different types of “credit” exist, and that this is not a homogeneous term. Credit and attribution may work differently for those who do not seek a traditional tenure-track road in the academy. University leadership is often wary of intervening in decisions about assessment as they do not wish to be seen to encroach upon academic freedom. We also found that there was an increasing sense of a need to hire new types of faculty/staff and to actively develop criteria for their assessment in order to maintain a global open research infrastructure.\n\nStakeholders in the credit and attribution space are many and range from: early-career, mid-career, late-stage researchers, faculty who sit on committees, university administrators, funders, publishers, and metric providers. Late-stage and tenured faculty have more chance to experiment in this domain since the consequences at their appraisals are far less serious than for those without secure employment.\n\nEconomic imbalances between different stakeholders are also present in this space. Metric providers and publishers, for instance, derive economic benefit from becoming evaluative frames. Apart from being poor scholarly practice, such evaluation of containers (presses, journals) restricts the type of researcher outputs that are incentivized. We need to move away from the journal impact factor or container name as a proxy for research evaluation as also suggested by previous initiatives, for example the San Francisco Declaration on Research Assessment (DORAb).\n\nPrinciples. We defined the following principles for credit and attribution:\n\nA1 Stakeholders (funders, universities, and researchers) should incentivize credit and attribution for a diverse range of research products and activities, such as research software, dissemination, infrastructure, data products and repositories.\n\nA2 Research should be assessed on its own merits, not based on its appearance in exclusive publication venues. This could be through the use of article-level metrics or narrativized impact measures. Journal impact factors should not be used to evaluate researchers.\n\nA3 Institutions that comparatively measure attention scores should evaluate attention measures with care, since such altmetrics may not correlate with measures of quality.\n\nA4 Institutions should provide appropriate career paths for staff working on open research and maintain the institutional infrastructure required for open research, including recognizing and valuing new, emergent forms of digital outcomes, such as software and data creation, curation and preservation, that are crucial to open research endeavors. These pathways may require rethinking existing classifications and assessments of tenure-track, non-tenure-track, and staff categories of university participants, and funder support for personnel in these categories.\n\nCommunities are the fabric of open research, and serve as the basis for development and sharing of best practices, building effective open source tools, and engaging with researchers newly interested in practicing open research. Effective communities often emerge from bottom-up interactions, and can serve as a support network for individual open researchers. These communities can consist of virtual clusters of like-minded individuals; they can include scholars, librarians, developers and technical staff or open research advocates at all levels of experience and with different backgrounds; the communities themselves can be short-lived and focused on a specific issue, tool, or approach, or they can have more long-term goals and aspirations. A key defining feature of these groups is that the principles of open scholarship permeate their practice, meaning they aim to be inherently inclusive, and aim to open up the process of scholarly exploration to the widest possible audience.\n\nExamples of success. In the meeting, we discussedc different examples of (successful) Open Scholarship Communities, and ways in which these have been developed. To begin with, although successful Open Scholarship Communities (OSCs) collaborate on infrastructure, they can still compete on science: this extensive collaboration does not mean that the science is de-scoped, or there is less competition between researchers. For a successful collaboration, the perceived value of participating has to be greater than fear of consequences. In other words, participation to Open Scholarship Communities should increase value (“I don’t have to reinvent this”) or decrease fear (“I can use a standard someone else invented!”). This is not an all-or-nothing step: it is important to praise incremental steps and make it easy (if not automatic) to continue ‘open’ behaviors.\n\nFor any of this to happen, and to make sure that the barriers to participation are not too great, systems and tools should emphasize the lowering of barriers to entry, resource efficiency and productivity: much can be gained in the tool/middleware layer. This also means that those creating those systems and tools get credit for it. We propose a new metric where work on infrastructure development is valued. One idea would be to rate researchers on a ‘FISH’ scale: Funding, Infrastructure, Science, and H-Index: four dimensions that validate orthogonal contributions to scientific progress. It is necessary that the credit and attribution system is in place to provide examples of ‘infrastructure leaders’ (akin to research leaders) to help overcome the notion that researchers who work on infrastructure and tool development suffer with respect to funding, promotion/tenure and such. In fact, we know of several examples (including some of the workshop participants) of researchers whose careers benefited strongly from being involved with open infrastructure, tooling, and community activities.\n\nRecommendations. After collecting a series of narratives on effective and intentional approaches to creating, growing, and nurturing such communities, we recommended the following actions for the different stakeholders to support the formation of adaptive and organic, bottom-up, distributed and open research communities:\n\nFor institutions, it is important, first of all, to provide the physical space and/or administrative support for community interactions. Since these practices are often not ingrained in the current research culture, institutions can support open scholarship by recognizing the need for explicit training in principles and practices of open research. This can and should include exploring what ‘design by a community’ looks like in areas where it’s not traditional, (e.g., mechanical engineering) and actively support changing views of what constitutes excellence in a discipline. Becoming more open is not a single step, but a process: it is imperative to reward incremental steps and provide incentives for engaging in different aspects of open scholarship at different levels. Engagement is more likely to occur if all steps are recognized as progress (some scholars will be happy to share their code, but not their data, or vice versa) and it’s easy to continue down a ‘sharing trajectory’ with incrementally greater levels and forms of openness.\n\nFor funders, it is important to recognize how ‘disciplinary shackles’ can hinder adoption of open scholarship practice. Development of common software, workflows and other community resources may not be respected as part of disciplinary work, but funders recognizing these non-traditional outputs can effect a culture change. A key component of openness is a focus on collaboration over competition: funders can contribute to making this happen by awarding grants to interdisciplinary and team efforts next to or instead of individual competitive efforts. Inclusivity is a defining feature of open scholarship, as well as extensibility and reproducibility. The goal is not solely to further individual rewards but to facilitate involvement of others: this means looking beyond ‘lock-in economics’ where the winner takes all, and exploring other reward systems. As with research institutions, funders should not adopt an exclusive definition of open scholarship, but reward incremental steps, by providing incentives for different aspects of open scholarship.\n\nPublishers and (research) platforms (data repositories, standards bodies and such) can support the trend towards openness in various ways. First, they can build the process of openness into the platform interface, by making openness the easy option, enabling open scholarship training materials into the platform, and building social networks and sharing opportunities into the fabric of the user interaction. Platforms can lower the entry barrier towards sharing practices by helping to build and define communities (e.g., similar to “My Facebook friends” you can have “My Jupyter Friends”). When platforms support the creation of communities around specific tools and practices, this helps build norms and codes of conduct into these platforms endemically. Community development can be further enhanced by supporting the development of platform specialists inside institutions (e.g., “JupyterHub guru on campus”) and supporting “pop-up open scholarship communities” around specific tools and practices (e.g., “open data hackathons”).\n\nAs a fourth and last stakeholder, community organizers can build openness into governance by recognizing the value of simple narratives for attracting people into community participation. This means identifying and funding ‘culture changers’: people who are tasked with changing, e.g., data dissemination processes/practices, and people who bring a culture and practice of open scholarship into the community and are happy to share their knowledge, toolset, and experience with community members. The community can and should reward incremental steps towards openness by community members, to easily allow new members to join. To ensure that diversity in background, culture and experience is acknowledged and maintained, communities should establish and maintain a code of conduct and set of expectations regarding community interactions.\n\nPrinciples. We defined the following principles for open scholarship communities:\n\nC1 Scholarly communities are both the target and the product of open scholarship and open research. Therefore, the fostering of communities is a key driver for open research.\n\nC2 These communities can take many forms. They may, for example, coalesce around tools, practices, shared interests, shared data or software, or shared hashtags. They can be short- or long-lived, explicitly funded, or emerge organically from shared interests among the participants. All of these variations are valuable components in the open research social ecosystem, and they must be supported as such by multiple resources, including travel funds, virtual networks, compensations for networking events, etc.\n\nC3 Community activities often serve to evaluate, encourage, and improve the use of tools, software, platforms, and data that are critical to open scholarship. All stakeholders must take steps to encourage these communities to develop, such as supporting common standards (and rewarding those who work on them), and funding projects that form a “connective tissue” between different communities. They should also actively encourage sharing practices for tools, and people across communities.\n\nThe practice of scholarship has a natural tendency to result in insular communities uniquely driven to produce new knowledge within narrow disciplinary bounds. The communication of research and research findings through traditional modes of journalism, such as through newspapers and television, is often limited to only high-profile work with broad public interest. At the same time, the communication of the rest of research is a critical component of having an informed public14. Further, since much of the funding for research and scholarship is derived from public money – as state or federal grants, tax funds, or student financial aid – the research community must find new and ever evolving ways to communicate with a diverse set of stakeholders, each of whom plays a role in ensuring that the societal and technological progress of our world is supported. This aligns well with the intent of the land-grant university under the Morrill Acts, which help define the missions of many higher education institutions and include a focus on outreach for an educated populace. On this basis, we proposed a set of guiding principles (see below) for outreach and engagement with the public. First of all, outreach and engagement, for public access to and understanding of research outcomes, depend on the audience and may encompass products, process, and dissemination. It is important to note that universities and researchers share a mutual interest in interactions with diverse audiences: therefore, outreach must be an institutionally valued, recognized, and supported component of university “service.” A key aspect of all of this is that access to products (research) and practitioners of research (scholars) builds trust in the products and process of research (scholarship).\n\nOperating within a changing landscape of scientific reporting, researchers often find themselves in a position of needing to fill communication gaps through outreach and engagement with stakeholders. These stakeholders are diverse and may include members of the public, the media, policymakers, educators, science enthusiasts, industry, students, other researchers, and university administration. Each of these diverse groups of people represent different and sometimes competing interests. Accordingly, the message disseminated must be crafted to match each audience and their needs. This could range from the raw data generated during an experiment to a broad explanation of the scientific process aimed at improving general scientific literacy. In most if not all cases, it is not appropriate to assume that by making our work available through open access journals we are doing enough to make our work accessible to the public. For our work to be accessible, it must be both available and comprehensible by the general public. Within this line of thinking, some labs and departments have implemented public-engagement policies aimed at improving public understanding of their research or field (e.g., Harvard University Department of Astronomy Public Outreach Projectd).\n\nThe ability to communicate on such a variety of topics to such a diverse set of stakeholders is generally not a part of a researcher’s training. This means that institutions must invest in outreach and engagement techniques and support their researcher’s efforts in building this skill set. Success in these efforts can potentially be evaluated through assessment of the reach of such communications. Alternative metrics such as social-media influence or publications in traditional media will help university communication offices track whether efforts are having the desired effect.\n\nPrinciples. We defined the following principles for outreach and engagement with the public:\n\nO1 Outreach and engagement, for public access to and understanding of research outcomes, depend on the audience and may encompass products, processes, and dissemination.\n\nO2 Universities and researchers share a mutual interest in interactions with diverse audiences.\n\nO3 Institutions should value, recognize, and support researchers who participate in outreach and engagement activities, crediting them as components of \"service\" to the institution.\n\nO4 Access to researchers (scholars) builds trust in the products and processes of research (scholarship).\n\nOpen scholarship can have an impact on improving education. In order to prepare students for emerging careers, accelerate discovery and reduce redundancy, incorporating open scholarship can bring about more opportunities and is critical for the survival of universities. The idea of Open Educational Resources goes back about 30 years, when advocates of “open content” proposed that principles of Full Option Science System (FOSS) could be applied to educational materials. (The term Open Educational Resources, OER, was coined at the 2002 UNESCO Forum.) Recurring topics in OER are reducing cost (for students), and increasing access.\n\nTo implement these ideas, we discussed examples that were relevant to these suggestions and were successfully adopted. As most of the workshop participants are University faculty, many of our recommendations were geared toward undergraduate level education. However, the sooner we can implement the idea of open scholarship, the greater the impact for future generations, and as university faculty, we should work with K-12 teachers to prepare and train them for implementation of open scholarship concepts. To do this at the university level, we first propose to embed open scholarship practices in the current curriculum for each major, as has successfully been demonstrated at UC Berkeley, aspects of which have been summarized in 15. Second, we propose to identify key faculty leaders in open science and incentivize engagement of open science with innovative methods such as funding course buy-outs. Third, we propose to develop low-barrier training and communities for students, staff, and faculty to engage in open scholarship and its benefit. A carpentry websitee that teaches foundational coding and data science worldwide perfectly embodies this concept. Last, we propose to introduce the concepts to K-12 teachers to prepare and introduce the concepts as early as possible.\n\nMany opportunities are available to drive universities towards open scholarship. Some, such as re-evaluating the university’s educational role, are more challenging, while others may pose lower barriers, such as positioning land-grant universities to provide a more contemporary role for education by merging research and education. A critical component for many students today is the gap between education and research. Most undergraduate students never have the opportunity to be exposed to research, which is critical for creativity and accelerating discovery, and can increase the relevancy of education.\n\nRevamping curricula comes with obvious challenges, such as changes that might involve university structure, training instructions, and changes in materials. However, due to the fast pace and constant change in research, textbooks in classrooms are becoming less useful, particularly in upper-level undergraduate courses. We have a window of opportunity, due to outdated textbooks, to introduce concepts of open scholarship. Open scholarship can help accelerate discovery by not depending on old literature, reducing redundancy, performing higher quality and more efficient research, and lowering barriers to collaboration by building resources for broader communities. Metrics are also needed that can be used to define the success of open scholarship for education to be incorporated in classrooms; in order to drive change, there needs to be a way to define the success of open scholarship in the current examples.\n\nOne example of open scholarship at the University of Arizona involves a researcher in drug discovery and basic sciences, May Khanna, who has created a course named “From Chemistry to Cure,” incorporating concepts of open scholarship. The course will begin with virtual docking of targets chosen by the students using cloud computing as has previously been donef. The students will continue the process of virtual drug discovery through the course, uploading their results on a live blog. The students will then complete the course by pitching their concepts to business students and results from pitches will continuously be shared. The course will include live student quizzes with such software as Poll Everywhere and apps that allow for cloud sharing of information. The idea of open scholarship in the course allows it to be integrated with other universities throughout the world, which will be done as the course matures in future years. This open scholarship course touches on several critical points that were discussed in the workshop: it merges research with education; it utilizes cloud computing for increased computing power through platforms like Google or Amazon and thus is not limited by local computers; it merges research goals with education using an open forum to teach students to be creative, open, and to accelerate discovery; and it shares materials with an international open forum, which breaks down the barrier between research and education. The course is completely driven by the students, which gives the students greater responsibility.\n\nPrinciples. We defined the following principles for education:\n\nE1 Every student should be guided to understand and learn how to access and use data and software to be well prepared for diverse, modern career paths.\n\nE2 Through open research training, universities should play an active role in increasing research by enabling evidence-based decisions, accelerating discovery, and extending impact to broader communities.\n\nE3 Universities should encourage their faculty to engage in open educational practices, including creating and assigning open educational resources, and reflecting open culture in their courses.\n\nReproducibility depends on transparently documenting and sharing all data products, protocols, and computational algorithms (with source code) used in the research. While advocates of open, reproducible scholarly research believe that it should be the norm—coupled with data sharing, reusability, and sustainability more generally—individual and institutional barriers hold back wide adoption of such practices and workflows. Currently, while some individuals and research groups feel strongly about openly sharing all research products (including data, written output, and software) and working in a reproducible way, they are largely motivated by personal beliefs. (Some scholarly communities have developed cultures with some aspects of openness, e.g., in physics the sharing of preprints via arXiv is the norm; but this is not widespread.) Thus, incentives at both institutional and wider community levels are needed to initiate change.\n\nCommunity leaders can and should make positive arguments for sharing and reuse of digital artifacts of research. These arguments could be more successful than negative ones around lack of reproducibility (the “crisis narrative”). Although arguments for open data and software often focus on increased citations, we can also argue for the greater overall impact and opening up of new, collaborative avenues of research. Similarly, while funders and publishers may support reproducible research in theory, in practice they currently provide few incentives. Funders could precipitate change by making reproducibility concerns part of the merit-review criteria; they could require compliance with data management plans for any future support, and extend such plans to consider software explicitly (e.g., “Data and Software Management Plans”). Publishers can award badges to articles that present open and reproducible workflows, for example, as is the case in the ACM Transactions on Mathematical Software16, or could go even further by encouraging editorial boards and reviewers to only consider submissions with such workflows. For example: the American Journal of Political Science contracts a third-party to verify that author-provided files are sufficient to reproduce the results in the paper. The Odum Instituteg, in this case, carries out reproducibility checks of accepted papers, and authors submit any required additional information before publication.\n\nReproducible workflows require ensuring access to open and reusable data as well as open, reusable, and sustainable software. Data should be preserved following established community standards, such as the FAIR principles – data should be Findable, Accessible, Interoperable, and Reusable17. An analogue to the FAIR principles for data does not exist for software, although recommendations have been made for the specific case of applying fair-use principles to allow preserving software for posterity18. However, for the purposes of reproducibility, research software needs to be made open source at the time of publication of the research results. Perhaps more importantly, it should be written from the outset with sharing and reuse in mind (ideally under an open development model). A new initiative, The Journal of Open Source Software, provides peer review on open code and promotes good practices for preservation (an OSI-approved license is enforced and software must be persistently archived before the paper is published). Leaders at all levels should also encourage scholars to appropriately cite data and software when used for a study, similar to citing literature articles19,20, to help standardize this behavior.\n\nIn addition to communicating the importance of preservation and reproducibility, training in skills for open and reproducible workflows needs to be emphasized, noting that skills for reproducibility are not the same as those for working openly in general. This could be another new role for libraries, but some work may be discipline-specific, and training by faculty should also be recognized as a service contribution. One good example of this is C. Titus Brown’s position at University of California Davis, which involves lower “traditional” teaching loads but more service in the form of computational and data science training for biologists and bioinformaticians.\n\nPrinciples. We defined the following principles for preservation and reproducibility:\n\nP1 The scholarly publication and communication ecosystem should support open and reproducible research, and enable credit for these efforts. Universities should encourage these initiatives by creating incentives (e.g., promotion and tenure categories, service recognition) for such activities.\n\nP2 Research funders should support open and reproducible research by making reproducibility part of their merit review criteria. They should also create new scholarly communication venues or support open scholarship efforts, and encourage, require, and reward reproducible research efforts.\n\nP3 Incentives that promote the public sharing and distribution of scholarly knowledge for open/transparent/reproducible research practices must be put in place. Publishers must require, when appropriate, submissions that provide open and reproducible workflows, and embed this requirement in their own workflows.\n\nP4 Universities should recognize the activities of faculty to educate and train researchers on open and reproducible research skills. Global and national bodies (e.g., National Academies) should promote this recognition across universities.\n\nEnabling open, collaborative scholarship that engages students, researchers and the public requires reducing the barriers to entry for both generating and distributing knowledge “products.” Academic endeavors—in both research and education—reap most benefit from adopting open source software in all technologies they adopt. Utilizing closed-source software creates a number of impediments to effective research, including reducing verifiability of the research products, reducing opportunities for synergistic collaborations, and imposing barriers to entry for reproducibility and dissemination of knowledge. In addition to these pragmatic considerations that favor open source software, we also identify that open source technologies provide intrinsic value to the entire research process that extends beyond a monetary value proposition. We therefore not only recognize that open source technologies should be preferred to accelerate open scholarship, but that researchers should be appropriately acknowledged and rewarded for participating in the ecosystem of open source and open data for scholarship.\n\nIn support of embedding open source technologies for open scholarship, we propose that institutions prioritize the identification and (ad hoc) endorsement of capabilities that meet several criteria. Firstly, the tools should be both interoperable and, to the extent possible, self-documenting. This provides the ability for individuals to communicate between different pieces of software and technical infrastructure, ensuring they are able to transport their work as the situation requires. An example of this is in data formats and storage methods, as well as in the ability of in-memory transfers between software libraries, or in the execution of virtual machines and containers on different cloud providers.\n\nThe second characteristic of tools that we highlight is that they should be selected to reflect a wide-ranging set of evaluating criteria. Rather than determining “winners,” these tools should be drawn from a diverse, evolving set of possibilities. Entrenching a single technological choice may serve to unduly influence future research studies and restrict growth of scholarly technologies.\n\nWe also recognize that frequently the ability of researchers to utilize cutting-edge open source and open scholarship infrastructure can be subject to the bottleneck of their own technical skills. This imposes a technical barrier to entry that we believe will detract from the utilization of open scholarship tools and software. To mitigate this, we propose that the University of Tomorrow provide foundational infrastructure, in the form of both technical resources (deployments, hardware, “glue” software) and human resources (support staff, contributing members of the open source community) to ensure that these tools and opportunities are made available widely across the university, coupled with learning opportunities that allow sharing of best practices.\n\nPrinciples. We defined the following principles for technologies:\n\nT1 Open source technologies, tools and platforms provide intrinsic value to researchers and educators and are an effective way of accelerating open scholarship. Academic institutions should favor and encourage open source solutions as much as possible.\n\nT2 A diverse and interoperable set of tools for open research should be known, shared, and clearly documented.\n\nT3 Institutions should provide and support foundational open scholarship infrastructure (technological and human) for all members of campus.\n\nT4 Institutions should recognize the contributions made by all members of campus to open scholarship infrastructure.\n\n\n5 Cross-cutting topics\n\nThree cross-cutting topics were identified in multiple group discussions: libraries, career paths, and campus information technology organizations & data repositories, though there are likely other parts of the university that should be involved in future discussions, such as policy and research offices.\n\nEvery research university has a library whose mission is to support research and education by collecting, organizing, managing, preserving, and ensuring long-term access to the products of research and the scholarly record for every discipline. Librarians and other library staff provide expert support to students and researchers at every career stage, universally and democratically. In many ways, libraries are the original core research facility, inseparable from the university itself, evolving alongside technological advances and other changes in research and educational models and methods. As we consider the future of research universities, open scholarship, and data-driven research, the unique role that the library plays in the university needs to be re-envisioned and perhaps broadened.\n\nMany of the challenges and opportunities facing research universities in embracing open scholarship and data-driven research revolve around “scholarly content” – recorded knowledge in books and journals but also in software and models, datasets and databases, visualizations and vast digital libraries. Libraries are already adapting to include these new forms of scholarship in their traditional functions of collecting, organizing, describing, preserving, and providing ongoing access. But digital content is different than print and other analog formats, providing greater challenges and opportunities at the intersection of research and content, on the production and consumption sides.\n\nSpecific ways in which libraries support open scholarship and data-driven research include the following.\n\nLibraries play a key role in helping institutions and individuals document their impact on knowledge creation. New forms of scholarship and research (e.g., scientific software or complex data creation) require new forms of credit and attribution to measure their impact in ways that can be aligned and integrated with traditional credit mechanisms: citation and attribution. The current scholarly record system was designed for authorship and is maintained by libraries and information companies (e.g., Web of Science), and libraries are collaborating with researchers to develop new citation standards and methods, promote new disciplinary norms, and create new tools and databases that interweave new and traditional scholarship. Libraries train students in citation practices and bibliographic tools, and collect the data that provide evidence of impact, including alternative metrics to traditional citation, such as ‘altmetrics,’ documenting both institutional and individual impact21.\n\nLibraries are natural homes for interdisciplinary research communities to form and collaborate. They provide central, neutral, and welcoming facilities, often with shared equipment and other research tools that are expensive and inefficient to duplicate in multiple departments (e.g., 3D printers, visualization tools, specialized software). Importantly, library spaces bring together researchers and students from across the university who are then exposed to each other’s research methods22. This is analogous to “browsing the stacks” of 20th Century libraries but with people and research as the objects of serendipitous discovery.\n\nThe fact that libraries are naturally omnidisciplinary allows them to organize quickly and fluidly around new research constellations without the need to form new “disciplines,” with consequent norms for publication, curricula, and excellence. They can attract and often hire academic staff to run new research areas (e.g., data science or spatial science researchers) and provide them with avenues for recognition and advancement, if not tenure.\n\nLibraries are logical stewards of and repositories for open scholarship technologies and data23. They provide expert support to all researchers and students for particular tools (e.g., bibliographic management tools, GIS software systems, 3D printers, text mining or bioinformatics) and practices (e.g., data curation, software publishing and archiving). They have the technical skills to catalog and organize the new products of open scholarship (e.g., data and software), individually and collectively via national consortia and programs24. For open scholarship to become the default, it will take coordinated international efforts to manage the tools and products of research, and libraries are well-positioned to expand current networks to meet that need.\n\nA core value of libraries is knowledge sharing – that everyone should have free and frictionless access to all knowledge for all time, whoever and wherever they are. Because of that value, libraries are strong advocates of open scholarship – open access, open data, open educational resources, open methodologies and research reproducibility – and promote them to their own communities and to the public at large. As an illustration of this advocate role, libraries are providing leadership in the development and support of open scholarship policies. Library professionals contribute expertise on open scholarship issues to government and university leaders as policies are being written. As new policies emerge and evolve, library staff are then often central to campus initiatives focused on meeting new policy requirements.\n\nLibraries can provide skills training at the point of need and in flexible ways, unlike traditional academic departments25. These training offerings can be credited (e.g., a semester-long seminar) or uncredited (e.g., a two-week boot camp or a single class) to allow for different incentives and rewards to students. The central locality of these training sessions exposes them to other students and researchers, providing a novel form of marketing and outreach.\n\nThe library’s mission to acquire, organize, and ensure permanent access to recorded knowledge means that it has already begun serious exploration of methods for digital preservation and research reproducibility. Since the 1990s, libraries have been developing best practices for describing and preserving many types of digital research products, e.g., CAD models, software libraries, images of many formats, audio and video. Libraries are key partners to research-generating agencies for ensuring long-term access to digital content, and to operationalizing support for that content over decades and longer26.\n\nThe main challenge that libraries face in building support for open scholarship and data-driven research is the same one facing their parent institutions: the need to balance supporting traditional modes and methods of conducting research and education, still critical for many functions and people across these institutions, with investment in new types of support for the new institutional goals27,28. Transitions can take decades and straddling two worlds is complex and expensive. To find the right balance of investment and pace of change, clear guidance is needed from both university administration and from the researchers and students themselves. Libraries already regularly build collaborations across campus units (e.g., with campus central IT/computing units and research administration offices) to ensure that research support services are coupled and coordinated. Such collaborations will continue to grow in importance.\n\nThe transition toward open scholarship is occurring unevenly across institutions and disciplines, so we find incoherent systems of practice today. Beginning with disciplines that are further along in the transition, defining global strategies for success, building collaborations, and focusing investment to implement those strategies, are good places to begin.\n\nNew models for supporting career paths that involve research, development, and campus service are emerging. Today, these positions can be considered along multiple axes, including professional status (postdoctoral researcher, staff, traditional tenure-track faculty, teaching faculty, research faculty, and faculty of practice), length of position (short term, long term, permanent, tenured), funding (soft/grant funded, institutionally funded), and organization (in one or more academic department(s), in an IT organization, in a center or institute, the library, or a combination). This is a large and diverse space of possibilities, and it is both possible and likely that universities may create new career paths with little overlap in their models. Hence, there is value in experimentation and in the insights gained from these experiments.\n\nA number of common elements appear, including:\n\nThe criteria used in hiring, evaluation, and career advancement.\n\nThe mix of job responsibilities, including teaching, research, service, and community engagement.\n\nThe positioning of software development and maintenance relative to knowledge creation, preservation, and dissemination. For example, it might be part of responsibilities in a traditional classification, or part of a new one.\n\nThe remuneration models and their relationship to those of other career paths at the same institution, and professionals with similar skills making a career in industry.\n\nTransition models into and from this career path, both within and across institutions.\n\nIn one example of such a model, from the University of Washington (UW), the Provost supported half faculty lines to help recruit data-oriented and software-oriented faculty across campus, and to help give them a community. In return, these faculty engage in teaching courses relevant campus-wide. Their tenure case is still owned by the home department. UW also explored chaired professorships in various fields funded through a Washington Research Foundation grant. These titles help reinforce community engagement and leadership among faculty around campus.\n\nPerhaps more uniquely, the UW eScience Institute has developed strong career paths for data scientists, typically with PhDs in STEM fields and strong software development expertise. These are very different roles than typical Research Software Engineers, which we find tend to be interpreted as service roles rather than senior researcher roles. The model for these data scientists is:\n\nPhD in domain science, with significant experience in data and/or software.\n\n50% work on Institute projects and initiatives, and 50% on their own research. (Typically 100% involves collaboration with others around campus.)\n\nAutonomy to choose which initiatives they work on.\n\nPI status for non-junior roles.\n\nAffiliate faculty roles, where that makes sense.\n\nIf they buy out their time by being part of a grant, they get some of the salary savings returned as research/travel/student budget.\n\nThey are seen as \"faculty peers.\" For example, sometimes working titles have significant influence: \"Director of Research in the Physical Sciences\" rather than just \"Data Scientist.\"\n\nThere is a community of these Data Scientists – shared governance, shared space, shared initiatives, social events. This avoids them disappearing into other peoples’ labs, where they would risk losing their autonomy and respected status.\n\nThis model appears to be replicable, perhaps by exploiting attrition in IT departments and libraries to build up a community of Data Scientists.\n\nUniversities collect massive amounts of data for research, teaching, learning, service, outreach, and strategic management. These data collections expose universities to new risks and create responsibilities that may converge and diverge in unexpected ways. Drawing on recent work29, this short section examines university concerns for research and “grey” data gathered for administrative and operational purposes.\n\nBy collecting data, institutions assume responsibility for managing those data in the short and long term. “Stewardship” is an overarching term that encompasses sustainability, curation, access, and preservation. Although “stewardship” is used in nuanced ways in the scientific, library, archival, and policy communities, it reflects a commitment to managing data in ways that they remain findable, accessible, and useful. For some kinds of data, stewardship requires indefinite preservation; for others, regular cycles of record disposal are needed. Modern data collections are dynamic, thus traditional archival approaches to sustaining access to static resources are unlikely to suffice. In an “age of algorithms” where datasets are in constant flux and can be disaggregated and re-aggregated continuously for multiple analytical purposes, new approaches are sorely needed30.\n\nUniversities have broad responsibilities for stewarding the data they collect, acquire, and hold. Despite the diffuse responsibility borne by institutions, some individual persons, offices, committees, or other entities must take specific actions, make investments, and manage the daily operations of data stewardship. Determining which entities have which responsibilities, based on what criteria and policies, is the process of governance. The University of California was among the first to address these processes in U.S. higher education, explicitly acknowledging the “distributed nature of information stewardship at UC, where responsibility for privacy and information security resides at every level”31. Universities are taking many approaches to governance, ranging from appointing “data czars” to assigning offices or committees to formulate generalized policies, agreements, and governance mechanisms.\n\nWhereas universities are generally held responsible by funding agencies for maintaining data, the responsibility for disposition and stewardship usually falls to the researchers who collected those data. They have vested interests in exploiting and protecting these data and they know the most about the data’s content and context. Local knowledge is essential to data management, given the vast array of data types, domain expertise, policies, and practices. However, the benefits of local control must be balanced with expertise and continuity. In units with external funding, graduate students and post-doctoral fellows conduct most of the data collection and perform most of the management tasks. They also write software code, scripts, and algorithms to analyze those data. Rarely are these domain experts also experts in data management or software engineering. Essential research tasks are being performed by short-term employees who are replaced every few years as students graduate, fellowships end, and grant projects are completed32.\n\nIn many academic domains, authors and investigators are responsible for releasing datasets associated with publications. Finding and funding access to their data for some specified number of years after the granting period is a looming challenge. Where data archives exist, deposit is usually the preferred solution, whether organized by discipline, data type, or institution, as these entities tend to have long-term commitments and staff responsible for curation. Archiving of digital research data has been under way for at least fifty years by entities such as the World Data Systems33, Harvard’s Institute for Quantitative Social Sciences and Dataverse34, and ICPSR35. Funding agencies vary considerably in support for sustaining access to findings. Some provide data archives, others require universities to maintain their own data archives as a condition of receiving grants, and yet others are agnostic on the disposition of datasets, as long as they remain accessible. Sustaining access to public archives is itself a challenge, as many of these are funded by research grants that expire on a cyclical basis.\n\n\nConclusions\n\nThe workshop “Imagining Tomorrow’s University” was a unique gathering of early career faculty and university leaders, united in their efforts to understand the implications of open scholarship on future universities. They worked together to derive the 22 principles listed above based on their personal goals as well as their mutual interest in seeing that universities take best advantage of opportunities brought about by current changes in scholarship and society, including increased digital products, increased sharing and transparency, public skepticism in authority, and the perceived reproducibility crisis.\n\nOther groups are also working in this area. Since the workshop, the National Academies has written a report on how to accelerate the movement towards open scholarship36 and the AAU-APLU Public Access Working Group released a report describing principles and recommendations for agencies and institutions to consider in implementing infrastructure for sharing research data37.\n\nActivities that follow on from workshop results take one of three forms. First, since the workshop, a number of workshop authors have further developed and published their white papers38–41 to make their reflections and recommendations more concrete. These authors are also conducting efforts to implement these ideas, and to make changes in the university system. For example, one of the current authors (Erin McKiernan) recently collaborated on a project to analyze review, promotion, and tenure (RPT) documents from a representative sample of 120 universities in the U.S. and Canada to learn how the public dimensions of faculty work, including aspects of open research, are currently valued and rewarded in university evaluations42.\n\nSecond, we propose to organize a follow-up workshop that focuses on how these principles could be implemented. This workshop could include 4–7 research institutions, some of which were represented at the workshop described in this report, and some of which would be new to the table. It would also include participants from a greater diversity of positions at these universities, including early career researchers, senior faculty, department chairs, deans, librarians, research data managers, as well as higher-level university leaders. This workshop would aim to work out the details by which a set of changes could be made, and conduct a trial to implement at least some of those changes at each of the participating institutions. The workshop would identify clear objectives and key results for each of these trials, as well as metrics by which they could be assessed.\n\nThird, we believe that the outcomes of this workshop support and are connected with recent theoretical work on the position and future of open knowledge institutions. In particular, the recently published online monograph “Open Knowledge Institutions: Reinventing Universities”43 advocates that universities “become Open Knowledge Institutions which institutionalise our world’s creative diversity in order to contribute to the stock of common knowledge.” The authors argue that such Open Knowledge Institutions “act as networks of knowledge, spanning common disciplinary boundaries and campus barriers in order to serve as agents for societal change.” The work proposes a theoretical change mechanism for knowledge institutions to become more open, which we believe can be a useful frame of thought, as well as a means of studying change in knowledge institutions towards greater openness. It is argued that the change from closed to an open (knowledge) institution happens in three stages:\n\n(i) policy and intent signaling (where action is desired and expressed);\n\n(ii) action and investment (where action is being taken); and\n\n(iii) measurable outcomes (where the result of action is assessed).\n\nTo make sure that actions undertaken do indeed lead to greater openness, three aspects should be taken into account:\n\nSince one of the core goals of open institutions is inclusion, Diversity is essential to this change, to ensure participation from a broad group of stakeholders, including nontraditional/unfunded/formerly peripheral actors (including the general public and local/nonlocal parties interested in university efforts);\n\nTo ensure a productive transition, extensive Coordination needs to be done between the (many) different groups involved, who all have different directives, backgrounds, requirements, and levels of knowledge and interest;\n\nTo make sure that Coordination occurs and Diversity is achieved, Communication is needed, both to transmit knowledge to the diverse communities and to support and engage diverse groups in the extensive, multi-stakeholder dialogues that characterize open institutions.\n\nConnecting these stages to these aspects of inclusion leads to a 3 x 3 table through which actions towards opening up knowledge institutions can be classified (e.g., see Table 1 in “Open Knowledge Institutions: Reinventing Universities”43). As a final follow-up to our workshop, we propose to use this framework to position and track efforts proposed and undertaken at the research institutions that take this work forward. We have commenced conversations with the authors of the monograph and will invite them to participate in these follow-up steps, to further enhance our collective understanding of this pivotal transition happening in academia today, from ivory towers to open knowledge institutions.\n\n\nData availability\n\nNo data is associated with this article.\n\nEarly Career Research Leaders\n\nDevin Berg (mechanical engineering), University of Wisconsin Stout\n\nHolly Bik (bioinformatics, genomics), University of California Riverside\n\nCarl Boettiger (theoretical ecology), University of California Berkeley\n\nC. Titus Brown (bioinformatics, genomics), University of California Davis\n\nMartin Eve (English literature), Birkbeck, University of London\n\nBrian Granger (physics, data science), Cal Poly State University San Luis Obispo\n\nAdina Howe (bioinformatics, genomics), Iowa State University\n\nBill Howe (computer science), University of Washington\n\nMay Khanna (structural biology, drug discovery), University of Arizona\n\nMatthew Mayernik (research data management and curation), National Center for Atmospheric Research\n\nErin McKiernan (biophysics, physiology, neuroscience), National Autonomous University of Mexico\n\nKyle Niemeyer (mechanical engineering; computational combustion and fluid dynamics), Oregon State University\n\nMatt Turk (astronomy, data science, information science), University of Illinois Urbana-Champaign\n\nUniversity Leaders\n\nRandy Burd, Associate Vice President for Research, Global Research Alliances, University of Arizona\n\nJean-Lou Chameau, President, King Abdullah University of Science and Technology\n\nJosé-Marie Griffiths, President, Dakota State University\n\nRandolph W. Hall, Vice President, Research, University of Southern California\n\nTimothy L. Killeen, President, University of Illinois\n\nSarah M. Nusser, Vice President for Research, Iowa State University\n\nCybele Raver, Senior Vice Provost for Academic Analytics and Graduate Academic Affairs, New York University\n\nDaniel A. Reed, Vice-President for Research and Economic Development, University of Iowa, now Senior Vice President for Academic Affairs at the University of Utah\n\nEdward Seidel, Interim Vice President for Research, University of Illinois\n\nJay Walsh, Vice President for Research, Northwestern University\n\nOther Stakeholders\n\nLaura Noren (organizational sociology), New York University\n\nMacKenzie Smith (research library), University of California Davis\n\nJeffrey Spies (research technology and methodology), Center for Open Science\n\nFunding Organization Representatives\n\nStuart Buck, Laura and John Arnold Foundation\n\nJosh Greenberg, Alfred P. Sloan Foundation\n\nChris Mentzel, Gordon and Betty Moore Foundation\n\nRajiv Ramnath, National Science Foundation (remote participation)\n\nCarol Shreffler, National Institutes of Health\n\nOrganizers\n\nGabrielle Allen, University of Illinois at Urbana-Champaign\n\nLorena Barba, George Washington University\n\nChristine L. Borgman, University of California Los Angeles\n\nAnita de Waard, Elsevier\n\nDaniel S. Katz, University of Illinois at Urbana-Champaign\n\nNirav Merchant, University of Arizona\n\nJohn Van Horn, University of Southern California\n\nEdward Seidel, University of Illinois\n\nFacilitators\n\nStavros Michailidis, KnowInnovation\n\nDonnalyn Roxey, KnowInnovation", "appendix": "Author contributions\n\n\n\nGA, DSK, JVH, and ES conceived the idea for this workshop. All authors attended the workshop and contributed to the discussion and the principles. DSK led the writing of this overall paper, with section writing led by MPE, AdW, DRB, MK, KEN, NM, MS, BH, and CLB.\n\n\nGrant information\n\nThis workshop was supported by awards from NSF (ACI-1645571, PI: DSK) and NIH (U24 ES026465-02; U24\n\nES026465-02S1, PI: JVH) and a gift from Elsevier. LN was supported by the Moore-Sloan Data Science Environment.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWorkshop attendees included all authors of this paper. In addition, Jean-Lou Chameau, José-Marie Griffiths, Randolph W. Hall, C. Cybele Raver, and Carol Shreffler attended and contributed to the workshop, and Pratim Sengupta contributed to the workshop via a white paper but did not attend. Facilitators were Stavros Michailidis and Donnalyn Roxey from KnowInnovation, whom we thank for their efforts in making the workshop run smoothly and effectively.\n\n\nFootnotes\n\nahttps://web.archive.org/web/20181020160025/https://research.usc.edu/rigor-transparency-and-reproducibility/.\n\nbhttps://sfdora.org\n\ncFor the full set of notes, see https://web.archive.org/web/20170709184934/http://ivory.idyll.org/blog/2017-open-science-communities.html.\n\ndhttps://web.archive.org/web/20180808184837/https://astronomy.fas.harvard.edu/public-outreach-project.\n\nehttps://carpentries.org/.\n\nfhttps://web.archive.org/web/20170912204801/https://cyclecomputing.com/improving-als-research-with-google-cloud-schrodinger-and-cycle-computing/.\n\nghttp://odum.unc.edu/.\n\n\nReferences\n\nBerg DR: Open research, open engineering, and the role of the university in society. engrXiv. 2017. Publisher Full Text\n\nBik H: Open science in the 21st century: Increasing scientists’ visibility and improving research efficiency. figshare. 2017. Publisher Full Text\n\nBoettiger C: Open science: Balancing individual incentives with common good. GitHub. 2017. Reference Source\n\nBrown CT: Blog: My thoughts for \"Imagining Tomorrow’s University. 2017. Reference Source\n\nEve MP: Imagining tomorrow’s university: Rethinking scholarship, education, and institutions for an open, networked era. figshare. 2017. Publisher Full Text\n\nHowe A, Howe MD, Kaleita AL, et al.: Imagining tomorrow’s university: open science and its impact. PeerJ Preprints. 2017; 5: e2781v1. Publisher Full Text\n\nHowe B, Grechkin M: Wide open science. zenodo. 2017. Publisher Full Text\n\nKhanna M: Answers to questions about open science for Imagining Tomorrow’s University. figshare. 2017. Publisher Full Text\n\nMayernik MS: Open data: Accountability and transparency. zenodo. 2017. Publisher Full Text\n\nMcKiernan EC: Imagining the ‘open’ university: Sharing scholarship to improve research and education. PeerJ Preprints. 2017; 5: e2711v3. Publisher Full Text\n\nNiemeyer KE: Open science and the future university researcher. engrXiv. 2017. Publisher Full Text\n\nSengupta P, Shanahan MC: Open science, public engagement and the university. arXiv. 2017. Reference Source\n\nRoyal Astronomical Society of Canada: Ten-year-old New Brunswick girl discovers exploding star. 2011. Reference Source\n\nPham D: Public engagement is key for the future of science research. NPJ Sci Learn. 2016; 1: 16010. Publisher Full Text\n\nNational Academies of Sciences, Engineering, and Medicine, Division of Behavioral and Social Sciences and Education, Board on Science Education, et al: Data Science for Undergraduates: Opportunities and Options. The National Academies Press, 2018. PubMed Abstract \| Publisher Full Text\n\nHeroux M: The TOMS initiative and policies for replicated computational results (RCR). Transactions on Mathematical Software (TOMS). 2015. Reference Source\n\nWilkinson MD, Dumontier M, Aalbersberg IJ, et al.: The FAIR Guiding Principles for scientific data management and stewardship. Sci Data. 2016; 3: 160018. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAssociation of Research Libraries: Code of best practices in fair use for software preservation. 2018. Reference Source\n\nStarr J, Castro E, Crosas M, et al.: Achieving human and machine accessibility of cited data in scholarly publications. PeerJ Comput Sci. 2015; 1: pii: e1. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSmith AM, Katz DS, Niemeyer KE, et al.: Software citation principles. PeerJ Comput Sci. 2016; 2: e86. Publisher Full Text\n\nRoemer RC, Borchardt R: Institutional altmetrics and academic libraries. Information Standards Quarterly. 2013; 25(2): 14–19. Reference Source\n\nMonastersky R: Publishing frontiers: The library reboot. Nature. 2013; 495(7442): 430–432. PubMed Abstract \| Publisher Full Text\n\nWitt M: Co-designing, co-developing, and co-implementing an institutional data repository service. Journal of Library Administration. 2012; 52(2): 172–188. Publisher Full Text\n\nErdmann C: Teaching librarians to be data scientists. Information Outlook. 2014; 18(3): 21–24. Publisher Full Text\n\nRaboin R, Reznik-Zellen RC, Salo D: Forging new service paths: Institutional approaches to providing research data management services. Journal of eScience Librarianship. 2012; 1(3): e1021. Publisher Full Text\n\nWalters T, Skinner K: New roles for new times: Digital curation for preservation. Technical report, Association of Research Libraries, 2011. Reference Source\n\nTenopir C, Hughes D, Allard S, et al.: Research data services in academic libraries: Data intensive roles for the future? Journal of eScience Librarianship. 2015; 4(2): e1085. Publisher Full Text\n\nJohnston LR, Carlson JR, Hswe P, et al.: Data curation network: How do we compare? A snapshot of six academic library institutions’ data repository and curation services. Journal of eScience Librarianship. 2017; 6(1): e1102. Publisher Full Text\n\nBorgman CL: Open data, grey data, and stewardship: Universities at the privacy frontier. Berkeley Technol Law J. 2018; 3(2): 365–412. Publisher Full Text\n\nLynch C: The rise of reading analytics and the emerging calculus of reader privacy in the digital world. First Monday. 2017; 22(4). Publisher Full Text\n\nUCOP Privacy and Information Security Initiative. 2013; [Accessed: 2018-08-27]. Reference Source\n\nBorgman CL: Big data, little data, no data: Scholarship in the networked world. MIT Press, 2015. Reference Source\n\nInternational Council of Scientific Unions: World Data System: Trusted Data Services for Global Science. [Accessed 2018-08-27]. Reference Source\n\nCrosas M: The Dataverse Network®: An Open-Source Application for Sharing, Discovering and Preserving Data. D-Lib Magazine. 2011; 17(1/2). Publisher Full Text\n\nICPSR - Inter-university Consortium for Political and Social Research. [Accessed: 2018-08-27.]. Reference Source\n\nNational Academies of Sciences, Engineering, and Medicine: Open Science by Design: Realizing a Vision for 21st Century Research. The National Academies Press, 2018. Publisher Full Text\n\nAAU-APLU Public Access Working Group Report and Recommendations. 2017. Reference Source\n\nMayernik MS: Open data: Accountability and transparency. Big Data & Society. 2017; 4(2): 2053951717718853. Publisher Full Text\n\nMcKiernan EC: Imagining the ‘open’ university: Sharing scholarship to improve research and education. PLoS Biol. 2017; 15(10): e1002614. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHowe A, Howe M, Kaleita AL, et al.: Imagining tomorrow’s university in an era of open science [version 2; referees: 3 approved]. F1000Res. 2017; 6: 405. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBerg DR, Niemeyer KE: The case for openness in engineering research [version 2; referees: 1 approved, 1 approved with reservations]. F1000Res. 2018; 7: 501. Publisher Full Text\n\nAlperin JP, Fischman GE, McKiernan EC, et al.: How significant are the public dimensions of faculty work in review, promotion, and tenure documents? Humanities Commons [preprint]. 2018. Publisher Full Text\n\nMontgomery L, Hartley J, Neylon C, et al.: Open Knowledge Institutions: Reinventing Universities. MIT Press, 2018. Publisher Full Text" }	[ { "id": "41794", "date": "14 Dec 2018", "name": "Lucy Montgomery", "expertise": [ "Reviewer Expertise Open Access", "Scholarly communication", "Creative Industries", "China" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper does what it says on the tin: it is a report on the perspectives and ideas generated and captured at an invitation-only workshop. As such, it presents the discussions and themes explored within the workshop, and points toward individual white papers published elsewhere. The use of references is appropriate to the form of the report. The inclusion of a survey element, which makes the perspectives of university leaders visible within the narrative, is helpful, and adds depth to the discussion.\nI would have found the document more comfortable to read if it has acknowledged the North American and Anglophone perspectives that it captures; and the limitations that this may imply.\n\nNonetheless, the report captures the energy, and the passion of the workshop participants; and points the way to important work that still needs to be done to ensure that the open knowledge potential of universities is realised.\nWhile I do not agree with every point made within the document, as an opinion article and a record of a series of discussions, I don't believe that it is necessary for me to do so. By making this document public via an open platform the authors are allowing for annotation and the continuation of conversations and collaborative knowledge building. I look forward to annotating the document itself and watching as differences in perspective continue to be negotiated by both the authors and a wider community as conversations about open affordances and the future of universities unfold.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Partly\n\nAre arguments sufficiently supported by evidence from the published literature? Partly\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes", "responses": [] }, { "id": "42058", "date": "14 Jan 2019", "name": "Olivia Guest", "expertise": [ "Reviewer Expertise computational cognitive neuroscience", "cognitive computational modeling" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper summarises the ongoing dialogue around open scholarship between ECRs (and others) and people in relatively more powerful positions within institutions based on the outcomes of a workshop, which included representatives from these two groups, held in March 2017 and in the USA. The points raised in the workshop and thus this paper are of value to the wider open scholarship community, to libraries, to institutions and funders and to society in general.\nIn my opinion, this article needs some minor polishing only. My comments below attempt to help the authors achieve a more clear version of their paper.\nThe Abstract and Summary could contain a sentence (near the top) that explicitly states what the article is about. I was a little lost without this. Something like \"This article is about X, Y, and Z.\" — not unlike the first sentence of my review. I wasn't sure if the workshop mentioned what was the main focus of the article or if the authors were themselves present until I checked what was being said further along in the manuscript. I wondered if the current article was going to elaborate on their viewpoints or add to them, etc. Along the same vain, in the paragraph that begins “In March 2017”, in the Introduction section, I think it would help the reader to know which of the attendees are authors? I know they are listed in the appendix, but I suppose a reader would have to manually check if all the names appear as authors to find out.\nThe Summary can be a little exhausting to read as a list. Perhaps a summary per area instead of a listing of the principles?\nIn Section 2.1 White papers, what does “open context” mean? I tried looking this up in the provided citation and it still remains unclear to me.\nIn Section 2 Workshop inputs, it might give a clearer picture instead of giving a paragraph to each author to instead give a paragraph to each theme they touch on. Giving space to talk to about each theme may tease apart where they agree and disagree in terms of emphasis on different aspects of openness. Thus, perhaps Section 2 could be able to provide a synthesis of all views for better comparison.\nIn Section 2.2. Leader survey, why are points 4 and 5 not presented in a visual way like the other questions? A histogram or other visual aid would work well here to show support for the various items.\nIn Section 4.5 Preservation and Reproducibility, I think there could be a lot more emphasis on what reproducibility means or at least more of an emphasis on how hard it is to pin down in a more general sense for all of science. Some citations here might help to get the point across that reproducibility is an umbrella term for many related concepts for different disciplines and doesn't only mean being able to rerun the same code. For example, you start the section with “Reproducibility depends on transparently documenting and sharing all data products, protocols, and computational algorithms (with source code) used in the research.” but of course that’s not the whole story. A scientific experiment can be fully documented and transparent and yet not reproducible (given the specific meaning of the word, as it can vary — see the citations I added for details on what I mean if need be).\nI cannot agree more with this sentence: “These arguments could be more successful than negative ones around lack of reproducibility (the “crisis narrative”).“ Transcending the crisis-style rhetoric seems like a sensible move.\nAnother useful part is where you stress inclusivity and diversity. Why make things open if we’re just sharing them with the same old networks of privileged people?\nTo conclude, I enjoyed reading this and I hope my feedback is useful.\n*** I have included some potential useful references.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1926
https://f1000research.com/articles/7-1924/v1	11 Dec 18	{ "type": "Systematic Review", "title": "Association between gout and atrial fibrillation: A meta-analysis of observational studies", "authors": [ "Keith Sai Kit Leung", "Mengqi Gong", "Yingzhi Liu", "Rachel Wing Chuen Lai", "Chengsheng Ju", "Fangzhou Liu", "Michael Huen Sum Lam", "Leonardo Roever", "Dong Chang", "Yunlong Xia", "Tong Liu", "Gary Tse", "Ka Hou Christien Li", "Keith Sai Kit Leung", "Mengqi Gong", "Yingzhi Liu", "Rachel Wing Chuen Lai", "Chengsheng Ju", "Fangzhou Liu", "Michael Huen Sum Lam", "Leonardo Roever", "Dong Chang", "Yunlong Xia", "Tong Liu" ], "abstract": "Background: Gout is a systemic inflammatory arthritis characterized by the deposition of monosodium urate crystals due to hyperuricemia. Previous studies have explored the link between gout and atrial fibrillation (AF). Given the increasing prevalence and incidence of gout, there is a need to quantify the relationship between gout and the risk of AF. Therefore, we conducted a systematic review and meta-analysis on this topic. Methods: PubMed and Embase were searched for studies that reported the association between gout and AF using the following search term: (‘Gout’ and ‘Arrhythmia’). The search period was from the start of the database to 3rd August 2018 with no language restrictions. Results: A total of 75 and 22 articles were retrieved from PubMed and Embase, respectively. Of these, four observational studies (three cohort studies, one case-control study) including 659,094 patients were included. Our meta-analysis demonstrated that gout was significantly associated with increased risk of AF (adjusted hazard ratio: 1.31; 95% confidence interval: 1.00-1.70; P = 0.05; I2 = 99%) after adjusting for significant comorbidities and confounders. Conclusions: Our meta-analysis confirms the significant relationship between gout and AF. More data are needed to determine whether this risk can be adequately reduced by urate-lowering therapy.", "keywords": [ "gout", "atrial fibrillation", "meta-analysis" ], "content": "Introduction\n\nGout is a systemic inflammatory arthritis characterized by the deposition of monosodium urate (MSU) crystals due to hyperuricemia. This is defined as serum uric acid levels above 6.8mg/dL and is due to impaired excretion and/or overproduction of uric acid1. Atrial fibrillation (AF) is a multifactorial condition whose prevalence increases with age2–4. Previous studies have suggested a correlation between pro-inflammatory conditions such as gout and cardiovascular diseases5–7, and have explored the link between gout and AF8–13. Given the increasing prevalence and incidence of gout14, there is a need to quantify the relationship between gout and the risk of AF.\n\n\nMethods\n\nThe meta-analysis was conducted in accordance with the MOOSE checklist. PubMed and Embase were searched for studies reporting the association between gout and AF using the search terms: (‘Gout’ and ‘Arrhythmia’). The search period was from the start of the database to 3rd August 2018 with no language restrictions. Two researchers (KSKL and MG) independently conducted the search. Any disagreement was resolved by adjudication from a third reviewer (GT). Study selection was carried out by screening titles of full publications to determine compliance with the following inclusion criteria: (1) retrospective or prospective cohort studies in human subjects with gout and non-gout control group; (2) AF occurrence was reported; (3) adjusted hazard ratio (aHR) with 95% confidence intervals (95% CI) was reported or could be calculated. As all of the included studies reported sufficient information, contact with the original study authors was not required.\n\nThe quality assessment of these studies included in our meta-analysis was performed using the Newcastle-Ottawa Quality Assessment Scale (NOS). The point score system evaluated the categories of study participant selection, comparability of the results and quality of the outcomes. The following characteristics were assessed: (1) representativeness of the exposed cohort; (2) selection of the non-exposed cohort; (3) ascertainment of exposure; (4) demonstration that outcome of interest was not present at the start of the study; (5) comparability of cohorts on the basis of the design or analysis; (6) assessment of outcomes; (7) follow-up period sufficiently long for outcomes to occur; and (8) adequacy of follow-up of cohorts.\n\nData were entered in prespecified spreadsheet in Microsoft Excel. The extracted data elements consisted of (1) surname of first author and publication year; (2) sample size of gout and non-gout cohorts; (3) follow-up duration; (4) population characteristics (age, gender, diabetes mellitus, hypertension, ischemic heart disease or coronary artery disease, chronic heart failure, hyperlipidemia, chronic obstructive pulmonary disease, liver disease).\n\nStatistical analysis was performed using Review Manager (Version 5.3). Heterogeneity was assessed by the I2 statistic. I2>50% reflects significant statistical heterogeneity. Therefore, a random-effects model with the inverse variance heterogeneity method was used. Subgroup analyses were not performed due to the fact that not all studies consistently reported the outcomes for the same subgroups.\n\n\nResults\n\nA total of 97 entries were retrieved from 75 and 22 from PubMed and Embase, respectively. Of these, four observational studies (three cohort studies, one case-control study) including 659,094 patients were included (Figure 1)4–7. The main characteristics of the studies are summarised in Table 1. The four different cohorts were recruited from the United Kingdom, United States and Taiwan. All included studies had NOS scores >= 7, indicating that they were of high quality (Table 2 and Table 3). Our meta-analysis demonstrated that gout was significantly associated with increased risk of AF (adjusted HR = 1.31; 95%CI: 1.00- 1.70; P = 0.05; I2 = 99%) (Figure 2). This was observed after adjusting for comorbidities and confounders of AF.\n\n*Data for Singh JA 2018 could not be extracted as data of AF and non-AF group were not separately displayed in the original paper and calculations are not viable.\n\n\nDiscussion\n\nAtrial cardiomyopathy, in particularly AF, is a significant clinical problem because it predisposes to stroke, which can be debilitating and increase mortality15–17. Numerous predictors of AF have been identified, including co-morbid conditions18, blood biomarkers19–26, and electrocardiographic predictors27–29. The main finding of this meta-analysis is that gout is associated with a 31% higher risk of AF after adjusting for significant comorbidities and confounders.\n\nSeveral mechanisms have been proposed to explain this relationship with reactive oxygen species as a critical player in the pro-inflammatory process13,30. High serum uric acid levels in gout patients overwhelm the uric acid transporter (URAT1) and their influx causes intracellular accumulation of urate, which in turn stimulates NADPH oxidase and subsequent reactive oxygen species (ROS) production. ROS produced activates the downstream ERK1/2 pathway and upregulates Kv1.5 protein expression, resulting in the attenuation of atrial cellular action potential and electrical remodelling of the left atrium, as illustrated in mice models31. Furthermore, the renin-angiotensin-aldosterone system that becomes activated by high urate acid levels in gout patients. Subsequent studies into gout treatment can prevent AF and lower its recurrence rate after pulmonary vein isolation (PVI) ablation32,33. In preclinical studies, urate-lowering therapy can reduce oxidative stress and prevent adverse remodelling of the cardiac chambers33,34. The NLRP3-inflammasome also contributes to the pathogenesis underlying AF in gout and can be activated by MSU crystals in gout35. Its activation triggers the maturation and production of the pro-inflammatory cytokine interleukin-1β (IL-1β), which upregulates the expression of TGF-β1, a key mediator for atrial fibrosis36, which can lead to atrial conduction abnormalities37,38, promoting AF by re-entry39,40.\n\nThe main strength is that it included the largest cohort of ~660000 patients. However, the following limitations remain. Firstly, as the included studies were retrospective, they are susceptible to bias as in all studies of this study design. Secondly, only four studies were included and future studies are needed to confirm the relationship between gout and AF. Thirdly, definitions of AF also differed between studies. It was based on physician diagnosis or ICD-9 criteria from data obtained using administrative databases. Finally, the type of AF, such as paroxysmal, sustained or permanent, was not reported. These limitations could explain the high heterogeneity observed in this meta-analysis.\n\nIn conclusion, our meta-analysis confirms the significant relationship between gout and AF. More data are needed to determine whether this risk can be adequately reduced by urate-lowering therapy.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nSchlesinger N: Diagnosis of gout: clinical, laboratory, and radiologic findings. Am J Manag Care. 2005; 11(15 Suppl): S443–50; quiz S65–8. PubMed Abstract\n\nZhu Y, Pandya BJ, Choi HK: Prevalence of gout and hyperuricemia in the US general population: the National Health and Nutrition Examination Survey 2007-2008. Arthritis Rheum. 2011; 63(10): 3136–41. PubMed Abstract \| Publisher Full Text\n\nMozaffarian D, Benjamin EJ, Go AS, et al.: Heart disease and stroke statistics--2015 update: a report from the American Heart Association. Circulation. 2015; 131(4): e29–322. PubMed Abstract \| Publisher Full Text\n\nKnuiman M, Briffa T, Divitini M, et al.: A cohort study examination of established and emerging risk factors for atrial fibrillation: the Busselton Health Study. Eur J Epidemiol. 2014; 29(3): 181–90. PubMed Abstract \| Publisher Full Text\n\nRidker PM: Inflammation, C-reactive protein, and cardiovascular disease: moving past the marker versus mediator debate. Circ Res. 2014; 114(4): 594–5. PubMed Abstract \| Publisher Full Text\n\nKorantzopoulos P, Letsas KP, Tse G, et al.: Inflammation and atrial fibrillation: A comprehensive review. J Arrhythm. 2018; 34(4): 394–401. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKorantzopoulos P, Letsas K, Fragakis N, et al.: Oxidative stress and atrial fibrillation: an update. Free Radic Res. 2018: 1–11. PubMed Abstract \| Publisher Full Text\n\nDixit S: Atrial fibrillation and inflammation: quest for the missing link in the etiopathogenesis of AF. Heart Rhythm. 2010; 7(2): 155–6. PubMed Abstract \| Publisher Full Text\n\nKuo CF, Grainge MJ, Mallen C, et al.: Impact of gout on the risk of atrial fibrillation. Rheumatology (Oxford). 2016; 55(4): 721–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKim SC, Liu J, Solomon DH: Risk of incident atrial fibrillation in gout: a cohort study. Ann Rheum Dis. 2016; 75(8): 1473–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSingh JA, Cleveland JD: Gout and the risk of incident atrial fibrillation in older adults: a study of US Medicare data. RMD Open. 2018; 4(2): e000712. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKuo YJ, Tsai TH, Chang HP, et al.: The risk of atrial fibrillation in patients with gout: a nationwide population-based study. Sci Rep. 2016; 6: 32220. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGiannopoulos G, Angelidis C, Deftereos S: Gout and arrhythmias: In search for causation beyond association. Trends Cardiovasc Med. 2018; pii: S1050-1738(18)30093-8. PubMed Abstract \| Publisher Full Text\n\nRoddy E, Doherty M: Epidemiology of gout. Arthritis Res Ther. 2010; 12(6): 223. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTse G, Wong CW, Gong M, et al.: Predictive value of inter-atrial block for new onset or recurrent atrial fibrillation: A systematic review and meta-analysis. Int J Cardiol. 2018; 250: 152–56. PubMed Abstract \| Publisher Full Text\n\nTse G, Lip GYH, Liu T: The CHADS2 and CHA2DS2-VASc scores for predicting healthcare utilization and outcomes: Observations on the Appropriate Use and Misuse of Risk scores. Int J Cardiol. 2017; 245: 181–82. PubMed Abstract \| Publisher Full Text\n\nHe J, Tse G, Korantzopoulos P, et al.: P-Wave Indices and Risk of Ischemic Stroke: A Systematic Review and Meta-Analysis. Stroke. 2017; 48(8): 2066–72. PubMed Abstract \| Publisher Full Text\n\nXiong Z, Liu T, Tse G, et al.: A Machine Learning Aided Systematic Review and Meta-Analysis of the Relative Risk of Atrial Fibrillation in Patients With Diabetes Mellitus. Front Physiol. 2018; 9: 835. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCheung A, Gong M, Bellanti R, et al.: Cancer antigen-125 and risk of atrial fibrillation: a systematic review and meta-analysis. Heart Asia. 2018; 10(1): e010970. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nQi W, Zhang N, Korantzopoulos P, et al.: Serum glycated hemoglobin level as a predictor of atrial fibrillation: A systematic review with meta-analysis and meta-regression. PLoS One. 2017; 12(3): e0170955. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShao Q, Korantzopoulos P, Letsas KP, et al.: Red blood cell distribution width as a predictor of atrial fibrillation. J Clin Lab Anal. 2018; 32(5): e22378. PubMed Abstract \| Publisher Full Text\n\nWeymann A, Ali-Hasan-Al-Saegh S, Popov AF, et al.: Haematological indices as predictors of atrial fibrillation following isolated coronary artery bypass grafting, valvular surgery, or combined procedures: a systematic review with meta-analysis. Kardiol Pol. 2018; 76(1): 107–18. PubMed Abstract \| Publisher Full Text\n\nWeymann A, Ali-Hasan-Al-Saegh S, Sabashnikov A, et al.: Prediction of New-Onset and Recurrent Atrial Fibrillation by Complete Blood Count Tests: A Comprehensive Systematic Review with Meta-Analysis. Med Sci Monit Basic Res. 2017; 23: 179–222. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWeymann A, Popov AF, Sabashnikov A, et al.: Baseline and postoperative levels of C-reactive protein and interleukins as inflammatory predictors of atrial fibrillation following cardiac surgery: a systematic review and meta-analysis. Kardiol Pol. 2018; 76(2): 440–51. PubMed Abstract \| Publisher Full Text\n\nZhang N, Fan C, Gong M, et al.: Leucocyte telomere length and paroxysmal atrial fibrillation: A prospective cohort study and systematic review with meta-analysis. J Clin Lab Anal. 2018; 32(9): e22599. PubMed Abstract \| Publisher Full Text\n\nZhang N, Tse G, Liang X, et al.: Telomere length: A newly marker for predicting atrial fibrillation? Int J Cardiol. 2017; 239: 21. PubMed Abstract \| Publisher Full Text\n\nAlexander B, Haseeb S, van Rooy H, et al.: Reduced P-wave Voltage in Lead I is Associated with Development of Atrial Fibrillation in Patients with Coronary Artery Disease. J Atr Fibrillation. 2017; 10(4): 1657. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTse G, Liu T, Baranchuk A: Authors' reply: Low prevalence of inter-atrial block in the general population from China: A possible reason for its low rates of atrial fibrillation. Int J Cardiol. 2018; 260: 112. PubMed Abstract \| Publisher Full Text\n\nZhang N, Gong M, Tse G, et al.: Prolonged corrected QT interval in predicting atrial fibrillation: A systematic review and meta-analysis. Pacing Clin Electrophysiol. 2018; 41(3): 321–27. PubMed Abstract \| Publisher Full Text\n\nTse G, Yan BP, Chan YW, et al.: Reactive Oxygen Species, Endoplasmic Reticulum Stress and Mitochondrial Dysfunction: The Link with Cardiac Arrhythmogenesis. Front Physiol. 2016; 7: 313. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMaharani N, Kuwabara M, Hisatome I: Hyperuricemia and Atrial Fibrillation. Int Heart J. 2016; 57(4): 395–9. PubMed Abstract \| Publisher Full Text\n\nDeftereos S, Giannopoulos G, Efremidis M, et al.: Colchicine for prevention of atrial fibrillation recurrence after pulmonary vein isolation: mid-term efficacy and effect on quality of life. Heart Rhythm. 2014; 11(4): 620–8. PubMed Abstract \| Publisher Full Text\n\nSakabe M, Fujiki A, Sakamoto T, et al.: Xanthine oxidase inhibition prevents atrial fibrillation in a canine model of atrial pacing-induced left ventricular dysfunction. J Cardiovasc Electrophysiol. 2012; 23(10): 1130–5. PubMed Abstract \| Publisher Full Text\n\nYang Y, Zhao J, Qiu J, et al.: Xanthine Oxidase Inhibitor Allopurinol Prevents Oxidative Stress-Mediated Atrial Remodeling in Alloxan-Induced Diabetes Mellitus Rabbits. J Am Heart Assoc. 2018; 7(10): pii: e008807. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHari A, Zhang Y, Tu Z, et al.: Activation of NLRP3 inflammasome by crystalline structures via cell surface contact. Sci Rep. 2014; 4: 7281. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYao C, Veleva T, Scott L Jr, et al.: Enhanced Cardiomyocyte NLRP3 Inflammasome Signaling Promotes Atrial Fibrillation. Circulation. 2018; pii: CIRCULATIONAHA. PubMed Abstract \| Free Full Text\n\nTse G, Lai ET, Yeo JM, et al.: Mechanisms of Electrical Activation and Conduction in the Gastrointestinal System: Lessons from Cardiac Electrophysiology. Front Physiol. 2016; 7: 182. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTse G, Yeo JM: Conduction abnormalities and ventricular arrhythmogenesis: The roles of sodium channels and gap junctions. Int J Cardiol Heart Vasc. 2015; 9: 75–82. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTse G, Lai ET, Yeo JM, et al.: Electrophysiological Mechanisms of Bayés Syndrome: Insights from Clinical and Mouse Studies. Front Physiol. 2016; 7: 188. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTse G: Novel conduction-repolarization indices for the stratification of arrhythmic risk. J Geriatr Cardiol. 2016; 13(9): 811–12. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "46171", "date": "26 Mar 2019", "name": "Wisit Cheungpasitporn", "expertise": [ "Reviewer Expertise Systematic reviews and meta-analysis" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nMy points below are for potential improvement of this manuscript, and I have no conflicts of interest. This systematic review needs several points to be significantly addressed as listed below:\n\nThere are very limited number of included studies (three cohort studies, one case-control study). This systematic reviews and meta-analysis has not been registered. Lack of transparency. Please add this point in the limitation. The reason why only 4 studies are included because literature search was too superficial. Hyperuriecemia should also be included in search term. Figure 1, suggest to use PRISMA 2009 Flow Diagram platform\n\nDetailed comments as below:\nIntroduction is too brief and not adequate enough. Need to describe in detail about the impact of atrial fibrillation and gout and how one can affect another This systematic review and meta-analysis has not been registered. Lack of transparency. Please add this point in the limitation. Search terms in PubMed and Embase are different. 4. Please attach search terms that were used in each database as supplement for Data source and search strategies in the manuscript. Please provide details search terms in supplementary documents. Please attach syntax used in each database as supplementary. When Pubmed is used for the search, MESH terms are always recommended to be included. Figure 1, suggest to use PRISMA 2009 Flow Diagram platform The reason why only 4 studies are included because literature search was too superficial. Hyperuriecemia should also be included in search term. There are high heterogeneity observed in this meta-analysis. Minor corrections are needed in English writing as below:\n\nUnited States should be “the United States” “Characteristics for the four studies” should be “Characteristics of the four studies”\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Partly\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Partly\n\nIs the statistical analysis and its interpretation appropriate? Partly\n\nAre the conclusions drawn adequately supported by the results presented in the review? Partly", "responses": [] }, { "id": "46723", "date": "04 Apr 2019", "name": "Per Wändell", "expertise": [ "Reviewer Expertise Cardio-metabolic research (cardio-vascular and metabolic diseases)", "including systematic reviews." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe review article is certainly of interest, but needs improvement on several points. As a complement to the earlier Reviewer (who certainly has given many good suggestions to improve the article) I add: Remarks:\nThe authors performed searches in PubMed and Embase. One way to extend the search is a backwards and forwards search, i.e. to look at the articles in the reference lists of included articles (and on earlier reviews), and to search in Google Scholar which articles have cited the included articles. If not performed, this ought to be done. As pointed out by Reviewer 1, the Introduction is too brief, and could certainly be extended. There are many connections between gout and atrial fibrillation that should be mentioned. For instance, hypertension is common in both conditions, with the use of many antihypertensive agents among certain medications may increase urate levels. As regards diagnosis of gout, there are different definitions of gout, e.g. the ARA criteria. The mentioned definition of gout in the article is too simplified, and needs to be extended. In epidemiologic research the definition of gout differ. The authors should mention how gout has been defined in the included studies. The authors could mention the known prevalence of gout and atrial fibrillation, respectively, in the world, and mention possible differences in different regions of the world, and different populations. This is important for the possibility to generalize their findings. In the Discussion, the connections between gout and atrial fibrillation should also be mentioned.\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Partly\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Partly\n\nIs the statistical analysis and its interpretation appropriate? Partly\n\nAre the conclusions drawn adequately supported by the results presented in the review? Partly", "responses": [] } ]	1	https://f1000research.com/articles/7-1924
https://f1000research.com/articles/7-1917/v1	10 Dec 18	{ "type": "Research Article", "title": "Influence of different stocking density on the growth, feed efficiency, and survival of Majalaya common carp (Cyprinus carpio Linnaeus 1758)", "authors": [ "Mir'atul Hayat", "Rudy Agung Nugroho", "Retno Aryani", "Mir'atul Hayat", "Retno Aryani" ], "abstract": "Background: Stocking density is key to successful Majalaya common carp (Cyprinus carpio Linnaeus 1758) culture which is a valuable fish culture in Indonesia. The aim of the present study was to evaluate the growth statues, feed utilization, and survival rate of Majalaya common carp (reared with different stocking density. Methods: In total, 1400 fish were randomly distributed into four replicates of four different groups of stocking density: 50, 75, 100, and 125 fish m−3. All fish were fed using a satiation method, three times per day with commercial diet for 12 weeks. At the end of the trial week, growth, feed utilization, and survival were determined. Water quality measures, such as dissolved oxygen (mg L-1), temperature (°C), pH, NH3 (mg L-1), and NO2 (mg L-1) were also measured once a week during the trial. Results: Similar weight gain and SGR were found in Majalaya common carp reared at stocking densities of 50 to 100 fish m3. However, 125 fish m-3 density may reduce weight gain and SGR. The average weekly and daily weight gain of Majalaya common carp significantly increased when reared from 50 to 100 fish m-3 and remained constant at 125 fish m-3 density. Meanwhile, feed conversion ratio and survival of Majalaya common carp were not affected by any stocking density. Conclusions: A stocking density of 100 fish m-3 exhibited significantly higher growth of Majalaya common carp in hapa net ponds among the treatment. Temperature ranges of 29.20-33.38°C, pH 7.47-8.22, DO 4.76-7.55 (mg L-1), NH3 0-0.5 mg L-1, and NO2 0-1 mg L-1 were found to provide optimum condition to the fish.", "keywords": [ "Majalaya Common carp", "stocking density", "growth" ], "content": "Introduction\n\nOne of the important factors related to fish culture productivity is stocking density1–4. Past research has found that growth, feed efficiency and survival can be optimized by considering stocking density in fish culture operations5–7. Besides stocking density, water quality is another factor that must be taken into consideration. Water quality is associated with stocking density in term of the availability of food and condition of the environment in fish culture8. Breeding of the Majalaya common carp (Cyprinus carpio Linnaeus 1758) is the result of selection conducted in Indonesia9. The Majalaya carp belongs to the Cyprinidae family and is an important fish to be cultured in Indonesia10. Though several research studies regarding stocking density in some fish have been conducted2,3,11–13, the influence of different stocking densities on the growth, feed efficiency, and survival of the Majalaya common carp in hapa fish ponds has never been determined. Thus, the purpose of the research was to evaluate the growth statues, feed efficiency, and survival rate of the Majalaya common carp, reared at different socking density, viz: 50, 75, 100, and 125 fish m−3 in the hapa fish pond.\n\n\nMethods\n\nIn total, 1400 Majalaya common carp (mean initial weight ±26.22 g, random sex) were distributed into four groups with four replications each groups and reared with different stocking densities: 50, 75, 100, and 125 fish m−3 in hapa fish ponds (1 x 1 x 1.2 m) for 12 weeks. During the trial, all fish were fed with a commercially available diet (PT Japfa Comfeed, No. reg. KKP RI IN 682072012, containing 30% protein, 5.5% fat, and 5% fibre). All fish were fed to satiation three times per day. At the end of the trial, growth parameters for each overall hapa fish pond, such as final weight, weight gain, average weekly weight gain (AWG), daily weight gain (DWG), specific growth rate (SGR), feed conversion ratio (FCR) were determined using an equation previously described by Abdel-Tawwab et al.14, Muchlisin et al.15, Tran-Ngoc et al.,16 Asriqah et al.17. Meanwhile, survival was calculated with equation as used by Nugroho et al.18.\n\nWater quality, such as dissolved oxygen (DO) (mg L-1) and temperature (°C) were assessed using a digital water checker (YSI™ Model 550A DO Meter; Fisher Scientific, USA). The pH was measured with a pH-meter (CyberScan pH 11; EuTech Instruments, Singapore), while NH3 (mg L-1), and NO2 (mg L-1) were detected using a Sera test kit (Sera GmbH D52518, Heinsberg, Germany). All water quality parameter were measured once a week during the trial.\n\nData were analysed using SPSS 22 (SPSS, Inc., USA). Growth, FCR, and survival were subjected to analysis of variance, followed by Duncan post hoc to evaluate significant differences among the groups. Water quality was descriptively analysed. All significant tests were at P<0.05.\n\n\nResults\n\nPresent study showed that stocking density from 50 to 100 fish m3 of Majalaya common carp in the hapa fish pond resulted in similar weight gain and SGR. However, stocking density higher than 100 fish m-3 may reduce weight gain and SGR. The AWG and DWG of Majalaya common carp showed a significantly increase trend when reared from 50 to 100 fish m-3 and remained constant at 125 fish m-3 density. Meanwhile, FCR and survival were not affected by any stocking density (Table 1; raw data available on OSF19). The high density (100 fish m−3) could be more profitable for the Majalaya common carp farms in Indonesia in terms of reduced land cost and facilities.\n\nDifferent superscript letters (a, b, c, d) indicate significantly different means for different group of diets at P<0.05. Initial and final weights are the biomass weights.\n\nWater parameters showed a suitable condition for culturing Majalaya common carp at different stocking density up to 125 fish m-3. The temperature ranged 29.20–33.38°C, pH range of 7.47–8.22, DO of 4.76–7.55 mg L-1, NH3 0–0.5 mg L-1, and NO2 0–1 mg L-1, respectively (Table 2). The data showing the growth parameters and water quality parameters can be seen on OSF19.\n\n\nDiscussion\n\nPrevious research indicated that high growth rates, high levels of survival and better FCR may be due to low feed competition and density2,20,21. The present study stated that a stocking density up to 100 fish m-3 resulted in similar weight gain and SGR, but this was reduced at the highest density (125 fish m-3). Meanwhile, FCR and survival were not affected by any stocking density. This finding is similar to previous research that survival and growth of fish were independent of the stocking density22,23. In addition, the growth and survival of fish in practical culture may also depend on the species. For example, the survival and growth rate of the catfish Rita rita, at different densities of 10, 20 and 30 fish per cistern, resulted in the highest survival and SGR in 20 fish per cistern. Further, no competition for feed and space observed at low density culture of this fish24. In contrast, a prior study revealed that survival rate in aquatic animals was negatively correlated with stocking which could be due to high competition and space for the fish2.\n\nExcess feed remaining in the pond, as well as stocking density, might change the water quality. In this research, the water quality parameters during the trials showed no effects on the growth and survival of fish culture during the trial. The present findings are concomitant with those of previous studies, which found that water quality measures such as temperature, DO, pH, NO2 and NH3 measured in similar current experimental setups are all within the acceptable value for culturing fin fish in tropical regions25,26. The data regarding the growth status and water quality mean, minimum and maximum values can be obtained in Dataset 1.\n\n\nConclusion\n\nThe Majalaya common carp can be reared at stocking density up to 100 fish m-3 without negative effects on the growth, FCR, and survival. The water quality is suitable condition and suggested for culturing Majalaya common carp in hapa fish ponds. Further research needs to be conducted to evaluate the fillet and carcass proximate composition, and immune system of Majalaya common carp when reared at high stocking density.\n\n\nData availability\n\nRaw data for Tables can be accessed on OSF, DOI: https://doi.org/10.17605/OSF.IO/TGC4519.\n\nData are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgments\n\nThe authors thank the PT Suri Tani Pemuka Unit Research and Development Cianjur, East Java, Indonesia for supporting this research with facilities. All authors also thank the Faculty of Mathematics and Natural Sciences, Mulawarman University, Samarinda, East Kalimantan. Our appreciation goes to all of our students who helped the authors during the trial in the field.\n\n\nReferences\n\nAkdemir F, Orhan C, Tuzcu M, et al.: The efficacy of dietary curcumin on growth performance, lipid peroxidation and hepatic transcription factors in rainbow trout Oncorhynchus Mykiss (Walbaum) reared under different stocking densities. Aquac Res. 2017; 48(8): 4012–4021. Publisher Full Text\n\nDaudpota AM, Kalhoro IB, Shah SA, et al.: Effect of stocking densities on growth, production and survival rate of red tilapia in hapa at fish hatchery Chilya Thatta, Sindh, Pakistan. J Fish. 2014; 2(3): 180–186. Publisher Full Text\n\nRoy D, Petrobich ZA, Aleksebich BB, et al.: Intensive polyculture of common carp (Cyprinus carpio), mirror carp (Cyprinus carpio carpio), silver carp (Hypophthalmichthys Molitrix) and grass carp (Ctenopharyngodon Idella) at different stocking densities. Bangladesh J Zool. 2018; 46(1): 71–80. Publisher Full Text\n\nVaishnav M, Sharma SK, Sharma BK, et al.: Growth performance of Pangasius Sp. cultured at different stocking density in floating net cages in Mahi Bajaj Sagar Dam of Banswara (Rajasthan). J Entomol Zool Stud. 2017; 5(5): 649–652. Reference Source\n\nNarejo N, Dayo A, Dars B, et al.: Effect of stocking density on growth and survival rate of Labeo rohita (Hamilton) fed with formulated feed. Sindh University Research Journal-SURJ (Science Series). 2010; 42(1).\n\nAli A, Rahman MR, Hossain MK, et al.: Stocking density effects on growth indices, survival and production of stinging catfish shing (Heteropneustes fossilis) in secondary nursing. Int J Fish Aquat Stud. 2017; 5(6): 269–274. Reference Source\n\nAbdel-Tawwab M: Effects of dietary protein levels and rearing density on growth performance and stress response of Nile tilapia, Oreochromis niloticus (L.). International Aquatic Research. 2012; 4(1): 3. Reference Source\n\nZahidah YAY, Zidni I: Effect of density ratio on performance of nile tilapia and catfish in polyculture fish farming system. Turkish J Zool. 2015. 39: 180–187.\n\nSumantadinata K: Present state of common carp (Cyprinus carpio L.) stocks in Indonesia. Aquaculture. 1995; 129(1–4): 205–209. Publisher Full Text\n\nArisuryanti T, Wibowo AT: Karyotype Ikan Mas (Cyprinus carpio Linnaeus 1758) Majalaya. Journal of Tropical Biodiversity and Biotechnology. 2016; 1(1): 15–19. Publisher Full Text\n\nNuwansi K, Verma A, Tiwari V, et al.: Standardization of the stocking density ratios of Koi carp (Cyprinus carpio var. koi): Goldfish (Carassius auratus) in Polyculture Aquaponic Recirculating System. Turk J Fish Aquat Sc. 2017; 17(6): 1271–1278. Publisher Full Text\n\nSharma J, Chakrabarti R: Larval rearing of common carp Cyprinus carpio: A comparision between natural and artificial diets under three stocking densities. J World Aquac Soc. 1999; 30(4): 490–495. Publisher Full Text\n\nAbdel-Tawwab M, Mousa MA, Sharaf SM, et al.: Effect of crowding stress on some physiological functions of Nile tilapia, Oreochromis niloticus (L.) fed different dietary protein levels. Int J Zool Res. 2005; 1(1): 41–47. Publisher Full Text\n\nAbdel-Tawwab M, Adeshina I, Jenyo-Oni A, et al.: Growth, physiological, antioxidants, and immune response of African catfish, Clarias gariepinus (B.), to dietary clove basil, Ocimum gratissimum, leaf extract and its susceptibility to Listeria monocytogenes infection. Fish Shellfish Immunol. 2018; 78: 346–354. PubMed Abstract \| Publisher Full Text\n\nMuchlisin ZA, Murda T, Yulvizar C, et al.: Growth performance and feed utilization of keureling fish Tor tambra (Cyprinidae) fed formulated diet supplemented with enhanced probiotic. [version 1; referees: 2 approved]. F1000Res. 2017; 6: 137. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTran-Ngoc KT, Dinh NT, Nguyen TH, et al.: Interaction between dissolved oxygen concentration and diet composition on growth, digestibility and intestinal health of Nile tilapia (Oreochromis niloticus). Aquaculture. 2016; 462: 101–108. Publisher Full Text\n\nAsriqah L, Nugroho RA, Aryani R: Effect of various organic acid supplementation diets on Clarias gariepinus BURCHELL, 1822: Evaluation of growth, survival and feed utilization [version 1; referees: 3 approved]. F1000Res. 2018; 7: 1465. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNugroho RA, Manurung H, Nur FM, et al.: Terminalia catappa L. extract improves survival, hematological profile and resistance to Aeromonas hydrophila in Betta sp. Archives of Polish Fisheries. 2017; 25(2): 103–115. Publisher Full Text\n\nNugroho RA: Stocking Density. OSF. Web. 2018.\n\nNarejo NT, Salam MA, Sabur MA, et al.: Effect of stocking density on growth and survival of indigenous catfish, Heteropneustes fossilis (Bloch) reared in cemented cistern fed on formulated feed. Pakistan J Zool. 2005; 37(1): 49–52. Reference Source\n\nDarmawan J, Tahapari E, Suharyanto S, et al.: Growth and survival of larva/patih jambal fish seed (Pangasius djambal) maintained solidly different distribution. Journal of Fisheries and Aquaculture Development. 2017; 2017(4): 1–4. Reference Source\n\nKhattab YA, Abdel-Tawwab M, Ahmad MH: Effect of protein level and stocking density on growth performance, survival rate, feed utilization and body composition of Nile tilapia fry (Oreochromis niloticus L.). In Proceedings of the Sixth International Symposium on Tilapia in Aquaculture. 2004.\n\nKeshavanath P, Gangadhar B, Ramesha T, et al.: Impact of substrates and fish stocking density on growth and production of the Indian major carp, Labeo rohita (Ham.). J Aqua Trop. 2015; 30(1/2): 1–14. Reference Source\n\nJalbani S, Khan P, Narejo NT, et al.: Stocking density and its effect on growth parameters of Catfish Rita rita (Hamilton) reared in cemented cisterns. Pak J Zool. 2018; 50(6): 2371–2373. Publisher Full Text\n\nNRC: Nutrient requirements of warmwater fishes and shellfishes. Washington D. C.: Subcommittee on Warmwater Fish Nutrition. National Research Council. National Academies. 1983. Reference Source\n\nOmosowone O, Dada A, Adeparusi E: Comparison of dietary butyric acid supplementation effect on growth performance and body composition of Clarias gariepinus and Oreochromis niloticus fingerlings. Iran J Fish Sci. 2018; 17(2): 403–412. Publisher Full Text" }	[ { "id": "41741", "date": "10 Jan 2019", "name": "Zainal Abidin Muchlisin", "expertise": [ "Reviewer Expertise Aquaculture" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nTitle: Sufficient.\n\nAbstract: Sufficient.\n\nKeywords: I think it would be better if the keywords are different from the words which already exist in the title.\n\nIntroduction:\nThe introduction should be started with the general information about the culture of Majalaya common carp Cyprinus carpio in Indonesia and the world, not directly with the problem statement. What are the advantages of this species, etc.?\n\nThis species is known as common carp, but in this study the authors add the “Majalaya” common carp as a common name, so the authors have to introduce why the Cyprinus carpio is called Majalaya common carp; maybe this is a new strain or variety which resulted from cross-breeding between species X and species Y.\n\nMethods:\nPlease clarify where the experimental fish come from - whether they come from the wild or from a hatchery. If from a hatchery, from which location? Etc.\n\nBefore distribution into the hapas, did you acclimatise the experimental fish? If yes, for how long and what feed were they fed on during acclimatisation?\n\nI suggest deleting the producer name of the feed; just say “commercial diet”.\n\nYou fed the fish three times a day, at what times exactly?\n\nHow many times was the weight gain measured? For example 2 weeks interval for 12 weeks? Etc.\n\nResults:\nPlease add information about the results of the ANOVA test, whether the treatment gave the significant effect or not to the measured parameters.\n\nThe FCR data are lower than 1 in all treatments; it means that to get 1 kg of fish we need 0.7 kg of feed. Are you sure about that? I think it is impossible except you used the additional feed for example plankton, or the plankton were available without attention, where the biomass of these plankton was not included in the calculation. Please clarify.\n\nDiscussion:\nExtend the discussion by comparing with other studies; maybe some studies are in agreement with your findings and some previous studies are contradictory, so that we have discussed both phenomena.\n\nConclusion:\nYou said that “The Majalaya common carp can be reared at stocking density up to 100 fish m-3 without negative effects on the growth” - be careful, can be reared until what size?", "responses": [] }, { "id": "41743", "date": "17 Jan 2019", "name": "Kim T. Tran-Ngoc", "expertise": [ "Reviewer Expertise aquaculture nutrition" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe experimental design and procedures sound good overall. The findings are obtained for better understanding of Majalaya common carp cultured with different stocking density.\n\nI have some comments that may improve your work:\nThe information about the feed is lacking: pellet size? Extruded pellets?\n\nIn Table 1: the initial weight and final weight can be confused between biomass or individual fish. Please add details for this.\n\nIn Table 1: please add the explanation for AWG, DWG, SGR and FCR in the footnote. The table and footnote should be understandable without distracting the reader from the main text.\nAdditional comments:\nComment 1: Table 1 shows that the final biomass weight was significantly different between treatments. You claim it was because of stocking density. From my view, since the initial biomass weight is already significantly different, that can be a main reason that lead to a difference in final weight. Then, either stocking density or initial weight has an effect on final weight. Could you please explain more about this?\n\nComment 2: In the discussion part “this finding is similar to previous research that survival and growth of fish were independent of the stocking density” that from of my view, it is much too subjective. You refer to 2 papers to support your findings.\nHowever, growth, survival and yield effects of stocking density on aquaculture are well known for a diversity of species and seem to impact production differently. Both growth performance and survival rate, for instance, tend to be higher in low stocking density in the African catfish, C. gariepinus (Hecht et al.,19961), Oreochromis spp. (Sorphea et al., 20102) and Thai climbing perch Anabas testudineus (Khatune-Jannat et al., 20123). Therefore, I suggest you should paraphrase this sentence.\n\nComment 3: The feeding was ad libitum in your method. Do you have excess feed remaining in the hapa? If yes, how do you manage it? How is feed intake calculated in this experiment? Does stocking density affect feed intake?", "responses": [] } ]	1	https://f1000research.com/articles/7-1917
https://f1000research.com/articles/7-1391/v1	03 Sep 18	{ "type": "Software Tool Article", "title": "NovoGraph: Genome graph construction from multiple long-read de novo assemblies", "authors": [ "Evan Biederstedt", "Jeffrey C. Oliver", "Nancy F. Hansen", "Aarti Jajoo", "Nathan Dunn", "Andrew Olson", "Ben Busby", "Alexander T. Dilthey", "Evan Biederstedt", "Jeffrey C. Oliver", "Nancy F. Hansen", "Aarti Jajoo", "Nathan Dunn", "Andrew Olson", "Ben Busby" ], "abstract": "Genome graphs are emerging as an important novel approach to the analysis of high-throughput sequencing data. By explicitly representing genetic variants and alternative haplotypes in a mappable data structure, they can enable the improved analysis of structurally variable and hyperpolymorphic regions of the genome. In most existing approaches, graphs are constructed from variant call sets derived from short-read sequencing. As long-read sequencing becomes more cost-effective and enables de novo assembly for increasing numbers of whole genomes, a method for the direct construction of a genome graph from sets of assembled human genomes would be desirable. Such assembly-based genome graphs would encompass the wide spectrum of genetic variation accessible to long-read-based de novo assembly, including large structural variants and divergent haplotypes. Here we present NovoGraph, a method for the construction of a genome graph directly from a set of de novo assemblies. NovoGraph constructs a genome-wide multiple sequence alignment of all input contigs and uses a simple criterion of homologous-identical recombination to convert the multiple sequence alignment into a graph. NovoGraph outputs resulting graphs in VCF format that can be loaded into third-party genome graph toolkits. To demonstrate NovoGraph, we construct a genome graph with 23,478,835 variant sites and 30,582,795 variant alleles from de novo assemblies of seven ethnically diverse human genomes (AK1, CHM1, CHM13, HG003, HG004, HX1, NA19240). Initial evaluations show that mapping against the constructed graph reduces the average mismatch rate of reads from sample NA12878 by approximately 0.2%, albeit at a slightly increased rate of reads that remain unmapped.", "keywords": [ "Genome graph", "de novo assembly", "alignment", "multiple sequence alignment", "population reference graph", "NovoGraph" ], "content": "Introduction\n\nSince the completion of the human reference genome in 2003, genomic sequencing has been established as a key tool for both fundamental research and personalized medicine. Sequencing costs have fallen dramatically, and the whole genomes of tens of thousands of individuals have been sequenced and analyzed. Although long-read sequencing is becoming more cost-effective and popular, the sequencing technologies that currently dominate cohort sequencing produce millions of short reads between 100 and 250 base pairs in length. As the first step of data analysis, these reads are typically mapped to the human reference genome to determine their genomic locations.\n\nThis approach works well for the large majority of reads; critically, however, it fails for reads that come from regions in the sequenced genome that are strongly divergent from the reference genome. Important examples include immunogenetic regions known to harbour important disease-associated variants like the major histocompatibility complex (MHC) and the killer-cell immunoglobulin-like receptor (KIR) genes (Kuśnierczyk, 2013; Trowsdale & Knight, 2013), as well as regions affected by large or complex structural variants, which together account for more than 50% of total base pair differences between individuals (Sudmant et al., 2015). The total proportion of the human genome inaccessible to classical reference-based analysis is estimated to be greater than 1% (Dilthey et al., 2015).\n\nInstead of mapping reads to a single reference genome, it is now possible to map reads to a reference genome graph (Computational Pan-Genomics Consortium, 2018; Paten et al., 2017). A reference genome graph can be thought of as a data structure that provides a unified representation of multiple genomes and their potential recombinants. As more genomes are added to the graph, the probability that any given region in a sequenced genome has a sufficiently close homolog in the graph (so as to allow for reliable mapping) increases. Technically, a genome graph is an acyclic or cyclic graph structure with nucleotide-labeled edges or nodes; each input genome can typically be reconstructed as a traversal of the graph, and nodes with more than one incoming or outgoing edge represent recombination points between the input genomes. Like linear reference genomes, genome graphs can serve as the basis for read mapping and variant calling.\n\nThe utility of reference genome graphs in the field of human genetics was first demonstrated in the field of immunogenetics and subsequently for the entire human genome. Specifically, a reference graph approach to model local haplotype structures enabled improved genotyping accuracy in the MHC (Dilthey et al., 2015) and, for the first time, reliable typing of the Human Leukocyte Antigen (HLA) genes from standard whole-genome sequencing data (Dilthey et al., 2016). More recently, multiple graph approaches and software toolkits suitable for genome-wide application have been published (Eggertsson et al., 2017; Garrison et al., 2017; Rakocevic et al., 2017; Sibbesen et al., 2018), showing, for example, that graph genome approaches can enable a fivefold reduction of missed SNP calls (Eggertsson et al., 2017) and enable the genotyping of thousands of additional variants longer than 50 base pairs per genome (Sibbesen et al., 2018).\n\nThese developments notwithstanding, the field is still in its infancy. One particularly important open question is how to integrate information from long-read sequencing into the graph construction process. In existing approaches, graph construction typically relies on call sets derived from short-read sequencing experiments. As discussed above, however, short-read sequencing has limited sensitivity in the hypervariable and structural-variation-rich regions where graph genomes can be expected to provide the greatest benefit. Therefore, graphs constructed via existing methods likely miss substantial proportions of relevant variation. By contrast, long-read-sequencing enables the assembly of complex sequences (Jain et al., 2018) in a reference-bias-free way and the detection of structural variants at high sensitivity (Sedlazeck et al., 2018). Even though the number of long-read-sequenced samples is still limited, rendering their sequences available via a genome graph would be highly desirable.\n\nHere we introduce NovoGraph, a pipeline for the direct construction of acyclic genome graphs from de novo assembly contigs. NovoGraph constructs a whole-genome graph by connecting the input assembly sequences at positions of homology; that is, it implements a model of homologous recombination between the input genomes. This approach has the advantage that the resulting graph will generally include the complete set of sequences present in the input assemblies, including (at base-pair resolution) the sequences that correspond to structural variants and divergent haplotypes. Graphs constructed by NovoGraph will therefore be comparatively enriched in large-scale structural and complex variants. In the spirit of modularity, constructed graphs are represented in VCF format, which enables them to be used with any of the established genome-wide graph toolkits. We also utilize the standard CRAM format for representing the output of intermediate steps, in particular a multiple sequence alignment of all input sequences.\n\nWe demonstrate NovoGraph by constructing a genome graph from seven ethnically diverse human genomes and the canonical reference. In a mapping experiment with vg (Garrison et al., 2017), we show that using this graph instead of the standard reference genome increases the average alignment identity of genome-wide short reads.\n\nThis project was initiated at an NCBI hackathon (Busby et al., 2016) held before the 2016 Biological Data Science meeting at Cold Spring Harbor Laboratory in October, 2016. The seven co-authors gathered for 3 days at CSHL to quickly develop and prototype the pipeline. As with all NCBI hackathons, the only stipulations for the event were (1) that the data be publicly available and (2) that any resulting software be open-source.\n\n\nMethods\n\nThe NovoGraph pipeline (Biederstedt et al., 2018) for constructing a genome graph from a set of assembly contigs consists of the following steps (see Figure 1):\n\n1. For each input contig, compute a global pairwise alignment to the GRCh38 primary assembly. This alignment determines the approximate placement of each input contig relative to the reference.\n\n2. Compute a global multiple sequence alignment (MSA) between all input contigs and the reference genome. This multiple sequence alignment embodies the joint sequence homology relationships between all input sequences and the reference genome. The pairwise contig-to-reference alignments from Step 1 are used to guide this process.\n\n3. Compute an acyclic directed graph structure from the global MSA, connecting contigs at homologous-identical positions.\n\nOverview of the genome graph generation pipeline presented here. In Step 1, each genome is aligned to a reference genome (GRCh38, shown here in black). In Step 2.1, the pairwise alignments are partitioned into smaller windows for multiple sequence alignment in Step 2.2. Multiple sequence alignments are concatenated into a single alignment in Step 2.3. Finally, in Step 3, the multiple sequence alignment is converted to a single graph representation of the genome, shown in gray. Each individual genome has a single, acyclic path through the genome graph (black, green, blue, and orange paths). The magenta path represents a “mosaic” genome—that is, a path through the graph which was not observed in any genome.\n\nThe outputs from Steps 1 and 2 are represented in SAM/CRAM format (Hsi-Yang Fritz et al., 2011; Li et al., 2009). The output from Step 3 is a VCF (Danecek et al., 2011), which may be provided as input to various existing graph genome frameworks.\n\nFor each input contig, we compute a global pairwise alignment between the input contig sequence and the GRCh38 primary assembly. This process is illustrated in Figure 2.\n\nSchematic of modified Needleman-Wunsch algorithm for global alignment of an input contig to a reference genome. The process starts with local alignments between the contig and the reference genome (blue diagonals; (A). All possible combinations of these local alignments are enumerated by realizing all paths connecting contigs from the upper left to lower right corner of the matrix (B). Each alignment is scored: matches contribute positive scores (dark blue lines in C), while indels (red) and mismatches (gold) incur a penalty (D). The alignment with the highest score is selected as the best global alignment (E) for the next step in graph genome creation; ties among global alignments are resolved arbitrarily.\n\nExact global alignment scales quadratically with the length of the input sequences and therefore quickly becomes computationally intractable as the input sequences increase in size. We therefore adopt a heuristic approach:\n\nFirst, we use bwa-mem (Li, 2013) to identify high-scoring local alignments between the input contig and the reference genome (GRCh38). These represent diagonal (or near-diagonal) moves in a global alignment matrix, i.e. regions of high pairwise alignment identity between the input contig and a reference genome. We refer to the identified local alignments as “diagonals”.\n\nNext, to obtain a global pairwise alignment, we identify the highest-scoring consistent combination of the identified diagonals into a global alignment by dynamic programming. Note that pairwise alignments by definition comprise two sequences in defined orientations; only diagonals that align to the same reference contig in the same orientation (strandedness) can therefore contribute to a consistent global alignment.\n\nWe now give a formal definition of the algorithm. For simplicity, we assume that all identified diagonals align to the same reference contig in the same orientation; if this is not the case, the following algorithm can be executed independently for all reference contig/orientation pairs and their corresponding diagonals, and the best global alignment between the input contig and the reference genome is the best identified alignment over all considered pairs of reference contigs and orientation.\n\nWe define a set P_ENTRY of “path entry” points and a set P_EXIT of “path exit” points. Each element (diagonal_id, (reference_coordinate, input_contig_coordinate)) of these sets consists of a diagonal identifier and a pair of coordinates that specify positions along the reference and input sequences, similar to the coordinates in the classical Needleman-Wunsch dynamic programming matrix. For example, the coordinate pair (3, 2) refers to a state in which 3 characters of the reference and 2 characters of the input sequence have been consumed. We also define the special points ORIGIN as (NA, (0, 0)) and TERMINUS as (NA, (n, m)), where n is the length of the reference sequence, m is the length of the input contig ID, and “NA” stands for an undefined diagonal identifier.\n\nWe populate the sets P_ENTRY and P_EXIT based on the identified diagonals. Each diagonal represents a local pairwise alignment between the reference and the input contig, and is therefore associated with two pairs of coordinates that specify the start and stop of the alignment in the reference and in the contig sequence. Specifically, let (d1, d2) denote the start coordinates of a given diagonal d in the reference and contig sequences, and let (d3, d4) denote the stop coordinates of the alignment in the reference and contig sequences. Both coordinate pairs are 1-based. To give an example, if diagonal d represents an alignment between positions 4 and 10 of the reference sequence and positions 3 and 11 of the contig sequence, d1 = 4, d2 = 3, d3 = 10, and d4 = 11. For each diagonal d, we add (d, (d1, d2)) as a member of the set P_ENTRY and (d, (d3, d4)) as a member of the set P_EXIT. We refer to these as “start-of-diagonal” entry and “end-of-diagonal” exit points. We also add “within-diagonal” path exit points that horizontally or vertically align with start-of-diagonal entry points of other diagonals, and “within-diagonal” path entry points that horizontally or vertically align with end-of-diagonal exit points of other diagonals. Specifically, we add a within-diagonal path exit point (d, (dx, dy)) for diagonal d if and only if (i) the coordinates (dx, dy) correspond to a column in the local alignment associated with d and (ii) there is another diagonal g with g1 = dx or g2 = dy. The definition of within-diagonal path entry points follows symmetrically. The different types of entry and exit points are illustrated in Figure 3.\n\nFocus on entry and exit points for obtaining global alignments. Diagonal blue lines represent local alignments between a reference sequence and an input sequence. Circles indicate types of entry and exit points used in the algorithm to define paths through the alignment space. See text for details of algorithms and formal definitions of entry and exit points.\n\nThe set of valid path traversals is defined as the set of sequences x0,x1,x2,...,xn that meet the following conditions:\n\n(i) for all i such that i is even, xi is a member of {ORIGIN ∪ P_EXIT}\n\n(ii) for all i such that i is odd, xi is a member of {TERMINUS ∪ P_ENTRY}\n\n(iii) x0 = ORIGIN and xn = TERMINUS\n\n(iv) for all xi=(g, (ga, gb)) and xi+1= (h, (ha, hb)), ga ≤ ha and gb ≤ hb\n\n(v) for all xi=(g, (ga, gb)) and xi+1=(h, (ha, hb)) with uneven i, g = h\n\n(vi) each element of the sequence is unique.\n\nEach traversal can be scored iteratively from left to right by combining the scores of the traversed diagonals with gap-incurred penalties from the jumps between exit and entry points. We initialize by setting score(ORIGIN) = 0. For uneven i with xi=(g, (ga, gb)) and xi-1=(h, (ha, hb)), we set score(xi) = score(xi-1) + gap_score x [(ha - ga)+(hb - gb)]. For even i with xi=(g, (ga, gb)) and xi-1=(h, (ha, hb)), g is equal to h by definition and we set score(xi) to be score(xi-1) plus the score of the local alignment on diagonal g between coordinates (ha,hb) and (ga,gb). In the current implementation, gap_score is -1, matches within local alignments are scored as +1, and mismatches/gaps within local alignments as -1. Jumps to ORIGIN and TERMINUS are not penalized along the reference dimension (i.e., ends-free alignment).\n\nA dynamic programming formulation for finding the highest-scoring traversal follows immediately from these definitions. In brief, order the union set S := {ORIGIN ∪ P_ENTRY ∪ P_EXIT ∪ TERMINUS} by coordinates and for the i-th element of the ordered set S, compute the maximum achievable score max_score(xi) of xi by\n\n(i) identifying the subset S’ ⊆ {x0 , .., xi-1} of possible predecessor elements\n\n(ii) for each s ϵ S’, scoring the transition from s to xi by replacing ‘score’ with ‘max_score’ in the definitions of the preceding paragraph\n\n(iii) selecting the maximum-scoring transition as the value for max_score(xi).\n\nWe now turn the pairwise input-contig-to-reference alignments created in Step 1 into a set of global multiple sequence alignments.\n\nWe split the GRCh38 reference contigs into non-overlapping windows of approximately 10,000 bases. A window size of 10,000 is chosen to be both sufficiently large to include the majority of human structural variants and small enough to allow for efficient processing of individual windows; see below for a precise definition of how window boundaries are determined. For each window, we extract the reference sequence and, based on the pairwise input-contig-to-reference alignments, the input contig sequences overlapping the window. We use MAFFT (Katoh & Standley, 2013) to generate an MSA for the sequences of each window (including the reference). This step is trivially parallelizable. After having computed an MSA for each window, we concatenate the per-window MSAs in the correct order. For each GRCh38 reference contig, this yields a combined MSA of the reference sequence and all input contigs initially aligned to it.\n\nIn this approach, the initial pairwise alignments determine in which window a given part of an input sequence ends up for the MSA computation. Ideally we would like to choose the window boundary positions so as to avoid regions of high uncertainty in the initial pairwise alignments. The placement of gaps in sequence alignments is often ambiguous and gaps are generally associated with increased alignment uncertainty. We therefore adopt a simple heuristic to avoid the crossing of gaps when choosing window boundaries: First we partition the reference into windows of exactly 10,000 bases in length. For each window boundary position independently, we scan the surrounding ± 100 reference positions. For each reference position, we identify the columns corresponding to that reference position in the pairwise sequence alignments, and count the number of gaps across the identified columns. We then choose the reference position with the lowest proportion of gaps as the final window boundary.\n\nThe output from this step is encoded in CRAM format. Reference gaps are represented using the ‘P’ CIGAR character.\n\nAs a last step, the multiple sequence alignment generated during the previous step is transformed into a graph. An important design decision for this operation is where to allow for recombination between the input sequences, i.e. where to allow for transitions between sequences encoded on different input contigs. The applied recombination rules shape the topology of the graph and determine the set of genomes that could be sampled from the graph.\n\nGraph topology is also constrained by our requirement that the constructed graph be, for interoperability reasons, representable in VCF format. Some third-party inference methods support fully general VCFs with overlapping variant alleles; other frameworks, for example gramtools (Maciuca et al., 2016), require that the encoded variants be non-overlapping. To achieve full interoperability with different downstream inference methods, NovoGraph therefore implements two separate algorithms for VCF generation.\n\nThe first graph generation algorithm, referred to as NovoGraph-Simple, implements a straightforward conversion of each individual sequence from the multiple sequence alignment into VCF format; that is, the positions at which an input sequence deviates from the reference are identified and represented as variant alleles in VCF format. Identical variant alleles from different MSA sequences are merged and sorted by position to obtain a valid VCF. This algorithm implements a recombination model that allows for recombination between pairs (a, b) of MSA sequences immediately prior to positions at which both a and b align to the reference in the joint MSA with either a match or a mismatch.\n\nThe second graph generation algorithm, referred to as NovoGraph-Universal, is more complex and ensures that the created VCF does not contain overlapping variant alleles; that is, it ensures universal compatibility with third-party inference methods.\n\nNovoGraph-Universal allows for recombination (A) between pairs of input contigs wherever input contigs start or end along the canonical reference, and (B) at positions at which all contig sequences agree with the canonical reference. The graph collapses into a uniformly homozygous state at positions whereby condition (B) applies. The resulting graph structure (composed of reference-identical, collapsed stretches interspersed with sets of alternative haplotypes) lends itself directly to representation in VCF format. Also note that criterion B (sequence identity across all input sequences) is stronger than the recombination condition (sequence identity across pairs of input sequences) of a related algorithm (Dilthey et al., 2015).\n\nAn overview of NovoGraph-Universal is given in Figure 4. At a high level, NovoGraph-Universal constructs a graph by processing the input MSA for each reference contig in a column-by-column fashion from left to right in the order of genomic position, accounting for the entry and exit of input contigs as well as for potential recombination between them. As the algorithm moves along the MSA, it keeps track of the set of haplotypes compatible with the input contigs and their potential recombinants. In the graph, each haplotype is generally represented as its own branch; however, these are collapsed at positions at which all haplotypes agree with the canonical reference. The sequences corresponding to this “collapsed homozygous” state are reference-identical and therefore not explicitly represented in VCF.\n\nGraph representation (left) and unique variants (right) produced by graph genome alignment. From the global multiple sequence alignment, all unique paths through the graph genome are enumerated and written to output. In this example, the reference genome (gray) serves as a scaffold to which all contigs (blue, green, and orange) are aligned. In the first “extension” phase, all unique paths through the graph are identified until deviation from reference genome terminates. At this point, all variant paths are output, or “flushed” to the genome graph output; in this implementation, the variants are written to a VCF file. In the second extension phase, the orange contig deviates again from the reference genome, producing another variant, which, following coalescence back to the reference genome, is “flushed” to the output file.\n\nIn the following, we provide a more detailed description of the algorithm:\n\nNovoGraph-Universal is executed for each GRCh38 reference contig independently. The set of input sequences for each reference contig is represented by the multiple sequence alignment constructed during Step 2. Each non-reference contig in the MSA has a first and a last column in which both the input contig and reference bases are non-gap. We refer to these columns as the entry and exit positions of the contig, and all bases outside the entry and exit columns are ignored during the following steps.\n\nWe keep a set of current haplotypes, denoted as R. Each element h of R (a “current haplotype”) consists of two elements: (i) the “current sequence” of h (which is updated as we move along the MSA) and (ii) the contig ID of the contig that the current haplotype is copying from (the “source haplotype”)—this can be either the reference or one of the input contigs. We initialize R such that R has one element that has a zero-length sequence and that is set to copy its sequence from the reference.\n\nWhen we process a column of the MSA, for each element h ∈ R, we append the corresponding MSA character of the source haplotype to the current sequence of h. This step is called “extension” (see Figure 4).\n\nAfter having carried out the extension step, the current sequences of the elements of R up to the second-last character are sent to the VCF generator if and only if (A) all appended characters are non-gap reference identical and (B) the length of the current sequences of the elements of R is greater than or equal to 2 non-gap bases. This step is called “flushing” (see Figure 4). The VCF generator writes a variant-encoding line to the output VCF file if the received sequences contain at least one non-reference sequence; otherwise, the output is empty. After processing by the VCF generator, the processed strings are removed from their corresponding source elements—that is, after flushing, the current sequences of all elements of R have a length of 1 (the last added base, which was not sent to the VCF generator). Note that R is a set so that, by definition, duplicate elements are collapsed. This process is carried out up to the rightmost column of the MSA, at which point the graph construction and VCF generation process is complete.\n\nTwo special cases corresponding to the entry and exit of non-reference contigs conclude the definition of the graph construction algorithm. First, if an MSA position being processed corresponds to the entry position of a contig, we duplicate all elements of R prior to the extension step, set the source haplotype of the duplicate elements to the ID of the starting contig, and add the modified duplicates to R. Second, if an MSA position being processed corresponds to the exit position of a contig, we execute the following algorithm after the extension step:\n\n1. Compile a list E of all elements of R which use the existing contig as their source haplotype (i.e. the elements R of affected by the contig exit).\n\n2. Compile a list C of non-reference contig IDs that a) are the source haplotype of any current element in R and b) don’t exit at the current MSA position (i.e. C is a list of non-exhausted current contig IDs).\n\n3. For each element (e, c) ∈ {E x C}, we add a new element to R with a) its current sequence set to the current sequence of e and b) its source haplotype set to c. After having processed all elements of the set {E x C} we set the source haplotypes of all e 𝜖 E to the reference.\n\nClearly the size of R increases as non-reference contigs enter and exit and, conversely, the size of R can only decrease during the flushing step. To limit computational demands, we impose an upper limit U1 on the size of R. If \|R\| ≥ U1, we prohibit the entry of new contigs, and when exiting a contig, we only allow the transition to the reference as source haplotype.\n\nFurthermore, due to the requirement of reference identity, gaps in the input MSA along the contig sequence dimension (i.e. corresponding to columns in the MSA in which the input contig sequence is a gap and the reference is not) prevent flushing. We therefore also place an upper limit U2 on the maximum number of contiguous contig gaps in the input alignments. If a contiguous gap along the input contig dimension in an input contig alignment exceeds U2 in size, we break the alignment, i.e. we split the alignment in two. U1 limits the complexity of the graph in terms of the number of per-site variant haplotypes, U2 limits the maximum size of deletions represented in the graph. In the current implementation, we use U1 = U2 = 5000 bp, but both parameters can be easily modified by the user.\n\nSteps 1 and 2 are implemented in Perl 5. Step 3 is implemented in C++, with a wrapper Perl script. Our pipeline utilizes bwa (version 0.7.15 and above), SAMtools (version 1.4 and above), and MAFFT (version 7). The minimum computational requirement for NovoGraph is a workstation computer with at least 32 Gb of RAM; we recommend, however, that the MSA generation steps be executed within a multi-node cluster environment. NovoGraph natively supports SGE-compatible grid environments, although this could be easily adapted to other platforms.\n\nWe used contigs from seven recent de novo assemblies of human genomes (Table 1), the data of which are publicly available. The total size and contig lengths of each input assembly are shown in Figure 5. In order to quantify the sequencing and alignment quality of each input assembly, we relied upon the edit distance (Levenshtein distance) encoded via the BAM NM tag, i.e. the number of nucleotide changes within each contig necessary to equal the reference. The results of dividing this value by the length of each aligned contig (NM/Length) are shown in Figure 6. We note that we have made no effort to classify variants within each assembly as genuine variation or errors.\n\nAssembly sizes (A) and contig length distributions (B) shown in units of base pairs for each input human assembly (see Table 1) used to demonstrate NovoGraph.\n\nReference divergence (edit distance divided by contig length; see text) for each contig within each individual assembly. No effort was made to classify variants within each assembly as genuine variation or errors.\n\nWe used the variation graph toolkit vg (Garrison et al., 2017) to assess the effect of mapping against the constructed human genome graph (based on the NovoGraph-Universal algorithm). Short-read sequencing data of sample NA12878 were obtained from the Platinum Genomes project (2 x 100bp paired-end sequencing reads; European Nucleotide Archive accession ERR194147) and randomly subsampled to 2% of read pairs. We mapped the subsampled reads to the genome graph constructed by us and against a genome graph constructed from the GRCh38 primary reference and assessed the resulting alignment metrics (alignment score, alignment identity, number of mapped reads).\n\n\nResults\n\nWe have presented NovoGraph, a pipeline for the construction of genome graphs from de novo assemblies and applied the pipeline to construct a genome graph from seven high-quality, ethnically diverse human assemblies (Biederstedt, 2018). The graph constructed by NovoGraph-Universal has a size of 17 Gb when stored in uncompressed VCF format and contains 23,478,835 bubbles (i.e. sites with multiple alternative alleles) representing 30,582,795 variant alleles. The graph constructed by NovoGraph-Simple has an uncompressed size of 1.2 GB in VCF format and contains 33,309,666 bubbles representing 34,519,145 variant alleles. Both graphs and intermediate files are available for download and can be used for genome inference with a variety of tools.\n\nWe manually assessed a small set of hyperpolymorphic regions in the human genome. Figure 7 shows an IGV-based visualization (Robinson et al., 2011; Thorvaldsdóttir et al., 2013) of the multiple sequence alignment of the input sequences in the HLA-B region of the MHC. HLA-B is the most polymorphic gene of the human genome and sequence polymorphisms are known to cluster around the peptide-binding-site encoding exons 2 and 3 (Marsh et al., n.d.); consistent with this, high rates of polymorphism are observed in our multiple sequence alignment around these loci.\n\nThe HLA-B region for the genome graph produced by our approach as visualized in the Integrated Genomics Viewer. (a) The coverage (gray bar) of the eight included assemblies (NA19240, HX1, etc.) and the alignment of each to the graph genome. Colored vertical lines indicate sequence variants (green = A, blue = C, orange = G, red = T), horizontal black lines indicate deletions, and vertical purple “I” characters show insertions. (b) Genomic annotations. High rates of polymorphism are observed around peptide-binding-site encoding exons 2 and 3.\n\nTo measure the extent to which mapping against the constructed graph influences alignment metrics, we used the variation graph toolkit vg to map a randomly selected subset of NA12878 reads (see Methods) against a) the genome graph constructed by us (based on the NovoGraph-Universal algorithm) and b) a simple non-branching reference graph constructed from the primary GRCh38 reference alone. Alleles longer than 10 kb in size were removed to ensure successful loading of the graphs into vg. Results of the mapping experiment are shown in Table 2; while mean alignment identity is increased by approximately 0.2%, the number of mapped reads decreases by 0.04%. This somewhat counterintuitive result is probably explained by greater alignment ambiguity for a subset of reads, caused by the presence of non-unique branches in the graph; reads with multiple optimal mapping locations will be assigned a mapping quality score of 0 and count as unmapped.\n\nA total of 2% of NA12878 Illumina Platinum reads were mapped against the NovoGraph-constructed genome graph (“Genome graph”) and against a GRCh38-equivalent genome graph (“Reference graph”; no ALT contigs used). As expected, mapping against the genome graph increases mean alignment scores and alignment identities, albeit at a small reduction in the number of mapped reads.\n\n\nConclusion and next steps\n\nNovoGraph enables the construction of a graph genome from multiple de novo assemblies. The pipeline is available under an open source license and will scale to at least a few dozen input assemblies without major modifications. It would also be straightforward to adapt NovoGraph to non-human species, given the appropriate reference and input assemblies.\n\nIt is instructive to contrast the MSA-based NovoGraph approach with possible alternative approaches in which one creates a separate VCF for each assembly and then builds a graph by combining the individual VCFs. First, carrying out the multiple sequence alignment prior to the VCF generation step enables the sharing of information across multiple samples during the alignment process, potentially improving overall alignment quality and providing more consistent variant definitions across samples. Secondly, the constructed multiple sequence alignment of all input assemblies can be repurposed for other applications, for example as an input to other graph construction algorithms like the Population Reference Graph (Dilthey et al., 2015). Finally, as the number of input genomes increases in size, it will become increasingly necessary to establish the mutual homology relationships between variant alleles from different samples and to represent these in the form of nested graphs; the MSA contains the information necessary for this. As an example, consider the case of two large insertion variants that differ from each other by a single base: in the field of graph genomes, these are most naturally represented as one large insertion with an additional SNP nested into it (instead of two near-identical branches). These points notwithstanding, multiple sequence alignments come at a computational cost, and might prove to be computationally prohibitive if the number of input genomes increases by more than one order of magnitude.\n\nThere are two important directions for future work. First, in the spirit of a hackathon, we have focused our efforts on the software development process. A comprehensive empirical evaluation of the constructed human genome graph is still outstanding. This could be achieved by loading the graph into multiple graph-based inference frameworks and by measuring genome inference accuracy. Secondly, it would be important to better understand the impact on the graph construction process of various parameter settings and trade-offs. For example, in the interest of simplicity, we implemented a simple gap scoring scheme that is neither affine nor convex; we relied on the default settings of MAFFT for the generation of the multiple sequence alignments; and we implemented a naive algorithm to split the reference genome into windows for MSA generation. Exploring alternative choices in each of these cases would be straightforward and could lead to valuable insights. A convex gap scoring scheme would probably improve the alignment of large and complex structural variants (Sedlazeck et al., 2018) and therefore be the most important point to address.\n\nThese limitations notwithstanding, we believe that NovoGraph represents a useful addition to the field of graph genomes. A strength of NovoGraph is its ability to generate genome graphs for all major genome graph approaches directly from de novo assembly data. The graphs constructed with NovoGraph are available for download and could, for example, inform comparisons of different genome graph construction methods and the improved calling of structural variation.\n\n\nData availability\n\nInput assemblies are publicly available and carry the NCBI assembly accession numbers GCA_001750385.2 (AK1), http://identifiers.org/ncbigi/GI:1078263188; GCA_001297185.1 (CHM1), http://identifiers.org/ncbigi/GI:929855629; GCA_001015385.3 (CHM13), http://identifiers.org/ncbigi/GI:953917559; GCA_001549605.1 (HG003), http://identifiers.org/ncbigi/GI:985741195; GCA_001549595.1 (HG004), http://identifiers.org/GI:985734877; GCA_001524155.1 (NA19240), http://identifiers.org/ncbigi/GI:1057722128; GCA_001708065.2 (HX1), http://identifiers.org/ncbigi/GI:1087879108. Full assembly data access details are given in Table 1.\n\nAll NovoGraph output data are available on OSF: https://doi.org/10.17605/OSF.IO/3VS42 (Biederstedt, 2018). The genome graphs of seven ethnically diverse human genomes in VCF format can be downloaded from https://osf.io/t5czk/?view_only=fedd8437d96c4d688f6c40150903d857 (constructed with NovoGraph-Universal) and https://osf.io/pgq52/?view_only=fedd8437d96c4d688f6c40150903d857 (constructed with NovoGraph-Simple). The global multiple sequence of all input sequences in CRAM format can be downloaded from https://osf.io/jhbwx/?view_only=fedd8437d96c4d688f6c40150903d857. OSF data are available under the terms of the Creative Commons Zero \"No rights reserved\" data waiver (CC0 1.0 Public domain dedication).\n\n\nSoftware availability\n\nSource code for the pipeline is available from: https://github.com/NCBI-Hackathons/NovoGraph.\n\nArchived source code at time of publication: https://doi.org/10.5281/zenodo.1342485 (Biederstedt et al., 2018).\n\nLicense: MIT license.", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by the Intramural Research Program of the National Library of Medicine, National Institutes of Health; by the Intramural Research Program of the National Human Genome Research Institute, National Institutes of Health; and by the Jürgen Manchot Foundation.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nThe authors would like to acknowledge the NCBI for facilitating the hackathon and thank Valerie Schneider, Justin Zook, and Lisa Federer for technical discussions. The authors thank Amazon Web Services for the Cloud Credits provided to hackathon participants during the October 2016 hackathon. The authors would also like to acknowledge Richard K. Wilson and the McDonnell Genome Institute, Washington University School of Medicine, for making available the assembly of NA19240.\n\n\nReferences\n\nBiederstedt E: NovoGraph. 2018. http://www.doi.org/10.17605/OSF.IO/3VS42\n\nBiederstedt E, Oliver J, Dunn N, et al.: NCBI-Hackathons/NovoGraph: NovoGraph 1.0.0 (Version v1.0.0). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1342485\n\nBusby B, Lesko M; August 2015 and January 2016 Hackathon participants, et al.: Closing gaps between open software and public data in a hackathon setting: User-centered software prototyping [version 2; referees: not peer reviewed]. F1000Res. 2016; 5: 672. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChaisson MJ, Huddleston J, Dennis MY, et al.: Resolving the complexity of the human genome using single-molecule sequencing. Nature. 2015; 517(7536): 608–611. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nComputational Pan-Genomics Consortium: Computational pan-genomics: status, promises and challenges. Brief Bioinform. 2018; 19(1): 118–135. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDanecek P, Auton A, Abecasis G, et al.: The variant call format and VCFtools. Bioinformatics. 2011; 27(15): 2156–2158. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDilthey A, Cox C, Iqbal Z, et al.: Improved genome inference in the MHC using a population reference graph. Nat Genet. 2015; 47(6): 682–688. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDilthey AT, Gourraud A, Mentzer AJ, et al.: High-Accuracy HLA Type Inference from Whole-Genome Sequencing Data Using Population Reference Graphs. PLoS Comput Biol. 2016; 12(10): e1005151. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEggertsson HP, Jonsson H, Kristmundsdottir S, et al.: Graphtyper enables population-scale genotyping using pangenome graphs. Nat Genet. 2017; 49(11): 1654–1660. PubMed Abstract \| Publisher Full Text\n\nGarrison E, Sirén J, Novak AM, et al.: Sequence variation aware genome references and read mapping with the variation graph toolkit. bioRxiv. 2017. Publisher Full Text\n\nHsi-Yang Fritz M, Leinonen R, Cochrane G, et al.: Efficient storage of high throughput DNA sequencing data using reference-based compression. Genome Res. 2011; 21(5): 734–740. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJain M, Koren S, Miga KH, et al.: Nanopore sequencing and assembly of a human genome with ultra-long reads. Nat Biotechnol. 2018; 36(4): 338–345. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKatoh K, Standley DM: MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Mol Biol Evol. 2013; 30(4): 772–780. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKuśnierczyk P: Killer cell immunoglobulin-like receptor gene associations with autoimmune and allergic diseases, recurrent spontaneous abortion, and neoplasms. Front Immunol. 2013; 4: 8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLi H: Aligning sequence reads, clone sequences and assembly contigs with BWA-MEM. arXiv E-prints. 2013. Reference Source\n\nLi H, Handsaker B, Wysoker A, et al.: The Sequence Alignment/Map format and SAMtools. Bioinformatics. 2009; 25(16): 2078–2079. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMaciuca S, del Ojo Elias C, McVean G, et al.: A Natural Encoding of Genetic Variation in a Burrows-Wheeler Transform to Enable Mapping and Genome Inference. In: Frith M, Storm Pedersen CN. eds. Algorithms in Bioinformatics. Lecture notes in computer science. Cham: Springer International Publishing: 2016; 222–233. Publisher Full Text\n\nMarsh SGE, Parham P, Barber LD: The' ' HLA factsbook. San Diego: Academic Press. Publisher Full Text\n\nPaten B, Novak AM, Eizenga JM, et al.: Genome graphs and the evolution of genome inference. Genome Res. 2017; 27(5): 665–676. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRakocevic G, Semenyuk V, Spencer J, et al.: Fast and Accurate Genomic Analyses using Genome Graphs. bioRxiv. 2017. Publisher Full Text\n\nRobinson JT, Thorvaldsdóttir H, Winckler W, et al.: Integrative genomics viewer. Nat Biotechnol. 2011; 29(1): 24–26. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchneider VA, Graves-Lindsay T, Howe K, et al.: Evaluation of GRCh38 and de novo haploid genome assemblies demonstrates the enduring quality of the reference assembly. Genome Res. 2017; 27(5): 849–864. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSedlazeck FJ, Rescheneder P, Smolka M, et al.: Accurate detection of complex structural variations using single-molecule sequencing. Nat Methods. 2018; 15(6): 461–468. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSeo JS, Rhie A, Kim J, et al.: De novo assembly and phasing of a Korean human genome. Nature. 2016; 538(7624): 243–247. PubMed Abstract \| Publisher Full Text\n\nShi L, Guo Y, Dong C, et al.: Long-read sequencing and de novo assembly of a Chinese genome. Nat Commun. 2016; 7: 12065. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSibbesen JA, Maretty L; Danish Pan-Genome Consortium, et al.: Accurate genotyping across variant classes and lengths using variant graphs. Nat Genet. 2018; 50(7): 1054–1059. PubMed Abstract \| Publisher Full Text\n\nSteinberg KM, Graves-Lindsay T, Schneider VA, et al.: High-Quality Assembly of an Individual of Yoruban Descent. bioRxiv. 2016. Publisher Full Text\n\nSteinberg KM, Schneider VA, Graves-Lindsay TA, et al.: Single haplotype assembly of the human genome from a hydatidiform mole. Genome Res. 2014; 24(12): 2066–2076. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSudmant PH, Rausch T, Gardner EJ, et al.: An integrated map of structural variation in 2,504 human genomes. Nature. 2015; 526(7571): 75–81. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nThorvaldsdóttir H, Robinson JT, Mesirov JP: Integrative Genomics Viewer (IGV): high-performance genomics data visualization and exploration. Brief Bioinform. 2013; 14(2): 178–192. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTrowsdale J, Knight JC: Major histocompatibility complex genomics and human disease. Annu Rev Genomics Hum Genet. 2013; 14(1): 301–323. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZook JM, Catoe D, McDaniel J, et al.: Extensive sequencing of seven human genomes to characterize benchmark reference materials. Sci Data. 2016; 3: 160025. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "38774", "date": "03 Oct 2018", "name": "Bjarni V. Halldorsson", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript describes a method to create a graph genome from a set of input sequences, such as whole genome assembly. This is a topic of current interest to the community and I am not aware of another implementation for this problem. The paper implementation is practical and likely to be useful, but it is also clear that that there is considerable room for improvement and work in this field. The provide an algorithm and software. The algorithm has three steps: 1) Whole genome alignment 2) Local multiple sequence alignment 3) Graph construction/VCF construction from graph. There is a large body of literature on whole genome alignment which might be cited and explained to the reader what the difference is between the current method and previously described ones. Multiple sequence alignment is also a well studied problem, this literature might be cited and it is not clear why MAFFT was chosen over other implementations.\nI would suggest the authors find another word to replace “recombination” in their text, recombination has a well defined meaning in genetics/meiosis which to me seems to be different from the meaning the authors assign to the word. The authors say the branching points in the graph are recombination points between the input genomes, but they might as well be due to SVs. Why do the authors choose to claim that the human reference genome was completed in 2003? There have been a number of updates since then and the reference genome is still filled with gaps. The authors are using GRCh38 which was released in December 2013. Genomic sequencing can be considered to be a key tool in research since shortly after invented, certainly since long before 2003. On page 4, step 2 states that a global MSA is computed, but from figure 2 it is clear that this is in fact a series of local MSAs. I would suggest using another term than global alignment or at least refer to it as inexact/approximate global MSA.\nFrom figure 1, step 2.1 it might be inferred that identifying windows was a more involved process than partitioning the genome into 10kb windows. I am not sure how NovoGraph-Simple is implemented, the details are not described in the manuscript. For NovoGraph-Simple to be useful, it would be very useful if the “The positions where an input sequence deviates from the reference is represented as variant alleles in a valid VCF” was better defined. For inserted sequence, particularly microsatellites/STRs, there are multiple valid VCF records that can be created. Conventions such as left aligning would be useful for the user. If I understand correctly, NovoGraph-Universal is a BFS (breadth first search) of the graph. The authors might make a note of this.\nIt is clear that NovoGraph-Universal is not a practical implementation for people studying a large number of genomes. Whenever a single sequence diverges from the reference all sequences are expanded. Once enough individuals have been sequenced this will lead to VCF files where each individual's’ chromosomes are the alternate allele. I have only evaluated the manuscript and not the software or the VCF files.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [] }, { "id": "38468", "date": "11 Oct 2018", "name": "Korbinian Schneeberger", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors present NovoGraph a method to build up a genome graph from whole-genome alignments between reference genome and de novo assemblies. Instead of building up a genome graph similar to an assembly graph, the genomes are assembled separately and then integrated into a graph based on reference-based whole-genome alignments. Practical application of NovoGraph is shown with a genome graph build up from seven human genome assemblies. The tool targets a practical problem filling a gap in a workflow with otherwise existing software. I would consider this idea as broadly useful and applicable in special cases. A main critic for this work would be its lack of scalability and flexibility (addition/removal of genomes) which might limit its applicability. The manuscript is well and clearly written.\nMajor (which the authors should consider)\n\nIn the title, the authors state that the method is applicable for “long-read de novo assemblies”. Are there any specific reasons that the authors decided to restrict the scope of their method? The method is geared to human genome comparison. This should be made clear in the title and abstract to avoid confusion as it might not be so straightforward to adapt it to other species if type and degree of sequence variation is different (see point 3 and 4). Otherwise, the authors could demonstrate how to adjust the tool for other genomes with higher degree of structural variation as well. If a “consistent global alignment” of a contig (step 1) can only be on one reference contig and can only consider one alignment direction, how are inversion breakpoints and cross-chromosome translocation breakpoints identified? How are large insertions (i.e. sequences not present in the ref seq) represented in step 2? How are the per-window MSA combined if different genomes have different orders of these 10kb windows (e.g. in translocations, inversions…)? Were there contigs that were too divergent to be aligned in the test cases? What if those exist? The authors mentioned that the graph construction is constrained by their requirement to generate VCF files, however, they don’t mention what could be the potential effects of this restriction. The authors have cited work which has not yet been published after peer review. Is this common for F1000research, if not, please mention that these citations are non-peer-reviewed preprints. In order to limit computation load, the method uses hard cut-offs for the number of haplotypes that can be analysed simultaneously. Consequently, new contigs are not added. However, the authors do not describe what happens to those contigs. Are they removed permanently? If yes, then would that lead to potential loss of alternate haplotypes that could be identified from available data?\nMinor\nWe agree with the other reviewer that the authors might want to consider the well-established definitions of “recombination” and “homologous recombination” and perhaps try to find different wording for the branching points in the graph. It is not clear what “homologous-identical recombination” (abstract) or “homologous identical positions” (methods). Similarly, it is conventional to write Directed Acyclic Graph instead of acyclic directed graph, and the authors might want to change that. “uneven” => “odd”? Figure 2 could be extended with cases where “entry” and “exit” points are within alignments as well.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1391
https://f1000research.com/articles/7-1484/v1	18 Sep 18	{ "type": "Systematic Review", "title": "Therapeutic interventions for osteoarthritis of the wrist: a systematic review and meta-analysis", "authors": [ "Benjamin Dean", "Shwan Henari", "Neal Thurley", "Chris Little", "Ian McNab", "Nicholas Riley", "Shwan Henari", "Neal Thurley", "Chris Little", "Ian McNab", "Nicholas Riley" ], "abstract": "Background: In order to evaluate the effectiveness of interventions for osteoarthritis of the wrist in adults we performed a systematic review and meta-analysis. Methods: The MEDLINE and EMBASE via OVID, CINAHL and SPORTDiscus via EBSCO databases were searched from inception to 25th April 2018.All randomised controlled clinical trials (RCTs) and any prospective studies of adults with wrist osteoarthritis investigating any intervention with a comparator were included. Data were extracted and checked for accuracy and completeness by pairs of reviewers. Primary outcomes were pain and function. Comparative treatment effects were analysed by random effects at all time points. Results: Three RCTs were identified for inclusion after screening and all had a high risk of bias. Two compared proximal row carpectomy (PRC) with four corner fusion (4CF) for post-traumatic osteoarthritis, while the other compared leather with commercial wrist splints in patients with chronic wrist pain, of which a small group had wrist osteoarthritis. Conclusion: There is no prospective study comparing operative to non-operative treatment for wrist osteoarthritis, while there is a paucity of prospective studies assessing the effectiveness of both non-operative and operative interventions. Further research is necessary in order to better define which patients benefit from which specific interventions. Registration: The review protocol was registered with PROSPERO under the registration number CRD42018094799.", "keywords": [ "wrist", "osteoarthritis", "ulnocarpal", "radiocarpal", "TFCC", "review", "intervention", "surgery" ], "content": "Introduction\n\nOsteoarthritis of the wrist is a diverse and poorly understood clinical condition, likely relating to the complexity of the anatomy and biomechanics of the human wrist joint1. Wrist pain is a relatively common clinical problem, accounting for an annual consultation prevalence rate of 58 in 10,000 patients in the UK2, which is around one-tenth the rate of consultation for back pain, the most common site of musculoskeletal pain. The prevalence of radiographic wrist osteoarthritis varies widely in the literature3–5, while there is a lack of epidemiological research relating pain to structural change in this condition.\n\nDifferent anatomical areas of the wrist may be affected by osteoarthritis; the radiocarpal, the distal radioulnar, the ulnocarpal, the midcarpal (including the scaphotrapeziotrapzoid (STT)) and the pisotriquetral joint6. Further complexity is added by the variety of terms used to describe ulnocarpal osteoarthritis, such as ulnocarpal abutment and ulnocarpal impaction syndrome, while tears and/or degeneration of the triangular fibrocartilage complex (TFCC) are intimately involved in ulnocarpal osteoarthritis and distal radioulnar joint synovitis7. Osteoarthritis of the wrist may also occur after trauma, such as scaphoid non-union advanced collapse (SNAC), while the pattern of osteoarthritis seen in scapholunate advanced collapse (SLAC) may be post-traumatic or degenerative8.\n\nNon-operative treatment includes activity modification, education, analgesia, physiotherapy, splintage and corticosteroid injections9. Operative treatment is reserved for patients who fail non-operative measures, the specifics of which are determined by the anatomical pattern of the osteoarthritis10. A wide array of operations are carried out on the wrist for osteoarthritis. Wrist arthroscopy has been increasingly performed in recent years, having both diagnostic and therapeutic components, including debridement and TFCC repair11. Other procedures include denervation, excision arthroplasty, partial and total arthroplasty, partial and total wrist fusion and osteotomies. There is a scarcity of published material relating to the number of elective surgical procedures carried out on the wrist. Melamed et al. have described trends in the US relating to total wrist fusion and wrist arthroplasty12. Jain et al. estimated that the number of wrist arthroscopies carried out per annum in the USA was approximately 25,000 in 2006, while the interest in and the number of surgeons performing wrist arthroscopy has been on the rise in recent years11.\n\nOur aim was to perform a systematic review of the effectiveness of available interventions for osteoarthritis of the wrist in terms of patient-reported outcome measures and to assess the rates of adverse outcomes associated with these interventions.\n\n\nMethods\n\nThe systematic review was developed in accordance with the PRISMA statement, using methodology described in the Cochrane Handbook for Systematic Reviews of Interventions. The protocol was developed prospectively and peer reviewed locally before registration on the PROSPERO database (CRD42018094799).\n\nA comprehensive search strategy was created in collaboration with a research librarian (N.T.) and was designed to capture all relevant articles pertaining to inventions for osteoarthritis of the wrist (Supplementary Material 1). The full search strategy is detailed on the PROSPERO website. The search strategy was applied to the following bibliographic databases from database inception until 24th April 2018: MEDLINE and EMBASE via OVID, CINAHL and SPORTDiscus via EBSCO.\n\nThe inclusion and exclusion criteria were defined prospectively during the protocol stage. All prospective studies relating to interventions for osteoarthritis of the wrist were included. Studies had to contain an intervention and a comparator (i.e. both non-randomised controlled trials, and randomised controlled trials, including semi/quasi randomised, cluster randomised trials and prospective comparative case series). Any therapeutic intervention or control treatments were included. We included ulnocarpal, distal radioulnar, radiocarpal, piso-triquetral, scaphotrapeziotrapezoid and midcarpal osteoarthritis. For the purposes of this review we have included all ulnocarpal disorders, such as ulnar abutment, ulnar impaction syndrome and triangular fibrocartilage complex degeneration, which is debatable but arguably fits within the most widely used definition of ‘osteoarthritis’13. Post-traumatic arthritis was also included (SLAC and SNAC). We excluded studies involving children and adolescents (age <18 years), and studies relating to inflammatory arthritis such as rheumatoid arthritis or psoriatic arthritis. Only studies published in a peer-reviewed English-language journal were included.\n\nDuplicates were removed and relevant studies identified from the search were imported into Covidence for screening. Studies were independently screened by title and abstract by two authors (B.J.F.D. and S.H.). This was followed by a full-text evaluation of the selected studies from the first selection step by these authors. Disagreement between the two reviewers was resolved by consensus involving a third author (N.R.)\n\nTwo reviewers (S.H. and B.J.F.D) independently extracted data. Data was extracted using a custom data extraction sheet in Covidence. Any inconsistencies between the two reviewers’ forms were resolved by consensus discussion. A third review (N.R.) was available for any disagreement that could not be resolved by this initial discussion.\n\nIf data was not available from full-text articles or trial registrations, authors were contacted to provide this information. If authors were not contactable for additional data, then this aspect of the study was excluded from the data synthesis. If contactable authors did not respond to initial requests, they were sent two subsequent reminders over a minimum of 6 weeks. If there was still no response for the additional data, then this aspect of the study was excluded from the data synthesis.\n\nIncluded studies were assessed for risk of bias by two independent raters (B.J.F.D. and S.H.) using the Cochrane Collaboration’s tool for assessing risk of bias in randomised trials14. This followed the description in the Cochrane Handbook for Systematic Review of Interventions, version 5.1 (Part 2: 8.5.1)14. Any disagreements between ratings were resolved by discussion between the raters. A third party (N.R.) was available in any case where disagreements persisted after discussion.\n\nDescriptive analysis was performed for all demographic, intervention and outcome data to facilitate narrative interpretation and comparison across studies. It was decided that a direct-comparison meta-analysis would only be performed if data was available for similar time-points, outcomes and interventions across two or more studies. As this was not possible with the identified studies, we conducted a narrative synthesis of the results based on the domains of interest.\n\n\nResults\n\nAfter duplicates were removed, 750 studies were identified by the search. After screening by full-text, three studies were identified as eligible for inclusion (Figure 1). These were all randomized controlled trials (RCTs). The number of studies identified and excluded at each stage is detailed in Figure 1.\n\nStudy characteristics of the included trials including the interventions and comparators are provided in Table 1. Of the three included randomised controlled trials, two compared proximal row carpectomy with four corner fusion for post-traumatic osteoarthritis15,16, while the remaining RCT compared leather with commercial wrist splints in patients with chronic wrist pain, of which a small group had wrist osteoarthritis17. Table 2 details the basic demographics of the intervention and comparator groups, as well as the details about the outcome data provided. The full details of all included studies and the forest plots are included within the Supplementary Material 3–7.\n\nSNAC, scaphoid non-union advanced collapse; RCT, randomised controlled trial; DASH, disabilities of arm, hand and shoulder; VAS, visual analogue score; SLAC, scapholunate advanced collapse; AUSCAN, Australian/Canadian Hand Osteoarthritis Index; COPM, Canadian Occupational Performance Measure.\n\nSNAC, scaphoid non-union advanced collapse; SLAC, scapholunate advanced collapse; OA, osteoarthritis.\n\nThe RCT by Aita et al. compared PRC (13 patients) with 4CF (14 patients) in adults with stage 2 SNAC. There were significant baseline differences between the two groups, the PRC group was significantly younger and had significantly greater range of movement pre-operatively. In terms of outcomes there was no observed difference in disabilities of arm, shoulder and hand (DASH), visual analogue score (VAS) pain and range of movement between PRC and 4CF, but a greater improvement in grip strength with PRC (SMD 0.61 95% CI 0.04-1.17) at final follow up.\n\nBisneto et al. showed similar outcomes in terms of DASH and VAS pain in an RCT comparing PRC with 4CF in patients with a diagnosis of either post-traumatic SNAC or SLAC (10 patients in each group). The 4CF group did have significantly greater grip strength than the PRC group on the affected side pre-operatively and this persisted post-operatively. Unfortunately, it was not possible to obtain further data from the authors to enable a more detailed analysis.\n\nThiele et al. conducted a single-blind crossover RCT comparing the use of leather and commercial wrist splints for treating chronic wrist pain in adults17. Of 25 included patients only six had the diagnosis of osteoarthritis, with the remaining 19 being inflammatory arthritis. Unfortunately, it was not possible to obtain the data pertaining to this osteoarthritis subgroup and further analysis was precluded.\n\nAita et al. reported one complication in each group16. One patient in the PRC group developed symptomatic radiocarpal osteoarthritis requiring total wrist arthrodesis, while one patient in the 4CF group developed a ‘pseudoarthrosis’ (non union of the intercarpal fusions) which did not require further intervention. Bisneto et al. reported one complication in the 4CF group (‘reflex sympathetic dystrophy’) and five complications in the PRC group (two cases of ‘reflex sympathetic dystrophy’ and three cases synovitis; none of these complications required further surgical intervention15.\n\nAll criteria were judged as low, high or unclear risk of bias. Overall, all studies were deemed to be at a high risk of bias, particularly in terms of allocation concealment, incomplete outcome data and selecting reporting. Full risk of bias assessment is available in Figure 2 and Figure 3. Specifically the studies by Aita et al. and Bisneto et al. were both single-centre studies, did not specify a primary outcome measure and did not fully report outcome data. None of the three studies mentioned power calculations and all studies were relatively small in terms of participant numbers. All three studies used a random method of sequence generation. The studies by Aita et al. and Thiele et al. did blind the outcome assessors; however, the blinding of outcome assessment was not specified by Bisneto et al. Blinding of the participants was not possible in the study by Thiele et al., while it was not made clear in the studies by Aita et al. and Bisneto et al.\n\nMeta analysis. As a result of the degree of heterogeneity in terms of study interventions and the incomplete outcome data, it was determined that a meta-analysis of the outcomes was not possible. We carried out a meta-analysis of adverse events for 4CF vs PRC, as these were reported by two trials15,16 and the forest plot of this data is shown in Figure 4. No significant difference was noted in the risk of adverse events between these two groups (risk ratio 3.08 95% CI 0.69-13.65).\n\n\nDiscussion\n\nBased on this systematic review of the published literature, we have identified that there is no prospective study comparing operative to non-operative treatment for wrist osteoarthritis, while there is a paucity of prospective studies comparing the effectiveness of both non-operative and operative interventions. All three included studies were at high risk of bias, especially regarding reporting and blinding, and had other significant methodological weaknesses. In our opinion, the available evidence to inform treatment choices for this common clinical condition is surprisingly limited.\n\nThe manner in which this lack of evidence results in a difficulty managing patients is saliently demonstrated by the example of the choice between PRC or 4CF for the treatment of SLAC or SNAC. Mulford et al. reviewed the literature in this area and remarked at the highly biased nature of the evidence base, the vast majority of which was retrospective case series18. Of the three studies included in this review, two compared PRC with 4CF; however, the biased and small nature of these studies means that these studies do not substantially aid clinicians in making an evidence based decision in managing this group of patients.\n\nThe management of many other relatively common subtypes of wrist osteoarthritis is similarly hampered by a lack of high-quality evidence. This is demonstrated in the case of degenerative ulnocarpal disorders, which may also referred to as ulnar abutment, or ulnar impaction syndrome where the available evidence base consists exclusively of low-quality and highly biased retrospective case series, which invariably demonstrate no meaningful difference in outcome between different surgical interventions19–22. Not only are there no prospective studies, but to our knowledge no study has compared non-operative with operative intervention. This leaves both patients and clinicians with a dilemma in which the evidence base makes it very hard to decisively support any specific intervention.\n\nThe prevalence of radiographic wrist osteoarthritis varies widely in the literature. The seminal epidemiological study by Kellgren and Lawrence demonstrated a prevalence of radiographic wrist osteoarthritis of approximately 10% in men and 5% in women, notably this was in the 55 to 64 age bracket4. Van Saase et al. demonstrated a prevalence of 9.1% and 12.4% in the 55–59 and the 60–64 year old age groups respectively5. A much lower prevalence was reported in the Framingham study of between 1% and 2% in a group of mean age 58.9 years3. Katayama reported a prevalence of radiographic ulnar sided wrist osteoarthritis in 12.8% of patients presenting to their orthopaedic wrist service with a mean age of 53.8 years; however, this study did not investigate the relationship between radiographic change and symptoms23. There is also a paucity of high quality epidemiological evidence relating to the association between radiological changes and patient reported symptoms such as pain and dysfunction. The only studies which have investigated how radiological changes may relate to wrist pain have been conducted in young gymnastic populations24.\n\nThe absence of high quality evidence presents a challenge to clinicians treating wrist osteoarthritis. Although wrist osteoarthritis is more heterogeneous and less common than osteoarthritis affecting other joints such as the hip and knee, this should not mean that prospective evidence is perpetually absent. Whether in the form of prospective case series, cohort studies or randomised clinical trials, it is clear that higher-quality evidence is needed to guide practice in the management of wrist osteoarthritis. This may relate to simple non-operative interventions such as splintage, all the way up to the more invasive procedures such as total wrist arthroplasty and total wrist arthrodesis.\n\n\nConclusions\n\nThere is no prospective study comparing operative and non-operative treatment for wrist osteoarthritis, and an overall paucity of prospective studies comparing the effectiveness of both non-operative and operative interventions. Further research is necessary in order to better define which patients benefit from which specific interventions.\n\n\nData availability\n\nDataset 1. Extracted study data, available as a RevMan 5 file. RevMan 5 can be downloaded from: https://community.cochrane.org/help/tools-and-software/revman-5/revman-5-download. DOI: https://doi.org/10.5256/f1000research.16218.d21762525.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nSupplementary material\n\nSupplementary Material 1. Full search histories.\n\nClick here to access the data.\n\nSupplementary Material 2. Completed PRISMA checklist.\n\nClick here to access the data.\n\nSupplementary Material 3. Forest plot of Aita et al.16 DASH in PRC versus four corner fusion.\n\nClick here to access the data.\n\nSupplementary Material 4. Forest plot of Aita et al.16 VAS in PRC versus four corner.\n\nClick here to access the data.\n\nSupplementary Material 5. Forest plot of Aita et al.16 Grip strength in PRC versus four corner fusion.\n\nClick here to access the data.\n\nSupplementary Material 6. Forest plot of Aita et al.16 ROM wrist in PRC versus four corner fusion.\n\nClick here to access the data.\n\n\nReferences\n\nKijima Y, Viegas SF: Wrist anatomy and biomechanics. J Hand Surg Am. 2009; 34(8): 1555–63. PubMed Abstract \| Publisher Full Text\n\nJordan KP, Kadam UT, Hayward R, et al.: Annual consultation prevalence of regional musculoskeletal problems in primary care: an observational study. BMC Musculoskelet Disord. 2010; 11: 144. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHaugen IK, Englund M, Aliabadi P, et al.: Prevalence, incidence and progression of hand osteoarthritis in the general population: the Framingham Osteoarthritis Study. Ann Rheum Dis. 2011; 70(9): 1581–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKellgren JH, Lawrence JS: Radiological assessment of osteo-arthrosis. Ann Rheum Dis. 1957; 16(4): 494–502. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nvan Saase JL, van Romunde LK, Cats A, et al.: Epidemiology of osteoarthritis: Zoetermeer survey. Comparison of radiological osteoarthritis in a Dutch population with that in 10 other populations. Ann Rheum Dis. 1989; 48(4): 271–80. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nvan Vugt RM, Bijlsma JW, van Vugt AC: Chronic wrist pain: diagnosis and management. Development and use of a new algorithm. Ann Rheum Dis. 1999; 58(11): 665–74. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiebenberg C, Velleman MD, Suleman FE: Ulnar impaction syndrome: A case series and imaging approach. SA orthop J. 2016; 15: 22–7. Publisher Full Text\n\nShah CM, Stern PJ: Scapholunate advanced collapse (SLAC) and scaphoid nonunion advanced collapse (SNAC) wrist arthritis. Curr Rev Musculoskelet Med. 2013; 6(1): 9–17. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLaulan J, Marteau E, Bacle G: Wrist osteoarthritis. Orthop Traumatol Surg Res. 2015; 101(1 Suppl): S1–S9. PubMed Abstract \| Publisher Full Text\n\nWeiss KE, Rodner CM: Osteoarthritis of the wrist. J Hand Surg Am. 2007; 32(5): 725–46. PubMed Abstract \| Publisher Full Text\n\nObdeijn MC, Tuijthof GJ, van der Horst CM, et al.: Trends in wrist arthroscopy. J Wrist Surg. 2013; 2(3): 239–46. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMelamed E, Marascalchi B, Hinds RM, et al.: Trends in the Utilization of Total Wrist Arthroplasty versus Wrist Fusion for Treatment of Advanced Wrist Arthritis. J Wrist Surg. 2016; 5(3): 211–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOARSI: OARSI: definition of osteoarthritis. 2015. Reference Source\n\nHiggins JPT, Green S: Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration, 2011. 2011. Reference Source\n\nBisneto EN, Freitas MC, Paula EJ, et al.: Comparison between proximal row carpectomy and four‐corner fusion for treating osteoarthrosis following carpal trauma: a prospective randomized study. Clinics (Sao Paulo). 2011; 66(1): 51–5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAita MA, Nakano EK, Schaffhausser HL, et al.: Randomized clinical trial between proximal row carpectomy and the four-corner fusion for patients with stage II SNAC. Rev Bras Ortop. 2016; 51(5): 574–82. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nThiele J, Nimmo R, Rowell W, et al.: A randomized single blind crossover trial comparing leather and commercial wrist splints for treating chronic wrist pain in adults. BMC Musculoskelet Disord. 2009; 10: 129. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMulford JS, Ceulemans LJ, Nam D, et al.: Proximal row carpectomy vs four corner fusion for scapholunate (Slac) or scaphoid nonunion advanced collapse (Snac) wrists: a systematic review of outcomes. J Hand Surg Eur Vol. 2009; 34(2): 256–63. PubMed Abstract \| Publisher Full Text\n\nOh WT, Kang HJ, Chun YM, et al.: Arthroscopic Wafer Procedure Versus Ulnar Shortening Osteotomy as a Surgical Treatment for Idiopathic Ulnar Impaction Syndrome. Arthroscopy. 2018; 34(2): 421–30. PubMed Abstract \| Publisher Full Text\n\nKim BS, Song HS: A comparison of ulnar shortening osteotomy alone versus combined arthroscopic triangular fibrocartilage complex debridement and ulnar shortening osteotomy for ulnar impaction syndrome. Clin Orthop Surg. 2011; 3(3): 184–90. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSmet LD, Vandenberghe L, Degreef I: Ulnar Impaction Syndrome: Ulnar Shortening vs. Arthroscopic Wafer Procedure. J Wrist Surg. 2014; 3(2): 98–100. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStockton DJ, Pelletier ME, Pike JM: Operative treatment of ulnar impaction syndrome: a systematic review. J Hand Surg Eur Vol. 2015; 40(5): 470–6. PubMed Abstract \| Publisher Full Text\n\nKatayama T, Ono H, Suzuki D, et al.: Distribution of primary osteoarthritis in the ulnar aspect of the wrist and the factors that are correlated with ulnar wrist osteoarthritis: a cross-sectional study. Skeletal Radiol. 2013; 42(9): 1253–8. PubMed Abstract \| Publisher Full Text\n\nDiFiori JP, Puffer JC, Aish B, et al.: Wrist pain, distal radial physeal injury, and ulnar variance in young gymnasts: does a relationship exist? Am J Sports Med. 2002; 30(6): 879–85. PubMed Abstract \| Publisher Full Text\n\nDean B, Henari S, Thurley N, et al.: Dataset 1 in: Therapeutic interventions for osteoarthritis of the wrist: a systematic review and meta-analysis. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16218.d217625" }	[ { "id": "38438", "date": "08 Oct 2018", "name": "Anju Jaggi", "expertise": [ "Reviewer Expertise Musculoskeletal pain", "shoulder pain", "shoulder instability", "rehabiliation", "physical therapy" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a well presented systematic review paper representing selection criteria, data for analysis, risk of bias very well diagrammatically.\nThe interpretation and conclusion is clear and justified in light of the results, there is a lack of evidence for both patients and clinicians to decide if operative or non-operative interventions are best for osteoarthritis of the wrist being reasonable how important this is in the context of osteoarthritis in general.\n\nI would accept this paper for publication and have no amendments to make however I do not have appropriate statistical knowledge to fully recommend the forest plot data in the study.\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Yes\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Yes\n\nIs the statistical analysis and its interpretation appropriate? I cannot comment. A qualified statistician is required.\n\nAre the conclusions drawn adequately supported by the results presented in the review? Yes", "responses": [] }, { "id": "40020", "date": "03 Dec 2018", "name": "Kirsty Challen", "expertise": [ "Reviewer Expertise Emergency Medicine and Health Services Research" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a well presented systematic review of the published literature on operative and non-operative interventions for wrist osteoarthritis. It could be strengthened with inclusion of grey literature (eg. searches of clinicaltrials.gov).\nI am not sure that it was appropriate to carry out a meta-analysis of adverse outcomes, as these seem to have been defined differently between the studies.\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Yes\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Yes\n\nIs the statistical analysis and its interpretation appropriate? Partly\n\nAre the conclusions drawn adequately supported by the results presented in the review? Yes", "responses": [ { "c_id": "4275", "date": "10 Dec 2018", "name": "Benjamin Dean", "role": "Author Response", "response": "Many thanks for taking the time to review our manuscript. The point regarding the meta-analysis is valid and we shall update the manuscript to include a description of this limitation." } ] }, { "id": "40811", "date": "03 Dec 2018", "name": "Katie Smith", "expertise": [ "Reviewer Expertise MSK research", "systematic reviews", "meta-analysis" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an interesting, well written, and well conducted systematic review highlighting the lack of prospective studies comparing operative and non-operative treatment for wrist osteoarthritis, and lack of evidence evaluating the effectiveness of any intervention for osteoarthritis of the wrist.\n\nThe introduction and method section is clear and appropriate, and the interpretation of the results seems suitable, given the lack of evidence and papers found.\n\nReviewing, and cross checking the PROSPERO, the search terms used seem suitable for the inclusion criteria specified.\n\nThere appears an error on the Exclusion: age >18 years in PROSPERO. Should read ‘age < 18’.\n\nI assume no attempt to search for unpublished data?\n\nAre there any numbers and percentages for agreement/disagreements for inclusion/extraction/scoring?\n\nOnly a discussion point. If ‘there is a lack of epidemiological research relating pain to structural change in this condition’, as mentioned in the introduction, is there an argument to be calling it ‘wrist pain’, rather than ‘wrist OA’?\n\nAre the rationale for, and objectives of, the Systematic Review clearly stated? Yes\n\nAre sufficient details of the methods and analysis provided to allow replication by others? Yes\n\nIs the statistical analysis and its interpretation appropriate? Yes\n\nAre the conclusions drawn adequately supported by the results presented in the review? Yes", "responses": [ { "c_id": "4274", "date": "10 Dec 2018", "name": "Benjamin Dean", "role": "Author Response", "response": "Many thanks for taking the time to review the manuscript.We have changed the erroneous '>' on Prospero to '<'.In terms of review agreement - this is not recorded on Covidence unfortunately.For the purposes of this review it was appropriate to use the term 'wrist osteoarthritis' as the specific scope of the review was to assess the evidence related to surgical intervention for this specific type of structural abnormality. Given the pack of evidence as regards the diagnosis and treatment of common wrist conditions, it is certainly worth considering how one should label these conditions in the future, both in the clinical and research settings. A review focused at 'wrist pain' instead of 'wrist osteoarthritis' would encompass a wider range of clinical conditions including diseases such as inflammatory arthritis, 'non-specific' wrist pain and tendinopathy, and this was not our aim." } ] } ]	1	https://f1000research.com/articles/7-1484
https://f1000research.com/articles/7-459/v1	13 Apr 18	{ "type": "Opinion Article", "title": "The challenges of designing a benchmark strategy for bioinformatics pipelines in the identification of antimicrobial resistance determinants using next generation sequencing technologies", "authors": [ "Alexandre Angers-Loustau", "Mauro Petrillo", "Johan Bengtsson-Palme", "Thomas Berendonk", "Burton Blais", "Kok-Gan Chan", "Teresa M. Coque", "Paul Hammer", "Stefanie Heß", "Dafni M. Kagkli", "Carsten Krumbiegel", "Val F. Lanza", "Jean-Yves Madec", "Thierry Naas", "Justin O'Grady", "Valentina Paracchini", "John W.A. Rossen", "Etienne Ruppé", "Jessica Vamathevan", "Vittorio Venturi", "Guy Van den Eede", "Mauro Petrillo", "Johan Bengtsson-Palme", "Thomas Berendonk", "Burton Blais", "Kok-Gan Chan", "Teresa M. Coque", "Paul Hammer", "Stefanie Heß", "Dafni M. Kagkli", "Carsten Krumbiegel", "Val F. Lanza", "Jean-Yves Madec", "Thierry Naas", "Justin O'Grady", "Valentina Paracchini", "John W.A. Rossen", "Etienne Ruppé", "Jessica Vamathevan", "Vittorio Venturi", "Guy Van den Eede" ], "abstract": "Next-Generation Sequencing (NGS) technologies are expected to play a crucial role in the surveillance of infectious diseases, with their unprecedented capabilities for the characterisation of genetic information underlying the virulence and antimicrobial resistance (AMR) properties of microorganisms. In the implementation of any novel technology for regulatory purposes, important considerations such as harmonisation, validation and quality assurance need to be addressed. NGS technologies pose unique challenges in these regards, in part due to their reliance on bioinformatics for the processing and proper interpretation of the data produced. Well-designed benchmark resources are thus needed to evaluate, validate and ensure continued quality control over the bioinformatics component of the process. This concept was explored as part of a workshop on \"Next-generation sequencing technologies and antimicrobial resistance\" held October 4-5 2017.\n\nChallenges involved in the development of such a benchmark resource, with a specific focus on identifying the molecular determinants of AMR, were identified. For each of the challenges, sets of unsolved questions that will need to be tackled for them to be properly addressed were compiled. These take into consideration the requirement for monitoring of AMR bacteria in humans, animals, food and the environment, which is aligned with the principles of a “One Health” approach.", "keywords": [ "Antimicrobial resistance", "bioinformatics", "next-generation sequencing", "benchmarking" ], "content": "1. Introduction\n\nNext-Generation Sequencing (NGS) technologies are increasingly regarded as an essential tool in modern regulatory frameworks. Monitoring schemes that rely on the characterisation of genetic information will gain considerably by utilising these technologies. Their importance for infectious diseases surveillance was highlighted by \"The Review on Antimicrobial Resistance\" in 2014, which stated that \"advances in genetics, genomics and computer science will likely change the way that infections and new types of resistance are diagnosed, detected and reported worldwide, so that we can fight back faster when bacteria evolve to resist drugs\"1.\n\nThis interest can be observed in the rapid expansion in recent years of whole-genome sequencing capacities in national public health infectious diseases surveillance laboratories, as recently reported in a European survey by the European Centre for Disease Prevention and Control (ECDC)2. Antimicrobial resistance (AMR), i.e. the ability of a microorganism to resist the action of an antimicrobial agent, is of particular importance in this surveillance program. Its observed rise places heavy burdens on healthcare systems, leading to prolonged treatment times, higher mortality and high economic impacts (see 3). In March 2017, the Joint Research Centre organised a meeting in order to better understand the state-of-the-art of the application of NGS technologies in the fight against AMR4. Although it is clear that the uses of NGS vary according to the specific need (e.g. to guide clinical intervention or to evaluate the environmental and human health risks of AMR genetic determinants), these discussions highlighted overlaps in the needs and the challenges of implementing NGS for the monitoring of AMR in humans, animals, food and the environment. Some of these were also highlighted in previous workshops organized by the European Food Safety Authority (EFSA) and the ECDC5,6.\n\nA full regulatory implementation of NGS technologies to monitor AMR will need to address many standardisation challenges throughout the process, which broadly includes sample preparation and DNA extraction, library preparation for sequencing, the use of an NGS instrument for generating the sequences, the bioinformatics analysis, and interpretation and reporting of results (see Figure 1). Focusing on the bioinformatics step, an important shared challenge is the need to correctly and reliably identify the known genomic determinants of AMR from a set of NGS reads produced from sequencing a sample. The ECDC study reported the requirement for sufficient bioinformatics expertise as one of the important hurdles to a more general implementation of NGS for routine testing2. This observation has also been expressed in recent case studies and reviews7–11.\n\nThe benchmark strategy discussed in the current article focuses on the bioinformatics steps, the pipeline converting the output of the sequencing experiment into a list of identified antimicrobial resistance genetic determinants (dashed rectangle).\n\nBy contrast, within the scientific research community the recent literature reflects widespread enthusiasm for the application of NGS approaches to the determination of AMR characteristics in bacteria. For the bioinformatics steps, many useful strategies have been published. These are, however, very varied in the approaches and resources they use. Some start with sequencing reads produced by the Illumina12,13, Ion Torrent14, PacBio15 or Nanopore16 platforms, just to give a few examples. To predict the resistance profile, interesting results were reported with very different strategies, including k-mer analysis of the reads17, sequence comparisons of individual reads to databases12,16, first assembling the reads into contigs using various software packages9,18 and building and comparing de Bruijn graphs of the sequenced sample reads and the reference database19. The reference set of genetic determinants of AMR used by the bioinformatics pipelines also varied, including databases such as ARGANNOT20, CARD16, ResFinder9, Resqu12, ARDB21, custom-generated from Genbank sequences18,22 or combinations of these14. Interestingly, the choice of the database was shown to greatly influence the interpretation of risk associated with AMR in public health23,24. Even individual steps, such as mapping sequenced reads to a reference, can be done with different tools, each carrying their own compromises (see 25–28).\n\nThis complex - and dynamic - reality poses a challenge for the implementation of bioinformatics pipelines in regulatory settings, where the demonstration of reliability and reproducibility is crucial (see also 11,29). Harmonisation approaches must face the variability described above in terms of technologies, strategies, and software used, each with their demonstrated success, limitations and caveats. A further factor influencing the complexity of applying a given bioinformatics pipeline is that new versions of the individual tools that perform tasks such as quality-checking, trimming or assembling the reads, are constantly being released, which may have unanticipated impacts on pipeline performance. Ready-made and/or commercially available solutions that aim to facilitate the implementation of a NGS-based pipeline by lowering the technical skill required (see, for example, 30,31) face the attendant \"black-box\" issues when proposed for regulatory purposes.\n\nIn response to this complex state-of-the-art and the fast-moving environment in which these technologies are developing, efforts for the standardisation and development of best practices have avoided the prescription of restrictive guidelines, methods or technologies in favour of a more flexible approach emphasising quality metrics and fitness-for-purpose32,33. For bioinformatics pipelines, the development of benchmark resources would play an important role in validating specific bioinformatics strategies and workflows, testing any update to the software underlying an established pipeline or allowing proficiency testing of individual laboratories33–35. These resources would need to include a set of inputs for the bioinformatics pipelines (\"in silico reference materials\") linked to a \"correct\" expected output, as well as consideration for the minimum performance requirements to be met by the pipelines. Different initiatives are ongoing to develop these benchmarking resources including, for example, the Critical Assessment of Metagenome Interpretation (CAMI) project for the evaluation methods for metagenome analysis36.\n\nOn the 5th of October 2017, the Joint Research Centre invited experts in the field of AMR monitoring in order to discuss the challenges involved in the development of such a benchmark strategy, for the specific purpose of evaluating the bioinformatics pipelines that transform a set of NGS reads to a characterised AMR profile. The conclusions of these discussions are summarised in Table 1, and discussed in this document.\n\nSee text for details.\n\n\n2. The challenges\n\nAlthough some of the challenges considered reflect the reality of NGS technologies in general, efforts were made to highlight the issues that are specific to the identification of AMR determinants. Broadly, the challenges can be grouped in different, often overlapping categories.\n\nHow should a benchmark strategy handle the current and expanding universe of NGS platforms? What should be the quality profile (in terms of read length, error rate, etc.) of \"in silico reference materials\"? Should different sets of reference materials be produced for each platform? In that case, how to ensure no bias is introduced in the process?\n\nAs described in the Introduction, different NGS technology platforms exist for the generation of sequence data serving as inputs for the bioinformatics processes used in the analysis of AMR determinants. Moreover, the technology continues to evolve rapidly with the advent of what is now termed \"third generation sequencing\" methods that can read the nucleotide sequences at the level of single molecules37. Focusing on validating the technology or the instrument itself is therefore not a useful approach to ensure the reliability of the bioinformatics steps, since it can reasonably be expected that sequencing technologies and protocols will undergo many changes over the coming years. Section 862.2265 of the FDA's Code of Federal Regulations Title 2138 regulates the general use of NGS instruments for clinical use; even when, in this context, devices are cleared as Class II exempt1, laboratories using these instruments must still establish a bioinformatics pipeline for their intended use39. Thus, an effective benchmark strategy will be independent of existing and upcoming NGS technologies, while avoiding any bias that would favour one technology to the detriment of others.\n\nThe proprietary nature of the different raw data outputs produced by the various technologies may not be a primary consideration for present purposes since standard file formats exist that can store raw reads and the associated metadata (ex. QC metrics) produced by the different sequencers. These include FASTQ40 and BAM41, and they have been successfully used in laboratory proficiency testing34,35,42. More recent platforms produce outputs using the HTF5 standard or variants of it; conversion into FASTQ would require an additional computational step, using one of the available tools. However, all platforms (as well as sequencer models and versions within each platform) have differences in the profile and amount of raw reads produced, with variations in their number, length, error rates, error types, etc.43,44. Attempting to create a single set of in silico reference materials would either introduce a bias towards a specific platform and/or create a dataset which is not representative. Creating individual sets of reads would increase the work (with no end in sight as platforms appear or evolve) and require careful consideration to avoid, once again, bias.\n\nAll this highlights a clear challenge, which is how to address both the evolution of the platforms, differences amongst instruments and run-to-run variabilities, in view of the need for benchmark datasets serving as the basis for the validation and harmonisation of NGS approaches in clinical and/or regulatory frameworks.\n\nShould in silico reference material be composed of the output of real experiments, or simulated read sets? If a combination is used, what is the optimal ratio? How is it possible to ensure that the simulated output has been simulated \"correctly\"? For real experiment datasets, how to avoid the presence of sensitive information?\n\nThe core component of a benchmark resource is, by definition, a set of inputs representative of what the benchmarked bioinformatics pipeline is expected to receive in normal, real-life use. A logical source for this dataset, then, is the actual output of laboratory sequencing experiments17,34. However, using data generated by real experiments assumes a high level of quality that will need to somehow be assessed and demonstrated. These experiments will need to be properly characterised in terms of the \"true\" conclusions the benchmarked pipeline is expected to reach. In addition, although there can be actions taken to ensure that most of the host DNA is filtered from the dataset, real experiments from a human source could lead to privacy problems, while samples from food should ensure the absence of information on patented genetically modified food potentially present in the sample8,45. Careful filtering against a standard “exclusion database”, or other adequate strategies, may be necessary to solve this issue - however, the risk is that the filtered dataset is no longer representative of a real experiment, which would contain a fraction of human reads. Experimental data could also be generated using pure cultures of bacteria present as well-characterised strains in biorepositories (see, for example, 46)\n\nThese concerns could be addressed by in silico-generated datasets, where the exact quantity of reads and genes from each source in the composite dataset can be better controlled. Many tools have been developed for this purpose, simulating reads from the different available platforms (see, for example 47–51). Once again, it will be important to properly understand these tools, agree on their applicability for the purpose of generating the desired benchmark datasets, and correctly set their parameters so that the resulting simulations are a correct representation of the \"real\" samples.\n\nRegarding the quality metrics in the benchmark datasets (e.g. error rate, read quality), should these values be fixed for all datasets, or fall within specific ranges? How wide can/should these ranges be?\n\nAvailable published studies of benchmarking NGS bioinformatics pipelines tend to focus on the performance of specific steps at various levels of input quality and/or complexity (SNP rate, GC content, error rate, quality of the reference sequences, contamination, etc.)26,52,53. This is different from a fit-for-purpose evaluation of a complete pipeline under conditions where the quality of the input is guaranteed through the application of best practices and quality control of the laboratory component of the procedure. An important consideration is the extent to which the benchmark should challenge the pipeline robustness by including varying levels of, for example, error rates or reads quality. It is likely that a pipeline that works best under optimal conditions would be sensitive to variation of the sequencing run quality. The extent of desired variation should be agreed upon and captured in the in silico reference material included in the benchmark.\n\nHow should the benchmark manage the different mechanisms by which bacteria acquire resistance? What is the set of resistance genes/mechanisms that need to be included in the benchmark? How should this set be agreed upon?\n\nSeveral mechanisms for the development of resistance to antimicrobials have been characterised54, including: 1) production of an enzyme that digests/metabolizes/modifies the antimicrobial; 2) production of efflux pumps that remove the drug from within the cell; 3) modification, through mutations or biochemical reactions, of the intracellular target of the antimicrobial so that their interaction is lost; 4) activation/upregulation/acquisition of alternate pathways that allow survival through the bypass of the pathway disrupted by the antimicrobial; and 5) downregulation of the expression of the pores through which the drug enters the bacteria.\n\nMechanisms 1), 2) and 4), often involve the acquisition of novel genes by the bacteria from its environment (horizontal transfer) and may be detected, for example, by mapping reads to reference sequence databases that compile such genes. The genetic determinants of mechanisms 3 to 5, however, vary on a case-by-case basis, and may require the detection of Single Nucleotide Polymorphisms (SNPs), insertions/deletions (indels) or variations of copy numbers. These represent different types of bioinformatics determinations which a comprehensive pipeline must be able to resolve, and the benchmark needs to reflect this reality by ensuring that the various types of AMR determinants are correctly represented in the dataset.\n\nMany recent evaluations on the use of NGS for the determination of AMR have emphasised the difficulty of establishing a curated knowledge base on drug resistance genetic determinants to be used as a reference database in NGS data analysis2,13,55. The same problem is mirrored in the design of a benchmark that would ensure all determinants are correctly detected. It is also of foremost relevance to consider that certain genetic determinants such as efflux pumps (mechanism 2 above) are notorious for giving false positive results, as they perform a variety of export functions not necessarily related to antibiotic resistance (see, for example, 56). Eliminating these from the search parameters of bioinformatics pipelines was shown to improve positive predictive value57. The results of testing a pipeline using a benchmark dataset involving all mechanisms must be interpreted with the aim of the pipeline in mind, and this should be taken into account when/if criteria are set (see also section 2.3).\n\nAlternatively, choosing to focus a benchmark dataset on specific resistance mechanisms could simplify the task, but these choices would need to be agreed upon, justified and the limitations clearly stated. This reflection is to be linked to ongoing extensive discussions on the generation of appropriate databases of resistance genes and correct interpretation of resistome profiles (see 24,58). An a priori statement can be made that the benchmark dataset should focus on mechanisms of acquired bacterial resistance. Similarly, for lack of being exhaustive in terms of the AMR genetic determinants it includes, a set of in silico reference materials can be composed of the resistance mechanisms most relevant for public and environmental safety, for example, focusing on certain specific plasmids and AMR genes which have been identified as being important in clinical infections. The decisions through which specific resistance mechanisms are included in/excluded from the benchmark should be clear, transparent, agreed upon and justified in order to ensure that the benchmark is relevant to the types of risks considered and achieves its purpose of quality assurance over time (see also section 2.4).\n\nShould datasets representing different sample types (e.g. isolated clones, environmental samples) be included in the same benchmark? Is a correct representation of different bacterial species (host genomes) important?\n\nThe preceding section focused on the nature of the genetic determinants to be included in the in silico datasets. These sequences (i.e. AMR genes), however, represent a very small fraction of the overall totality of the sequence data generated from biological materials (i.e. bacterial genomes) in a given experiment. The nature of these majority \"background\" reads (bacterial host genomes, other contaminants in the sample etc.) in the components of a proper benchmark dataset thus needs to be carefully considered, as they can influence the accuracy of the pipelines.\n\nThe detection of drug resistance in clinical settings is often performed by sequencing pure cultured isolates18,59,60. Pathogens of particular concern in the context of nosocomial infections will, accordingly, need to be properly represented in the in silico datasets. Lists of AMR pathogens presenting significant risks are maintained (see 61) and include the ESKAPE pathogens (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumanii, Pseudomonas aeruginosa and Enterobacter sp.) and Escherichia coli, among others.\n\nCulture-dependent methods cannot be systematically applied to environmental samples for various reasons, including the fact that most environment bacteria are not recovered under standard culture conditions62. Culture-independent approaches (metagenomics) can then be used to analyse the human and environmental resistomes within complex bacterial populations13,25,63. These approaches have also been proposed for clinical purposes, greatly reducing the time necessary for characterisation8,16. For these samples, agreeing on a realistic genetic diversity within a benchmark64 - a set of communities which can be considered \"representative\" - is a significant challenge as there is tremendous variability in the species composing the microbiomes of different communities13,65–67.\n\nHow can the \"true\" value of the samples, against which the pipelines will be evaluated, be guaranteed? What is needed to demonstrate that the original sample has been correctly characterised, in case real experiments are used?\n\nOne of the objectives of validating a bioinformatics pipeline is to demonstrate that its accuracy is above an acceptable value, with low instances of false negative and false positive results produced68. Antimicrobial susceptibility testing using traditional methods is, in itself, a complex procedure subject to differences in methodologies and interpretations69; hence they have required (and will require) validation and standardisation70–72. There have been reports where discrepancies between NGS-based predictions and susceptibility testing were caused by isolates with inhibition zones close to the susceptibility breakpoint. It was suggested that the results could have been concordant if the susceptibility testing had been performed under different culture conditions, for example, with a different culture medium73. The extent to which these \"borderline\" cases should be included in the benchmark or not, and the final \"correct\" prediction that will be attached to them will need to be carefully considered. It should also be discussed what the most relevant endpoint in this context is, between, for example, the MIC prediction and resistance levels above wildtype/type strain.\n\nThe realities of veterinary medicine, with specific modalities of antimicrobial administration, mean that susceptibility MIC breakpoints may differ between humans and animals74. Thus, the definition of science-based clinical MIC-breakpoints (CBPs) is relevant to interpret results and to harmonise the results of antimicrobial susceptibility testing of veterinary pathogens. Currently, this issue is being discussed in different working groups led by VETCAST. This may cause difficulties in assigning a universal \"correct\" label to some datasets that would apply to both humans and animals.\n\nReference samples of metagenomics experiments are even more complex in this regard, with each sample containing numerous instances of genetic AMR determinants12,14,75. Metagenomics analyses can detect genes (genotype), which are not necessarily translated into resistance (phenotype); expression of the protein(s), which is not directly revealed by DNA sequencing, is important in this context. Assigning accurate profiles to components of a reference dataset will be challenging, as there is no existing pipeline recognised as the 'gold standard' to do so8. Spiked samples or simulated reads may be a necessary initial step in this context.\n\nHow should the target performance thresholds (e.g. specificity, sensitivity, accuracy) for the benchmark suite be set? What is the impact of these targets on the required size of the sample set?\n\nValidation of a process involves the determination of various performance parameters, such as specificity, sensitivity, accuracy, etc.32. When used specifically for the detection of antimicrobial resistance the benchmark resources need to include strict performance thresholds, and whether these should be set a priori along with the levels of these thresholds are subjects for consideration. One also needs to clarify how the process can cope with cases where more than one type of resistance needs to be identified in a single sample, in particular for metagenomics studies.\n\nThese performance parameters will be important, not only as information to be included in the benchmark, but also because they generally have a significant influence on the size of the in silico dataset needed (see, for example, 76,77). Understanding the target performance characteristics of a valid pipeline will be necessary to guide decisions as to how many samples will be needed in the in silico dataset, with respect to the presence or absence of AMR genetic determinants. Finally, not all parameters are equally important for all samples - for example, considerations of sensitivity are generally not relevant in the case of cultured isolates as the bacteria are present in high numbers, but may be crucial for metagenomics experiments where the proportion of the target(s) relative to the background is variable and unknown. Targeted metagenomics seem promising approaches for the accurate detection of minority genes in complex samples13, and challenging the sensitivity of bioinformatics pipelines with a benchmark dataset would be of added value in this context.\n\nHow can the benchmark stay relevant when new resistance mechanisms are regularly characterised? How is the continued quality of the benchmark dataset ensured?\n\nAn important fact concerning antimicrobial resistance - and one of the reasons it represents a global health emergency - is that novel mechanisms of resistance are constantly being reported and new genes and/or vectors of transmission regularly emerge58,78. Assuming that a benchmark resource can be produced covering the existing complexity of AMR determinants (section 2.2), adapting this resource to new information is a challenge that will need to be addressed in order to ensure that its utility does not diminish with time. Criteria for inclusion of new in silico datasets, and the mechanisms by which these decisions should be taken, need to be discussed and agreed upon when developing the resource.\n\nNewly identified genetic determinants can also impact the information linked to existing datasets in the benchmark resources. These datasets will need to be re-evaluated in view of new information to ensure that their AMR determinants are properly characterised. As an example, this issue was evidenced in 2015 with the identification of mcr-1 as a plasmid-borne colistin resistance gene79; re-analysis of existing NGS data from E.coli isolates from food, feed and hospitalised patients for the previous years in Denmark revealed previously characterised samples containing this gene80,81.\n\nWho should generate the benchmark resource? How can it be efficiently shared?\n\nCurrent guidelines and recommendations place the responsibility of validating the bioinformatics pipelines (and ensuring reliability after update of any of its components) with the operator/quality manager of the test facility32,33,39. In fact, thus far, many different sets of benchmark materials and resources have been produced for local use or within collaborative endeavours (see 34). Benchmark datasets have also been used to compare different methods or tools17,82,83. The extent to which these datasets address the concerns described in this document is the subject of a case-by-case evaluation that may become crucial for a wide implementation of NGS technology for routine and regulatory use. An open and inclusive discussion on the different issues (described here or arising upon more detailed considerations) will be important for the development of a resource that can gain wide acceptance and use.\n\n\nConclusions\n\nThe aim of this document is to summarise a list of challenges that were identified at the meeting organised by the Joint Research Centre on the 4th and 5th of October 2017 for the creation of a benchmark resource. The specific objective of this benchmark would be to challenge the bioinformatics step of a workflow to identify antimicrobial resistance in samples, using NGS technologies. It is clear that this covers only a fraction of the work necessary to fully implement this technology in a regulatory context, which will also need to cover additional steps such as the sampling, library preparation, sequencing run, and interpretation of the AMR profiles (see Figure 1). However, this resource would facilitate the implementation of the NGS technology in routine laboratory analyses by:\n\nEnsuring confidence in the implementation of the bioinformatics component of the procedure, a step currently identified as limiting in the field2,8–10.\n\nAllowing evaluation and comparisons of new/existing bioinformatics strategies, resources and tools.\n\nContributing to the validation of specific pipelines and the proficiency testing of testing facilities.\n\n\"Future-proofing\" bioinformatics pipelines to updates and replacement of the tools and resources used in their different steps.\n\nSome of the challenges in building such a resource are common to all NGS-based methods. Many reports on standardisation, quality management and good laboratory practice have focused on clinical testing and the detection of germline sequence variants linked to cancer or other diseases and could guide some of the decisions to be taken. In this context, reference materials were highlighted as necessary for test validation, QC procedures and proficiency testing68. However, many of the challenges also reflect the reality of antimicrobial resistance monitoring and are specific to this framework. How much of the available resources can be directly applied or used to guide future efforts in this field will need evaluation and, eventually, complementation.\n\nAs it was made apparent in the previous sections, many of the challenges are due to the large heterogeneity behind the reality of detecting AMR using NGS. Some of this heterogeneity will require the development of separate benchmark datasets (e.g. the different sequencing platforms) while some will obviously gain by being combined into a single resource (e.g. human and veterinary medicine). Other cases will require more discussions and evaluations of feasibility/added value in being considered together vs separately (e.g. samples composed of isolates vs metagenomics).\n\nWhatever the final composition and number of the benchmark resource(s), the proper path will ensure a holistic view of the problem that also reflects current public health data. This decision-making process should include expertise in AMR characterisation in humans, animals, food and the environment, in order to maximise its impact on the establishment of an AMR surveillance framework that is in line with the principles of a “One Health” approach.\n\n\nDisclaimer\n\nThe contents of this article are the views of the authors and do not necessarily represent an official position of the European Commission.\n\n\nData availability\n\nNo data is associated with this article.", "appendix": "Competing interests\n\n\n\nJOG receives some research funding from Oxford Nanopore Technologies. ER is consultant for Pathoquest.\n\n\nGrant information\n\nThe \"Next-generation sequencing technologies and antimicrobial resistance - Working groups kick-off\" meeting (4–5 October 2017) was funded by the European Commission's Joint Research Centre (JRC), Ispra, Italy.\n\n\nAcknowledgments\n\nWe would like to thank Marc Struelens (European Centre for Disease Prevention and Control, ECDC), Ernesto Liebana Criado and Valentina Rizzi (European Food Safety Authority, EFSA) for their participation to the workshop discussions. We are also grateful to Maddalena Querci and Alex Patak (Joint Research Centre) for their help during the workshop and their review of the manuscript.\n\n\nFootnotes\n\n1See, for example, https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfRL/rl.cfm?lid=427645\n\n\nReferences\n\nO’Neill J: Antimicrobial resistance: tackling a crisis for the health and wealth of nations. Rev Antimicrob Resist. 2014; 1–16. Reference Source\n\nRevez J, Espinosa L, Albiger B, et al.: Survey on the Use of Whole-Genome Sequencing for Infectious Diseases Surveillance: Rapid Expansion of European National Capacities, 2015–2016. Front Public Health. 2017; 5: 347. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nG7 Health Ministers: Berlin Declaration on Antimicrobial Resistance.2015; [cited 2018 Jan 25]. Reference Source\n\nAngers A, Petrillo M, Patak A, et al.: The role and implementation of next-generation sequencing technologies in the coordinated action plan against antimicrobial resistance. Publications Office of the European Union; 2017. Publisher Full Text\n\nEuropean Food Safety Authority (EFSA): EFSA’s 20th Scientific Colloquium on Whole Genome Sequencing of food-borne pathogens for public health protection. EFSA Support Publ. 2015; 12(2). [cited 2018 Feb 20]. Publisher Full Text\n\nEuropean Centre for Disease Prevention and Control: Expert opinion on whole genome sequencing for public health surveillance. Stockholm: ECDC; 2016. Reference Source\n\nDeurenberg RH, Bathoorn E, Chlebowicz MA, et al.: Application of next generation sequencing in clinical microbiology and infection prevention. J Biotechnol. 2017; 243: 16–24. PubMed Abstract \| Publisher Full Text\n\nForbes JD, Knox NC, Ronholm J, et al.: Metagenomics: The Next Culture-Independent Game Changer. Front Microbiol. 2017; 8: 1069. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLesho E, Clifford R, Onmus-Leone F, et al.: The Challenges of Implementing Next Generation Sequencing Across a Large Healthcare System, and the Molecular Epidemiology and Antibiotic Susceptibilities of Carbapenemase-Producing Bacteria in the Healthcare System of the U.S. Department of Defense. PLoS One. 2016; 11(5): e0155770. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPunina NV, Makridakis NM, Remnev MA, et al.: Whole-genome sequencing targets drug-resistant bacterial infections. Hum Genomics. 2015; 9(1): 19. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRossen JWA, Friedrich AW, Moran-Gilad J, et al.: Practical issues in implementing whole-genome-sequencing in routine diagnostic microbiology. Clin Microbiol Infect. 2018; 24(4): 355–360. PubMed Abstract \| Publisher Full Text\n\nBengtsson-Palme J, Boulund F, Fick J, et al.: Shotgun metagenomics reveals a wide array of antibiotic resistance genes and mobile elements in a polluted lake in India. Front Microbiol. 2014; 5: 648, [cited 2017 Apr 5]. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLanza VF, Baquero F, Martínez JL, et al.: In-depth resistome analysis by targeted metagenomics. Microbiome. 2018; 6(1): 11, [cited 2018 Feb 20]. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNoyes NR, Yang X, Linke LM, et al.: Characterization of the resistome in manure, soil and wastewater from dairy and beef production systems. Sci Rep. 2016; 6(1): 24645, [cited 2017 Apr 5]. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYu CY, Ang GY, Chong TM, et al.: Complete genome sequencing revealed novel genetic contexts of the mcr-1 gene in Escherichia coli strains. J Antimicrob Chemother. 2017; 72(4): 1253–5. PubMed Abstract \| Publisher Full Text\n\nSchmidt K, Mwaigwisya S, Crossman LC, et al.: Identification of bacterial pathogens and antimicrobial resistance directly from clinical urines by nanopore-based metagenomic sequencing. J Antimicrob Chemother. 2017; 72(1): 104–14. PubMed Abstract \| Publisher Full Text\n\nClausen PT, Zankari E, Aarestrup FM, et al.: Benchmarking of methods for identification of antimicrobial resistance genes in bacterial whole genome data. J Antimicrob Chemother. 2016; 71(9): 2484–8. PubMed Abstract \| Publisher Full Text\n\nStoesser N, Batty EM, Eyre DW, et al.: Predicting antimicrobial susceptibilities for Escherichia coli and Klebsiella pneumoniae isolates using whole genomic sequence data. J Antimicrob Chemother. 2013; 68(10): 2234–44, [cited 2017 Apr 5]. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBradley P, Gordon NC, Walker TM, et al.: Rapid antibiotic-resistance predictions from genome sequence data for Staphylococcus aureus and Mycobacterium tuberculosis. Nat Commun. 2015; 6: 10063. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGupta SK, Padmanabhan BR, Diene SM, et al.: ARG-ANNOT, a new bioinformatic tool to discover antibiotic resistance genes in bacterial genomes. Antimicrob Agents Chemother. 2014; 58(1): 212–20. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYang Y, Li B, Zou S, et al.: Fate of antibiotic resistance genes in sewage treatment plant revealed by metagenomic approach. Water Res. 2014; 62: 97–106. PubMed Abstract \| Publisher Full Text\n\nGordon NC, Price JR, Cole K, et al.: Prediction of Staphylococcus aureus antimicrobial resistance by whole-genome sequencing. J Clin Microbiol. 2014; 52(4): 1182–91. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBengtsson-Palme J, Larsson DG: Antibiotic resistance genes in the environment: prioritizing risks. Nat Rev Microbiol. 2015; 13(6): 396. PubMed Abstract \| Publisher Full Text\n\nMartínez JL, Coque TM, Baquero F: What is a resistance gene? Ranking risk in resistomes. Nat Rev Microbiol. 2015; 13(2): 116–23. PubMed Abstract \| Publisher Full Text\n\nBengtsson-Palme J, Larsson DGJ, Kristiansson E: Using metagenomics to investigate human and environmental resistomes. J Antimicrob Chemother. 2017; 72(10): 2690–703. PubMed Abstract \| Publisher Full Text\n\nHatem A, Bozdağ D, Toland AE, et al.: Benchmarking short sequence mapping tools. BMC Bioinformatics. 2013; 14(1): 184. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMartínez JL, Coque TM, Lanza VF, et al.: Genomic and metagenomic technologies to explore the antibiotic resistance mobilome. Ann N Y Acad Sci. 2017; 1388(1): 26–41. PubMed Abstract \| Publisher Full Text\n\nZhou J, He Z, Yang Y, et al.: High-throughput metagenomic technologies for complex microbial community analysis: open and closed formats. mBio. 2015; 6(1): pii: e02288–14. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEndrullat C: Standardization in next-generation sequencing-Issues and approaches of establishing standards in a highly dynamic environment. PeerJ Preprints. 2017; 5: e2771v1. Publisher Full Text\n\nInouye M, Dashnow H, Raven LA, et al.: SRST2: Rapid genomic surveillance for public health and hospital microbiology labs. Genome Med. 2014; 6(11): 90. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRowe W, Baker KS, Verner-Jeffreys D, et al.: Search Engine for Antimicrobial Resistance: A Cloud Compatible Pipeline and Web Interface for Rapidly Detecting Antimicrobial Resistance Genes Directly from Sequence Data. PLoS One. 2015; 10(7): e0133492. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEndrullat C, Glökler J, Franke P, et al.: Standardization and quality management in next-generation sequencing. Appl Transl Genom. 2016; 10: 2–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLambert D, Pightling A, Griffiths E, et al.: Baseline Practices for the Application of Genomic Data Supporting Regulatory Food Safety. J AOAC Int. 2017; 100(3): 721–731, [cited 2017 Apr 7]. PubMed Abstract \| Publisher Full Text\n\nDavies KD, Farooqi MS, Gruidl M, et al.: Multi-Institutional FASTQ File Exchange as a Means of Proficiency Testing for Next-Generation Sequencing Bioinformatics and Variant Interpretation. J Mol Diagn. 2016; 18(4): 572–9. PubMed Abstract \| Publisher Full Text\n\nDuncavage EJ, Abel HJ, Pfeifer JD: In Silico Proficiency Testing for Clinical Next-Generation Sequencing. J Mol Diagn. 2017; 19(1): 35–42. PubMed Abstract \| Publisher Full Text\n\nSczyrba A, Hofmann P, Belmann P, et al.: Critical Assessment of Metagenome Interpretation-a benchmark of metagenomics software. Nat Methods. 2017; 14(11): 1063–1071. PubMed Abstract \| Publisher Full Text\n\nLee H, Gurtowski J, Yoo S, et al.: Third-generation sequencing and the future of genomics. BioRxiv. 2016; 048603. Publisher Full Text\n\nFood and Drug Administration: Code of Federal Regulations Title 21—Food and Drugs. Parts. 1970; 10: 121–200. Reference Source\n\nGargis AS, Kalman L, Bick DP, et al.: Good laboratory practice for clinical next-generation sequencing informatics pipelines. Nat Biotechnol. 2015; 33(7): 689–93. PubMed Abstract \| Publisher Full Text\n\nCock PJ, Fields CJ, Goto N, et al.: The Sanger FASTQ file format for sequences with quality scores, and the Solexa/Illumina FASTQ variants. Nucleic Acids Res. 2010; 38(6): 1767–71. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLi H, Handsaker B, Wysoker A, et al.: The Sequence Alignment/Map format and SAMtools. Bioinformatics. 2009; 25(16): 2078–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMellmann A, Andersen PS, Bletz S, et al.: High Interlaboratory Reproducibility and Accuracy of Next-Generation-Sequencing-Based Bacterial Genotyping in a Ring Trial. J Clin Microbiol. 2017; 55(3): 908–13. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nQuail MA, Smith M, Coupland P, et al.: A tale of three next generation sequencing platforms: comparison of Ion Torrent, Pacific Biosciences and Illumina MiSeq sequencers. BMC Genomics. 2012; 13(1): 341. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSuzuki S, Ono N, Furusawa C, et al.: Comparison of sequence reads obtained from three next-generation sequencing platforms. PLoS One. 2011; 6(5): e19534. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGutmann A, Wagner J, Ali Y, et al.: Privacy and progress in whole genome sequencing. Pres Comm Study Bioethical. 2012; (2012). Reference Source\n\nATCC: The Rise of Multidrug-Resistant Strains and Need for New Therapeutic Approaches (White Paper). 2016. Reference Source\n\nHuang W, Li L, Myers JR, et al.: ART: a next-generation sequencing read simulator. Bioinformatics. 2012; 28(4): 593–4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMotro Y, Moran-Gilad J: Microbial Metagenomics Mock Scenario-based Sample Simulation (M3S3). Clin Microbiol Infect. 2018; 24(3): 308.e1–308.e4. PubMed Abstract \| Publisher Full Text\n\nOno Y, Asai K, Hamada M: PBSIM: PacBio reads simulator--toward accurate genome assembly. Bioinformatics. 2013; 29(1): 119–21. PubMed Abstract \| Publisher Full Text\n\nShcherbina A: FASTQSim: platform-independent data characterization and in silico read generation for NGS datasets. BMC Res Notes. 2014; 7(1): 533. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYang C, Chu J, Warren RL, et al.: NanoSim: nanopore sequence read simulator based on statistical characterization. Gigascience. 2017; 6(4): 1–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDrezen E, Lavenier D: Quality metrics for benchmarking sequences comparison tools. In: Brazilian Symposium on Bioinformatics. Springer; 2014. 144–53. Publisher Full Text\n\nSmolka M, Rescheneder P, Schatz MC, et al.: Teaser: Individualized benchmarking and optimization of read mapping results for NGS data. Genome Biol. 2015; 16(1): 235. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLin J, Nishino K, Roberts MC, et al.: Mechanisms of antibiotic resistance. Front Microbiol. 2015; 6: 34. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEllington MJ, Ekelund O, Aarestrup FM, et al.: The role of whole genome sequencing in antimicrobial susceptibility testing of bacteria: report from the EUCAST Subcommittee. Clin Microbiol Infect. 2017; 23(1): 2–22. PubMed Abstract \| Publisher Full Text\n\nDeininger KN, Horikawa A, Kitko RD, et al.: A requirement of TolC and MDR efflux pumps for acid adaptation and GadAB induction in Escherichia coli. PLoS One. 2011; 6(4): e18960. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKnowles M, Stinson S, Lambert D, et al.: Genomic Tools for Customized Recovery and Detection of Foodborne Shiga Toxigenic Escherichia coli. J Food Prot. 2016; 79(12): 2066–77. PubMed Abstract \| Publisher Full Text\n\nBengtsson-Palme J, Kristiansson E, Larsson DGJ: Environmental factors influencing the development and spread of antibiotic resistance. FEMS Microbiol Rev. 2018; 42(1): fux053 [cited 2018 Feb 27]. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEyre DW, Golubchik T, Gordon NC, et al.: A pilot study of rapid benchtop sequencing of Staphylococcus aureus and Clostridium difficile for outbreak detection and surveillance. BMJ Open. 2012; 2(3): pii: e001124. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSherry NL, Porter JL, Seemann T, et al.: Outbreak investigation using high-throughput genome sequencing within a diagnostic microbiology laboratory. J Clin Microbiol. 2013; 51(5): 1396–401. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGarner MJ, Carson C, Lingohr EJ, et al.: An assessment of antimicrobial resistant disease threats in Canada. Conly J, editor. PLoS One. 2015; 10(4): e0125155. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBerendonk TU, Manaia CM, Merlin C, et al.: Tackling antibiotic resistance: the environmental framework. Nat Rev Microbiol. 2015; 13(5): 310–7. PubMed Abstract \| Publisher Full Text\n\nCrofts TS, Gasparrini AJ, Dantas G: Next-generation approaches to understand and combat the antibiotic resistome. Nat Rev Microbiol. 2017; 15(7): 422–434. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLindgreen S, Adair KL, Gardner PP: An evaluation of the accuracy and speed of metagenome analysis tools. Sci Rep. 2016; 6: 19233. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nForster SC: Illuminating microbial diversity. Nat Rev Microbiol. Nature Publishing Group; 2017; 15(10): 578. PubMed Abstract \| Publisher Full Text\n\nGibson MK, Forsberg KJ, Dantas G: Improved annotation of antibiotic resistance determinants reveals microbial resistomes cluster by ecology. ISME J. 2015; 9(1): 207–16. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchlaberg R, Chiu CY, Miller S, et al.: Validation of Metagenomic Next-Generation Sequencing Tests for Universal Pathogen Detection. Arch Pathol Lab Med. 2017; 141(6): 776–86. PubMed Abstract \| Publisher Full Text\n\nGargis AS, Kalman L, Berry MW, et al.: Assuring the quality of next-generation sequencing in clinical laboratory practice. Nat Biotechnol. 2012; 30(11): 1033–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVar SK, Hadi R, Khardori NM: Evaluation of regional antibiograms to monitor antimicrobial resistance in hampton roads, Virginia. Ann Clin Microbiol Antimicrob. 2015; 14(1): 22. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrown D, Canton R, Dubreuil L, et al.: Widespread implementation of EUCAST breakpoints for antibacterial susceptibility testing in Europe. Euro Surveill. 2015; 20(2): pii: 21008. PubMed Abstract \| Publisher Full Text\n\nKahlmeter G: Defining antibiotic resistance-towards international harmonization. Ups J Med Sci. 2014; 119(2): 78–86. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLeclercq R, Cantón R, Brown DF, et al.: EUCAST expert rules in antimicrobial susceptibility testing. Clin Microbiol Infect. 2013; 19(2): 141–60. PubMed Abstract \| Publisher Full Text\n\nAanensen DM, Feil EJ, Holden MT, et al.: Whole-Genome Sequencing for Routine Pathogen Surveillance in Public Health: a Population Snapshot of Invasive Staphylococcus aureus in Europe. MBio. 2016; 7(3): pii: e00444–16. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nToutain PL, Bousquet-Mélou A, Damborg P, et al.: En Route towards European Clinical Breakpoints for Veterinary Antimicrobial Susceptibility Testing: A Position Paper Explaining the VetCAST Approach. Front Microbiol. 2017; 8: 2344. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCaucci S, Karkman A, Cacace D, et al.: Seasonality of antibiotic prescriptions for outpatients and resistance genes in sewers and wastewater treatment plant outflow. FEMS Microbiol Ecol. 2016; 92(5): fiw060. PubMed Abstract \| Publisher Full Text\n\nBujang MA, Adnan TH: Requirements for Minimum Sample Size for Sensitivity and Specificity Analysis. J Clin Diagn Res. 2016; 10(10): YE01–YE06, [cited 2018 Jan 12]. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRao Y, Lee Y, Hsu JC: Determination of sample size for validation study in pharmacogenomics. preparation. 2009. Reference Source\n\nBlair JM, Webber MA, Baylay AJ, et al.: Molecular mechanisms of antibiotic resistance. Nat Rev Microbiol. 2015; 13(1): 42–51. PubMed Abstract \| Publisher Full Text\n\nLiu YY, Wang Y, Walsh TR, et al.: Emergence of plasmid-mediated colistin resistance mechanism MCR-1 in animals and human beings in China: a microbiological and molecular biological study. Lancet Infect Dis. 2016; 16(2): 161–8. PubMed Abstract\n\nHasman H, Hammerum AM, Hansen F, et al.: Detection of mcr-1 encoding plasmid-mediated colistin-resistant Escherichia coli isolates from human bloodstream infection and imported chicken meat, Denmark 2015. Euro Surveill. 2015; 20(49), [cited 2017 Apr 5]. PubMed Abstract \| Publisher Full Text\n\nKluytmans-van den Bergh MF, Huizinga P, Bonten MJ, et al.: Presence of mcr-1-positive Enterobacteriaceae in retail chicken meat but not in humans in the Netherlands since 2009. Euro Surveill. 2016; 21(9): 30149. PubMed Abstract \| Publisher Full Text\n\nPhelan J, O’Sullivan DM, Machado D, et al.: The variability and reproducibility of whole genome sequencing technology for detecting resistance to anti-tuberculous drugs. Genome Med. 2016; 8(1): 132. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nXavier BB, Das AJ, Cochrane G, et al.: Consolidating and Exploring Antibiotic Resistance Gene Data Resources. Kraft CS, editor. J Clin Microbiol. 2016; 54(4): 851–9. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "33450", "date": "11 May 2018", "name": "Enrico Lavezzo", "expertise": [ "Reviewer Expertise Microbiology", "bioinformatics", "NGS data analysis" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this opinion article, Alexandre Angers-Loustau and colleagues address a hot topic in Diagnostic Microbiology regarding the implementation of next generation sequencing platforms for the identification of antimicrobial resistances. In particular, they focus on a specific step of the long pipeline which goes from sample collection to data interpretation and management, namely the bioinformatics analysis of sequencing data.\n\nIt's been more than 10 years since the introduction of the first NGS instrument, but an adequate standardization and harmonization level of the many different platforms is not yet accomplished.\nThe article is very well written; it is concise but covers all the important issues that need to be considered when willing to implement such technologies for the purpose of AMR detection.\n\nI only have some minor points for the authors:\nIn section 2.1, page 5, regarding the potential presence of sensitive information (e.g. reads of human origin) when including real experiment datasets in the benchmark. I don't see the problem here. How the availability of anonymous DNA sequence information could compromise the patient's privacy? Of course, patient's identity must be separated from sample's ID, but the de-identification process is a standard procedure and others methods have also been developed for this purpose (see for example this paper from Qamar et al. (2014)1). Please discuss about this.\n\nSince this is an opinion article, I would like the authors to 'expose' themselves a little bit more on some points. The challenges are presented by explaining the state-of-the-art, but it is not easy to me to understand authors' thinking about the possible solutions. Obviously they can't have answers for all the challenges, but they could express some personal positions (for example, what about \"Who should generate the benchmark resource\"? To me, there should be a centralized action aimed, as far as possible, at the definition of a global standard (similar for example to the WHO HPV LabNet Proficiency Studies for Evaluating HPV DNA Typing Methods). What do the authors think about this?\n\nFinally, no reference is made to microorganisms other than bacteria, but AMR may be related also to viruses and eukaryotic microorganisms. Are there any difference or peculiarity regarding the bioinformatics analysis of sequencing data coming from them?\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes", "responses": [] }, { "id": "35036", "date": "29 Jun 2018", "name": "Jason C. Kwong", "expertise": [ "Reviewer Expertise Clinical infectious diseases", "microbial pathogen genomics", "antimicrobial agents and resistance", "clinical metagenomics" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe opinion article by Alexandre Angers-Loustau and colleagues addresses an increasingly encountered question by clinical and public health microbiology laboratories - what are the considerations in developing a benchmark for validating methods to identify antimicrobial resistance determinants from high-throughput sequencing data? This is an important step in using sequencing data for this purpose, and this article reports back on the discussions by a panel of experts in the field of antimicrobial resistance surveillance.\nIn general, the article is well-suited to publication in F1000Research, offering insight into an issue of global importance, and I thank the authors for making the discussions from the October 2017 meeting organised by the Joint Research Centre public. It is well written and explores each of the discussion questions with supporting literature. My suggestions for improvement are minor and addressing them is left up to the authors' discretion.\nMy overall comment is that while many of the relevant considerations are raised, as a clinician and researcher, I am left with the burden of all these concerns, but no clear direction in how to address the challenges (other than a few comments in the conclusion). I appreciate the purpose of the paper was to summarise the challenges identified at the meeting, but I think this (opinion) article could be enhanced and have a greater impact if more expert opinions and ideas on how to proceed were included.\nMinor comments: Introduction (page 4) It would be nice to know a little more about the panel of experts - for example, what mix of background or fields did the experts come from e.g. clinical or public health microbiology, bioinformaticians, epidemiologists, veterinary microbiology etc.?\nSection 2.1: real experiments vs simulated read sets (page 5) This section highlights some important concerns (e.g. filtering host DNA or patented genetically modified food DNA), but these are more relevant to identification of antimicrobial resistance from metagenomic sequencing data - this is alluded to in the text, but not explicitly stated. Given the other technical challenges with this, I wonder if the discussion should focus more on the \"real experimental\" data from cultured bacterial organisms – the more likely initial validation before validation proceeds to metagenomic datasets.\nSection 2.2: how should the set of resistance genes/mechanisms be agreed upon (page 6) The considerations here are well explored, but I wonder if there could be some mention of variation in benchmark datasets depending on the purpose of resistance gene/mechanism detection. For example, detection of AMR genes from environmental or livestock samples may have a different focus, and thus different benchmark targets to detection of AMR genes in organisms from human samples. Even among human samples, there may be differences in the mechanisms prioritised for detection between testing for clinical treatment and testing for epidemiological surveillance (mcr-1 is a good example).\nSection 2.3: how can the \"true\" value of the samples be guaranteed (page 7) Great section and discussion. Along the lines of the discussion, I am interested to hear the authors' thoughts on whether it is sufficient to validate NGS for predicting AMR against standard susceptibility testing (e.g. MIC), or whether clinical studies evaluating treatment outcomes based on NGS detection are required (i.e. treating based on genotype vs phenotype). My only comment for this section is that perhaps some of the commentary is more directed at validation of NGS for \"identifying antimicrobial resistance\" rather than \"identifying antimicrobial resistance determinants\" as stated in the title - two subtly different targets. Perhaps some of the discussion could be clarified here.\n“HTF5 format” (page 4) should actually be “HDF5 format”.\n\nMIC should be spelt out at the first usage - minimum inhibitory concentration.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-459
https://f1000research.com/articles/7-474/v1	18 Apr 18	{ "type": "Research Article", "title": "Predicting lithium treatment response in bipolar patients using gender-specific gene expression biomarkers and machine learning", "authors": [ "Andy R. Eugene", "Jolanta Masiak", "Beata Eugene", "Jolanta Masiak", "Beata Eugene" ], "abstract": "Background: We sought to test the hypothesis that transcriptiome-level genes signatures are differentially expressed between male and female bipolar patients, prior to lithium treatment, in a patient cohort who later were clinically classified as lithium treatment responders. Methods: Gene expression study data was obtained from the Lithium Treatment-Moderate dose Use Study data accessed from the National Center for Biotechnology Information’s Gene Expression Omnibus via accession number GSE4548. Differential gene expression analysis was conducted using the Linear Models for Microarray and RNA-Seq (limma) package and the Random Forests machine learning algorithm in R. Results: In pre-treatment lithium responders, the following genes were found having a greater than 0.5 fold-change, and differentially expressed indicating a male bias: RBPMS2, SIDT2, CDH23, LILRA5, and KIR2DS5; while the female-biased genes were: HLA-H, RPS23, FHL3, RPL10A, NBPF14, PSTPIP2, FAM117B, CHST7, and ABRACL. Conclusions: Using machine learning, we developed a pre-treatment gender- and gene-expression-based predictive model selective for lithium responders with an ROC AUC of 0.92 for men and an ROC AUC of 1 for women.", "keywords": [ "lithium", "treatment response", "gene expression", "machine learning", "microarray", "transcriptome", "precision medicine", "pharmacogenomics" ], "content": "Introduction\n\nLithium is the most well-established mood-stabilizer in the practice of psychiatry (Jermain et al., 1991; Landersdorfer et al., 2017). A recent propensity-score adjusted and matched longitudinal cohort-study evaluating the effectiveness of the newer mood stabilizers: olanzapine (n=1477), quetiapine (n=1376), and valproate (n=1670), in comparison to lithium (n=2148), found that patients treated with lithium experienced reduced rates of both unintentional injury and self-harm (Hayes et al., 2016). However, due to lithium's narrow window, 0.5-1.2 mEq/mL, of maximal effectiveness and safety (i.e. therapeutic index), Therapeutic Drug Monitoring is the standard-of-care to ensure patient safety in medical practice (Hiemke et al., 2011). Further, divergent clinical response rates have been reported among male and female patients diagnosed with bipolar disorder and treated with lithium (Viguera et al., 2000).\n\nIn a 1986, Zetin and colleagues published the results of a study that evaluated four methods for predicting lithium daily dosages, and the final equation resulted in a 147.8mg/day increased dosage-adjustment for male patients (Zetin et al., 1986). Similarly, a later study by Lobeck and colleagues corroborated the 147.8 mg/day male increase dose requirement for the lithium maintenance dose in bipolar patients (Lobeck et al., 1987). However, neither do the current dosing guidelines recommend a gender-based dose adjustment via clinical pharmacometrics, to avoid toxicity, nor are gender-specific gene expression screening panels available to predict lithium efficacy currently available and implemented.\n\nA recent large-scale meta-analysis of human body-tissue gene expression reported that the body organ with the most abundant gender-biased gene expression is the anterior cingulate cortex within the frontal cortex of the brain (Mayne et al., 2016). Thus, these findings suggest that therapeutic drug response may be influenced not only via drug absorption, distribution, metabolism, and elimination, but also within the underlying gene signatures across the human transcriptome and mechanisms of gene-gene interactions that regulate physiology. Beech and colleagues conducted a study to identify gene expression differences from the peripheral blood in patients classified as lithium responders and non-responders (Beech et al., 2014). However, the study reported that no significant gender-biased gene expression differences were found (p-value=0.941) in patients who were randomized to optimal therapy (control), defined as one FDA-approved mood stabilizer, versus patients treated with lithium plus optimal therapy (Beech et al., 2014). Despite these initially reported findings, a recent study by Labonté and colleagues, which used RNA-Seq to evaluate the transcriptome in patients diagnosed with major depressive disorder (MDD), concluded that gender dimorphism exists at the transcriptome-level in MDD patients and that gender-specific treatments should be investigated (Labonté et al., 2017).\n\nTherefore, there is an urgent clinical need to improve behavioral healthcare by understanding gene expression variability that may lead to personalizing medicine in patients with mania. These findings may improve prediction of clinical drug response of lithium prior to initiating pharmacotherapy in patients with bipolar or schizoaffective disorders, who cannot risk drug inefficacy for obvious safety reasons. Therefore, the overall aim for our study is to define gender-specific transcriptional-level regulators of lithium treatment response that may influence treatment of bipolar or schizoaffective disorders. We will test the hypothesis that biologically plausible gene expression differences exist, prior to lithium treatment, in patients diagnosed with bipolar disorder in the following three patient subgroups: (1) male and female patients who were later clinically classified as lithium treatment responders; (2) male-responders versus male-non-responders; (3) female-responders versus female-non-responders.\n\n\nMethods\n\nDNA microarray data analyzed in this study are originally referenced from the Lithium Treatment-Moderate dose Use Study placed in the National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO) via accession number GSE45484 with the Illumina HumanHT12 V4.0 expression Beadchip GPL10558 platform file to associate gene names and descriptions. The original multisite clinical study recruited patients from Case Western Reserve University, Massachusetts General Hospital, Stanford University, Yale University, and the Universities of: Pittsburgh, Texas Health Science Center at San Antonio, and Pennsylvania (Beech et al., 2014). From the original 120 peripheral blood samples used to generate probe and gene expression profiles, from patients diagnosed with bipolar disorder, the clinical phenotype of being either a treatment- responder or non-responder was assessed using the Clinical Global Impression Scale for Bipolar Disorder-Severity (CGI-BP-S) (Spearing et al., 1997).\n\nTo assess for gender-specific differential gene signatures, in our first analysis we grouped patients based on gender alone and not on any other variables (i.e. optimal treatment versus lithium, or responder versus non-responder status). From the results of the gender-specific transcriptome signatures in first analysis, we set the top two-hundred and fifty genes as controls that would be excluded from all results that would be reported in subsequent gene expression analyses to identify genes with lithium-specific gene expression differences between genders associated with response to lithium treatment. In our second analysis, we only selected patients who were classified as lithium treatment-responders, at baseline, and the results from the gene expression differences are reported excluding the sex-specific control genes identified in the first experiment. In our third and fourth analyses, we compared: male-responders vs. male non-responders, and female-responders vs. female non-responders, respectively.\n\nDifferential gene expression analysis of the microarray data was conducted using the Empirical Bayes method implemented within the limma package (version 3.34.5) and utilizes the Biobase package (version 2.38.0) which both run within the R for Statistical Programming environment (version 3.4.3; R Foundation for Statistical Computing, Vienna, Austria) (Ritchie et al., 2015; Team, 2013). Due to multiple testing of the peripheral blood transcriptome, the False-Discovery Rate was adjusted using the Benjamini-Hochberg method. The Decision Tree and Random Forests machine learning algorithms were used to assess gender using transcriptional signatures and for predictive modeling using the discovered microarray genes to select for the gender-specific lithium responders. A p-value of less 0.05 was considered to be statistically significant and a differential gene expression threshold of 0.5 was used and reported during the machine learning process. Further methods detailing the Random Forests decision processes for male- and female-responders are located in Supplementary File 1.\n\n\nResults\n\nTable 1 provides the patient age and sample sizes used during subgroup analyses. In our first analysis, which aimed to group patients based on gender alone and not based on clinical variables detailed in the original study, data-driven gene analytics identified four female-labeled patient samples with gene expression levels similar to that found in male patients for the following Y-chromosome genes: RPS4Y1, EIF1AY, KDM5D, RPS4Y2; and the XIST gene located on the X-chromosome. Therefore, all subsequent hypothesis-testing were analyzed with the updated male-gender classification for the following NCBI GEO patient samples: GSM1105526 (baseline lithium-non-responder), GSM1105528 (1-month lithium-non-responder), GSM1105546 (baseline lithium-non-responder), and GSM1105548 (1-month lithium-non-responder). Figure 1 illustrates the gene expression findings resulting in re-assignment for the aforementioned patient samples from females to males using a decision-tree approach that evaluated if the RPS4Y1 gene had an expression level of greater than or equal to 9.6 resulting in: yes=male (31%) and no=female (69%). After proceeding with the machine learning analysis of both the ‘training’ and ‘validation’ datasets, the final ‘test’ dataset resulted in the following diagnostic test evaluation parameters: Sensitivity=100% (95% C.I. 66.37%-100.00%), Specificity=100% (95% C.I. 78.20%-100.00%), and an area under the receiver operator characteristic (ROC) curve of 1. Figure 2 illustrates the variable importance plots used in the machine learning process.\n\nNote: United States Food and Drug Administration approved Mood Stabilizers.\n\nMales (n=41) and Females (n=39).\n\nVariable importance ratings of genes selective (above) male lithium responders versus the entire population of treated and untreated patient men and women; and (below) female lithium responders versus the entire population of treated and untreated men and women.\n\nTable 2 provides the results for the gender-specific differentially expressed genes from the entire study population using a fold-change (FC) threshold of 0.5. A total of five genes met the a priori FC requirements and were found to be RPS4Y1, EIF1AY, KDM5D, RPS4Y2, and EIF1AY. These five down-regulated male-biased genes were all found on the Y-chromosome. Contrastingly, a total of 10 upregulated female-biased genes were found to be: XIST, S100P, IFIT3, TNFAIP6, IFITM3, IFIT2, CHURC1, ANXA3, ADM, and PROK2. The RPS4Y1 gene in males (FC= -4.9807, p=7.36E-47) and the XIST gene (FC=1.7615, p=2.98E-36), found on the X-chromosome, in females resulted in the greatest expression changes between genders. The male-favored genes resulted in a larger expression change than compared to the females.\n\nTable 3 provides the results for the differentially expressed genes that were found between male and female responders prior to initiation of lithium and optimal therapy, meeting the FC criteria of at least 0.5. In male lithium responders, we found 5 differentially expressed and down-regulated genes while the RNA binding protein with multiple splicing 2 (RBPMS2) gene ranked with the greatest FC of -1.351 (unadjusted p=0.00111). Whereas, 9 up-regulated genes were associated with female lithium responders, with greatest expression change being the major histocompatibility complex class-1-H (HLA-H) at 1.602 (unadjusted p-value=0.00099). The neuroblastoma breakpoint family member-14 (NBPF14) gene met the Benjamani-Hochberg adjusted p-value criteria and resulted with an expression change of 0.586 (adjusted p=0.0462). Figure 3 illustrates the heat-map and dendrogram overview of the two-way unsupervised hierarchical cluster analysis of the reported differentially expressed genes among male and female responders to lithium therapy at baseline that correspond to values reported in Table 3.\n\nNotes: *The NBPF14 gene reached the Benjamani-Hochberg adjusted p-value.\n\nUsing the baseline blood sample microarray data, the predictive modeling results for identifying lithium-responders from the complete study population of male and female controls and treatment samples, resulted in a validation/test sample cohort for males of: Sensitivity=95.83% (95% C.I. 78.88%-99.89%), Specificity=not calculated due sample size of test dataset, and an ROC curve AUC = 0.92 using the RBPMS2 and LILRA5 genes. Likewise, in the test dataset for females: Sensitivity=91.67% (95% C.I. 61.52%-99.79%), Specificity= not calculated due sample size of test dataset, and an ROC curve AUC = 1 with the ABRACL and NBPF14 genes. Therefore, we developed a 2-gene predictive model for men and likewise for women predicting lithium response in bipolar patients from a general population of bipolar patients using transcriptional signatures at baseline.\n\nTable 4 provides the list of 10 differentially expressed genes found in male lithium responders (5-genes) and male lithium-non-responders (5-genes). The RNA binding protein with multiple splicing 2 (RBPMS2) gene (FC= -1.326, unadjusted p=0.001358) in male lithium responders and the Ribosomal protein S23 (RPS23) gene (FC=1.521, unadjusted p=0.013306) were found to result in the largest expression change differences between subgroups. However, in female responders and female non-responders, the Family with Sequence Similarity 117 Member B (FAM117B) gene (FC=0.5257, unadjusted p=0.0048554) and the Golgin B1 (GOLGB1) gene (FC= -0.6536, unadjusted p=0.0003716) were differentially expressed, respectively and shown in Table 5.\n\n\nDiscussion\n\nThe purpose of this investigation was to define gender-specific transcriptome-level regulators of lithium treatment response prior to the initiation of lithium treatment. We first established the gender-relevant transcriptional control genes across all study-participant blood samples and specifically to male- and female-responders using a differential gene expression threshold of 0.5. We found this to be adequate and corroborated with similar studies that used a similar threshold for establishing gene transcription signatures (Jansen et al., 2014; Mayne et al., 2016). However, when comparing the male-responders to male non-responders, as well as, the female responders to female non-responders, we set an inclusion fold-change threshold to 0.3. This approach is not unusual, since it is already established that both large and subtle expression changes produce to significant biological and physiological processes (Wurmbach et al., 2002). Our analysis is both hypothesis-generating, and establishes a computational methodology that provides insight to the importance of subgroup analysis in genomic medicine, irrespective of patient sample-sizes. The end-goal of such analyses serves as a testing methodology for establishing gene screening panels to improve personalized medicine in vulnerable and high-risk patient populations. In these patient populations, it is often not feasible to wait for weeks to determine whether a prescribed medication will work and in some cases manic patients are neither able to fully comprehend and be objectively assessed using the CGI-BP-S (Spearing et al., 1997).\n\nWhen reviewing the heat-map and dendrogram hierarchical cluster analysis patterns, specifically the numerous non-responders clinically-labeled and illustrated in Figure 4, they suggest that the underlying etiology resulting in clinical symptoms (e.g. mania) that led to the diagnosis of bipolar disorder may need re-classification. Further, the subsequent treatments may need to be tailored in data-driven computational psychiatry approaches. In Figure 4, for the females, the samples in the center cluster illustrates that a group of patients are clear non-responders while the patients clustered in the far-right are partial-responders, from a molecular perspective. The natural questions that arise are: (1) How to best convert the non- and partial-responders to treatment-responders? (2) Is a behavioral intervention, in this select group of patients, for whom lithium is not effective, the best answer because the symptoms maybe of a different etiology? If indeed the symptoms are of a different etiology (e.g. inflammatory), from the lithium treatment-responders, then other diagnostic (e.g. electrophysiological neuroimaging) tools may be warranted and corresponding most efficacious treatments sought.\n\nHeat-map and dendrogram overview of the two-way unsupervised hierarchical cluster analysis of differentially expressed genes prior to lithium treatment in (above) male lithium responders (n=3, RESP_Male) and male lithium non-responders (n=7, NR_Male); and (below) female responders (n=6, RESP_Fem) and female non-responders (n=14, NR_Fem).\n\nWhen differentiating between male and female patients, we found that the Ribosomal Protein S4, Y-linked 1 (RPS4Y1, adjusted p-value=7.36E-47) male-linked gene and the X Inactive Specific Transcript (XIST, adjusted p-value=2.98E-36) female-linked gene were the most differentially expressed among genders, which is consistent with previously published studies (Guillén et al., 2014; Jansen et al., 2014; Mayne et al., 2016). The genes that are specific to male lithium responders, relative to female lithium responders, are RBPMS2, SIDT2, CDH23, LILRA5, and KIR2DS5. Using the same methodology, genes identifying female lithium responders, relative to male lithium responders, are HLA-H, RPS23, FHL3, RPL10A, NBPF14, PSTPIP2, FAM117B, CHST7, and ABRACL. The Neuroblastoma Breakpoint Family Member 14 (NBPF14, adjusted p-value=0.0462, Fold-change=0.586) achieved the Benjamani-Hochberg adjusted p-value of 0.0462, and has been reported to be associated with cortical neurogenesis (Suzuki et al., 2017).\n\nComputational psychiatry, as advocated by the National Institute of Mental Health’s Research Domain Criteria (RDoC), may need data to drive the classification, diagnosis, and treatment response status, especially in patients with developmental delay, language difficulty, and condition of a potentially different etiology than traditionally taught (Clark et al., 2017; Eugene & Masiak, 2016). Ideally, in such cases, alternative FDA-approved mood stabilizers may be initially selected prior to any pharmacological intervention by simply using a blood test. Perhaps, a gene expression screening panel at baseline, prior to the initiation of lithium and/or other FDA-approved mood stabilizer, may be better in high-risk patient populations.\n\nThese findings suggest that when implementing genomic medicine, clinical research teams should move beyond the single-gene approach when screening for treatment responders or non-responders. This approach is currently the standard when screening for patient toxicity at standard doses in poor or ultra-rapid metabolizers; however, as more transcriptional factors are discovered that regulate the cytochrome (CYP) P-450 system of genes, multi-gene pharmacokinetic panels are inevitable and may be included in future Clinical Pharmacogenetics Implementation Consortium (CPIC) guidelines. Next, medical management of patients with mania and psychosis either with pharmacotherapy and/or behavioral intervention should be tailored to biological gender due to known neuronal circuitry differences in age-matched patients with psychosis (Eugene et al., 2015). Further, as a result of lithium not being hepatically metabolized, but rather transported and renally excreted, as well as, the known myriad drug-drug interactions, patient dose selection may benefit from clinical pharmacometrics modeling by board-certified/eligible pharmacologists (Perera et al., 2014; Zetin et al., 1986). This approach may be implemented to ensure drug pharmacokinetic safety.\n\nThe limitations of our analysis and in most genetic studies are understandably due to multiple-comparison p-value adjustments and patient sample size (Dudoit et al., 2003). The fundamental aims of our research questions were designed to answer biological questions of gender and clinical response to lithium and not meant to be driven exclusively by multiple comparisons adjusted p-values. This approach has led to various successes in genomic medicine, specifically, in genome-wide association studies; however, understandably, the limitations are thoroughly acknowledged. In reference to patient sample sizes, 9 out of the 28 patients who received lithium and optimal therapy were classified as lithium treatment responders. Further, 30% of men and 33% of women, who were treated with lithium, were found to be responders at the respective gender categories (Beech et al., 2014). However, the strengths of our findings are in the gender-gene screening ability for lithium treatment-responders in the general population of 60 patients at baseline, minus the tested responder group. Opportunities exist for prospective clinical trials and application of the methods outlined in this text for other therapeutic agents across several medical specialties.\n\n\nConclusion\n\nWe explored the Lithium Treatment-Moderate dose Use Study clinical trial gene expression data with the aim of identifying gender-specific transcriptome-level regulators of lithium treatment response. Using machine learning, we successfully developed a pre-treatment gender- and gene-expression-specific predictive model selective for lithium responders with an ROC AUC of 0.92 for men and an ROC AUC of 1 for women. Further, by using well-established Bayesian statistical methods, to identify differentially expressed genes and then machine learning, we discovered 5-genes selective for men and 9-genes that are selective for women that will inform the physicians and clinical staff of whether the patient will respond to lithium prior to being prescribed the drug. With the small number of patient responders from the clinical trial, our results should be confirmed. Lastly, in an overall context, our results suggest that the methodology used in this analysis may be extended to other therapeutic drug classes and provides insight to the gender-based gene transcriptome differences influencing lithium pharmacodynamics.\n\n\nData availability\n\nData used in this study are available from https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE45484", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nThe authors gratefully acknowledge the patients in the original clinical trial, the medical staff, and the NCBI GEO database accession GSE4548.\n\n\nSupplementary material\n\nSupplementary File 1: Supplementary methods.\n\nClick here to access the data.\n\n\nReferences\n\nBeech RD, Leffert JJ, Lin A, et al.: Gene-expression differences in peripheral blood between lithium responders and non-responders in the Lithium Treatment-Moderate dose Use Study (LiTMUS). Pharmacogenomics J. 2014; 14(2): 182–91. PubMed Abstract \| Publisher Full Text\n\nClark LA, Cuthbert B, Lewis-Fernández R, et al.: Three Approaches to Understanding and Classifying Mental Disorder: ICD-11, DSM-5, and the National Institute of Mental Health’s Research Domain Criteria (RDoC). Psychol Sci Public Interest. 2017; 18(2): 72–145. PubMed Abstract \| Publisher Full Text\n\nDudoit S, Shaffer JP, Boldrick JC: Multiple Hypothesis Testing in Microarray Experiments. Statist Sci. 2003; 18(1): 71–103. Publisher Full Text\n\nEugene AR, Masiak J: Identifying Treatment Response of Sertraline in a Teenager with Selective Mutism Using Electrophysiological Neuroimaging. Int J Clin Pharmacol Toxicol. 2016; 5(4): 216–19. PubMed Abstract \| Free Full Text\n\nEugene AR, Masiak J, Kapica J, et al.: Electrophysiological Neuroimaging Using sLORETA Comparing 22 Age Matched Male and Female Schizophrenia Patients. Hosp Chron. 2015; 10(2): 91–98. PubMed Abstract \| Free Full Text\n\nGuillén IA, Fernández JR, Palenzuela DO, et al.: Analysis of Gene Expression Profile for Gender in Human Blood Samples. International Journal of Innovation and Applied Studies. 2014; 7(1): 329–42. Reference Source\n\nHayes JF, Pitman A, Marston L, et al.: Self-harm, Unintentional Injury, and Suicide in Bipolar Disorder During Maintenance Mood Stabilizer Treatment: A UK Population-Based Electronic Health Records Study. JAMA Psychiatry. 2016; 73(6): 630–7. PubMed Abstract \| Publisher Full Text\n\nHiemke C, Baumann P, Bergemann N, et al.: AGNP Consensus Guidelines for Therapeutic Drug Monitoring in Psychiatry: Update 2011. Pharmacopsychiatry. 2011; 44(6): 195–235. PubMed Abstract \| Publisher Full Text\n\nJansen R, Batista S, Brooks AI, et al.: Sex differences in the human peripheral blood transcriptome. BMC Genomics. 2014; 15(1): 33. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJermain DM, Crismon ML, Martin ES 3rd: Population pharmacokinetics of lithium. Clin Pharm. 1991; 10(5): 376–81. PubMed Abstract\n\nLabonté B, Engmann O, Purushothaman I, et al.: Sex-specific transcriptional signatures in human depression. Nat Med. 2017; 23(9): 1102–11. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLandersdorfer CB, Findling RL, Frazier JA, et al.: Lithium in Paediatric Patients with Bipolar Disorder: Implications for Selection of Dosage Regimens via Population Pharmacokinetics/Pharmacodynamics. Clin Pharmacokinet. 2017; 56(1): 77–90. PubMed Abstract \| Publisher Full Text\n\nLobeck F, Nelson MV, Evans RL, et al.: Evaluation of Four Methods for Predicting Lithium Dosage. Clin Pharm. 1987; 6(3): 230–33. PubMed Abstract\n\nMayne BT, Bianco-Miotto T, Buckberry S, et al.: Large Scale Gene Expression Meta-Analysis Reveals Tissue-Specific, Sex-Biased Gene Expression in Humans. Front Genet. 2016; 7: 183. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPerera V, Bies RR, Mo G, et al.: Optimal sampling of antipsychotic medicines: a pharmacometric approach for clinical practice. Br J Clin Pharmacol. 2014; 78(4): 800–814. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRitchie ME, Phipson B, Wu D, et al.: Limma Powers Differential Expression Analyses for RNA-Sequencing and Microarray Studies. Nucleic Acids Res. 2015; 43(7): e47. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSpearing MK, Post RM, Leverich GS, et al.: Modification of the Clinical Global Impressions (CGI) Scale for Use in Bipolar Illness (BP): The CGI-BP. Psychiatry Res. 1997; 73(3): 159–71. PubMed Abstract \| Publisher Full Text\n\nSuzuki IK, Gacquer D, Van Heurck R, et al.: Hominin-Specific NOTCH2 Paralogs Expand Human Cortical Neurogenesis through Regulation of Delta/Notch Interactions. bioRxiv. Cold Spring Harbor Laboratory, 2017; 221358. Publisher Full Text\n\nTeam R: R Development Core Team. R: A Language and Environment for Statistical Computing. 2013. Reference Source\n\nViguera AC, Tondo L, Baldessarini RJ: Sex differences in response to lithium treatment. Am J Psychiatry. 2000; 157(9): 1509–11. PubMed Abstract \| Publisher Full Text\n\nWurmbach E, González-Maeso J, Yuen T, et al.: Validated genomic approach to study differentially expressed genes in complex tissues. Neurochem Res. 2002; 27(10): 1027–33. PubMed Abstract \| Publisher Full Text\n\nZetin M, Garber D, De Antonio M, et al.: Prediction of lithium dose: a mathematical alternative to the test-dose method. J Clin Psychiatry. 1986; 47(4): 175–78. PubMed Abstract" }	[ { "id": "33941", "date": "21 May 2018", "name": "Ming-Fen Ho", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors demonstrated that sex-differences gene expression might contribute to lithium treatment response using microarray expression data.\nMajor comments:\nA samples size of 60 might be too small to determine the sex effects. Can the sample size n=60 provide adequate power for data interpretation, especially separated men and women for study sex-effect on gene expression?\n\nThe authors stated that their predictive model for lithium responders with an ROC AUC 0.92 for men, and 1 for women. If the prediction accuracy is so significant, what are the potential biological mechanisms beyond these genes? More discussion regarding the biology of those genes should be included in the paper. Once again, if the prediction accuracy is so significant, it is needed a replication study using different data sets? In summary, the authors claimed the prediction model with very high accuracy; it should be included either functional validation of those genes or a replication study population.\n\nSpecific comments:\nMethods - study design, it might be better to use a flow chart to demonstrate the study design.\n\nMethods - study design, please clarify the rationale of filtering out “250” genes.\n\nTable 1 shows total study population n=60, but figure 1 legend shows male: n=41, female: n=39?\n\nFigure 2: please elaborate the data presented in Figure 2. The key results for each of the four panels should be summarized in Results.\n\nTable 2 and Table 4, the log FC threshold of 0.5 or 0.3 might be too low. The changes in gene expression are very subtle in Table 4.\n\nTable 2, are there any gene up-regulated in males? /downregulated in females?\n\nLimitations of the study should be addressed in Discussion.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "3661", "date": "29 May 2018", "name": "Andy Eugene", "role": "Author Response", "response": "Major Comment Responses:Response 1: This point is well noted; however, it is important to realize that our gender-effects of gene expression is consistent with other studies noted within the paper and shown below:Jansen, Rick, et al. \"Sex differences in the human peripheral blood transcriptome.\" BMC genomics 15.1 (2014): 33.Mayne, Benjamin T., et al. \"Large scale gene expression meta-analysis reveals tissue-specific, sex-biased gene expression in humans.\" Frontiers in genetics 7 (2016): 183.Further, our gender-specific results met the Benjamini-Hochberg multiple comparisons criteria adjustment due to multiple comparisons.Comment Response 2: We welcome and thank the reviewer’s comments on the biological mechanisms beyond these genes. Clearly, it is well noted and cited in the paper that in clinical practice there is a wide inter-individual variability in the treatment and response to treatments of biplolar disorder. Moreover, these patients were not treated with lithium monotherapy, alone, and therefore further insight into the biological mechanisms were left out due to these patients were treated with an “Optimal Therapy” that includes a variety of other FDA-approved mood stabilizers.In reference to the comment regarding the prediction accuracy, we agree that the study may warrant functional validation in a laboratory; however, it is beyond the scope of our computational psychiatry study and we will leave the functional genomics characterization of the genes to investigators seeking to pursue the findings from our results." }, { "c_id": "3665", "date": "29 May 2018", "name": "Andy Eugene", "role": "Author Response", "response": "Specific Comment Responses: We thank you for your specific comments and have addressed several of the pertinent points in your review. For all differentially expressed results reported throughout tables within the manuscript, we changed the wording from genes up-regulated or down-regulated in males or females to a clearer description statement that of genes-associated with males or females. However, we thought not necessary to include an extra figure, but rather encourage the reader to (1) review the study design section within the methods to better understand the computational approach used in our analysis and (2) read the systematic tabular reporting of the results in the manuscript text as well to understand that study approach. For the caption in Figure 1, we thank you for the comment and have updated the sample sizes for males and female patients. The updated Figure 1 text reads: Males (n=20; with 40 pre- and post-treatment samples) and Females (n=40; with 80 pre- and post-treatment samples). The comments regarding: (1) the fold-change of 0.5 and 0.3 being subtle and (2) the study limitations, are already specifically addressed within the original version of the manuscript. Again, it is well established and referenced within the text that small changes in gene expression have already been reported to result in major functional outcomes in human physiology. We will update the variable importance illustration shown in Figure 2 and that will be added to the updated version of the manuscript." } ] } ]	1	https://f1000research.com/articles/7-474
https://f1000research.com/articles/7-1905/v1	07 Dec 18	{ "type": "Research Article", "title": "Pre-treatment loss to follow-up among patients with rifampicin-resistant tuberculosis in Baluchistan, Pakistan, 2012-17: a retrospective cohort study", "authors": [ "Shoaib Aziz Kurd", "Ahmed Wali", "Razia Fatima", "Aashifa Yaqoob", "Dawood Khan", "Sultan Lehri", "Ahmed Wali", "Razia Fatima", "Aashifa Yaqoob", "Dawood Khan", "Sultan Lehri" ], "abstract": "Background: Patients with rifampicin-resistant TB (RR-TB) pretreatment loss to follow-up continue to be a global health challenge. Although the accuracy of diagnosis significantly increased with the implementation of Xpert MTB/RIF assay, which is a rapid molecular based test and more sensitive than conventional microscopy which detects MTB even present in small limit of 136 MTB/ml of sputum, but still data suggest a wide treatment initiation gap among diagnosed. This study was done to assess the proportion of patients with RR-TB pretreatment lost to follow-up and the socio-demographic factors associated with this in Balochistan, Pakistan. Methods: This was a retrospective cohort study based on review of the routinely managed program records. The data included all patients with RR-TB detected at Fatima Jinnah Chest & General Hospital Quetta and District Head Quarter Hospital Loralai, Xpert sites and enrolled at programmatic management of drug resistant TB (PMDT) sites during 2012-2017. Data collected was double-entered, validated and analyzed using EpiData. Results: Of the 396 patients with RR-TB detected during 2012-17, 78 (19.8%) underwent pre-treatment lost to follow-up. The mean age of those detected with RR-TB was 37 years (SD ±16.98); 189 (48%) were of age group 15-34, while 60% were female. Among 84 individuals referred out to other facilities, only 6 started treatment. Almost half of the ‘pretreatment lost to follow-up’ patients were from age group 15-34, while 43 were from within the Quetta and Loralai districts. Conclusions: The high proportion of patients with RR-TB that were pre-treatment lost to follow-up in Balochistan needs immediate strategies to establish linkages between Xpert and PMDT sites for the timely management of patients to prevent the spread of RR-TB infection.", "keywords": [ "Rifampicin Resistant", "tuberculosis", "Pre-treatment loss to follow-up." ], "content": "Introduction\n\nRifampicin resistant tuberculosis (RR-TB) continued to be a global health challenge. In 2017, the estimated incidence of RR-TB cases was 0.5 million, but only 0.13 million were notified to National TB Control Programs (NTPs), meaning 0.37 million RR-TB were not identified and notified. About 87% of these notified cases were enrolled for treatment, resulting in attrition of about 13% cases of RR-TB from system and high burden countries like India and China alone contribution was 40%. In 2017, the estimated RR-TB incidence in Pakistan was 15,000. The number of laboratory-confirmed cases was 3475, of which 3016 were enrolled for treatment1.\n\nAn estimated 13% of RR-TB patients are missed from care in Pakistan in 20171. There could be many reasons for missing cases of RR-TB, such as a lack of patient accessibility to health care facilities, patients reaching hospital but not being properly diagnosed, patients being diagnosed but not enrolled, and patients being privately diagnosed and treated, but not notified to the NTP. We defined patients as pre-treatment loss to follow-up, as “any RR-TB patient detected by Xpert MTB/RIF assay but not initiated on RR-TB treatment with a TB control program’s setup (programmatic management of drug resistant TB (PMDT) site)2. Such patients, if untreated, are likely to die and/or continue to transmit the RR-TB infection in the community3.\n\nThe World Health Organization (WHO) recommend that the Xpert MTB/RIF assay should be used rather than conventional microscopy as the initial diagnostic test in presumptive TB cases (PTC), which is endorsed in the majority of laboratories worldwide for rapid and improved diagnosis4. Only one in five people with RR-TB gain access to treatment5. Data from PMDT sites suggest that RR-TB cases are regularly being detected by Xpert testing, but not all are being enrolled for management at PMDT sites. In 2014; about 3243 cases of RR-TB were detected in Pakistan, while 2662 were enrolled for treatment6.\n\nStudies from neighboring countries like Bangladesh and India reported pretreatment lost to follow-up rates of 8–21 %7–9. Another study from Vietnam showed that only 18.7% (948/5065) of RR-TB cases were enrolled for treatment10. Studies from Zimbabwe and South Africa reported 44% and 53% RR-TB patients started treatment, respectively11. However; in Pakistan we find limited data regarding enrollment of RR-TB patients, which are a potential source for the spread of DR-TB in the community3. Hence; this important issue needs to be addressed from both a patient and public health perspective. Therefore this study was done to assess the magnitude of pre-treatment loss to follow-up of RR-TB patients detected and enrolled for treatment and factors associated, that could be investigated thoroughly.\n\n\nMethods\n\nThis is a retrospective cohort study based on review of the routinely managed program data and records.\n\nBalochistan is one of the five provinces of Pakistan, and is situated on the southwest part of the country. It is the largest province and covers an area of 347,190 km212. It constitutes approximately 44% of the total land area of the country and is comprised of 33 districts13. In 2017, the population was estimated at 1.2 million14, which is scattered across difficult-to-reach terrain. The capital of the province is Quetta, the ninth largest city of Pakistan, which located in the northwest of the province near the Pakistan-Afghanistan border and is densely populated (with a population of 2 million).\n\nTB care facilities established by the Provincial TB Control Program (PTP) through an integrated approach at the existing primary, secondary and tertiary health care facilities are providing free-of-cost diagnosis and treatment services to TB patients. There were three Xpert sites in the province during the study period, where Xpert MTB/RIF assay services were available for diagnosis of RR-TB patients. The PTP had also the PMDT sites for the management of the diagnosed DR-TB patients namely; Fatima Jinnah Chest and General Hospital Quetta, District Head Quarter Hospital Loralai and District Head Quarter Hospital Turbat.\n\nThe data from Fatima Jinnah Chest and General Hospital (Quetta) and DHQ Hospital (Loralai) sites was included in study. The PMDT site in Turbat was excluded from the study because it was not functional during the study period.\n\nThe study population included all RR-TB patients detected at Xpert sites and enrolled at PMDT sites from 2012–17. All RR-TB patients referred out for enrollment at other than the study PMDT sites were also included. Patients detected at Xpert site that died before enrollment at PMDT site were excluded from pretreatment lost to follow-up.\n\nData were extracted from the RR-TB registers of the Xpert site’s program database and was validated with the Electronic Nominal Registration System (ENRS) at PMDT sites. Data was entered on a structured data collection form. Socio-demographic variables, including age, sex, address of patient (within and out of district) and distance from PMDT site, were collected to find out any association with outcome variable pre-treatment loss to follow-up.\n\nData of patients was collected on a designed data collection form and was kept confidential in password protected computer in soft and lockable cabinet in hard. The demographic characteristics of patients was not revealed in study except address, as it was requirement of study to find out association with enrollment of patient. This data is only be accessible to principle investigator and will be maintained securely for five years after completion of study.\n\nThe data being utilized for the research projects is program data routinely collected, validated and processed by the principal investigator, and an ethical clearance request letter from program manager TB control program was obtained, which stated that a specific local ethical clearance was not required in utilizing this data. There was no direct contact with the patient, so requirement for patient consent were waived.\n\nData collected was double- entered, validated and analyzed using EpiData version 3.1 for entry and version 2.2.2.183 for statistical analysis. Descriptive analysis was used for the proportion of patients with RR-TB. The association of socio-demographic factors with pre-treatment loss to follow-up was assessed using a chi-square test. The level of significance was set at P<0.05.\n\n\nResults\n\nA total of 78 (18.9%) out of 396 detected patients with RR-TB were pre-treatment loss to follow-up. Of the detected RR-TB patients, 98% were from the Xpert site at Fatima Jinnah Chest and General Hospital (Quetta) and 60% were females. The mean age was 37 years (SD-16.98) and 189 were of age group 15–34. About 55% were from outside the district, with 10 patients from out of the country. The median distance of the patient’s residence from PMDT sites was 78 km (range, 2–782 km) and only 6 patients started treatment among 84 individuals referred out to other facilities. A significant association was found between address and distance of patient’s residence with pre-treatment lost to follow-up (P<0.05) (Table 2). Raw data for this study are available on OSF15.\n\nRR+ve, rifampicin-resistance positive.\n\nPMDT, programmatic management of drug resistant TB; PTLF, Pre-treatment lost to follow-up. *Significant association\n\nOut of 78 pretreatment lost to follow up patients, 55% belonged to the 15–24 age group and females were almost 58%. About 51% patients were from within the district while 13% from outside of the country and 43 patients (55 %) were within 50 km of PMDT sites. A significant association was found between address and distance of patient’s residence with pre-treatment lost to follow-up (P<0.05); (Table 2).\n\n\nDiscussion\n\nThe study reported that 19.8% of RR-TB patients were pretreatment loss to follow-up among RR detected patients at selected PMDT sites of Balochistan. The possible reasons for pretreatment loss to follow-up may be due to poor coordination among Xpert and PMDT sites3, lack of awareness about disease and treatment; however, studies in other settings show enough knowledge among individuals about RR-TB as a disease16–18, indicating the need to assess the knowledge and attitude of individuals about TB in Pakistan. Also observed has been treatment refusal from the patient’s side due to the stigma surrounding TB in society19,20.\n\nWe found an association between pretreatment loss to follow-up with address and patient’s residence distance from PMDT sites. It is evident that the majority of patients those who were lost to follow up were from Quetta district and areas which were within 50 km of PMDT sites, which indicated that patients might give the wrong address at time of registration for their convenience and requirement for enrollment. Patients lost from outside the country were from Afghanistan, and were considered pretreatment loss to follow-up because we couldn’t find any documented proof of their treatment initiation at PMDT sites in the country of residence.\n\nA large proportion of RR-TB patients and pretreatment loss to follow-up belong to the younger age group (15–35 years). One reason seems to be that young patients are more exposed to the outside world and are in contact with individuals. Secondly, due to Islamic and Pakistani culture, young individuals facilitate activities for their old family members in many aspects of life without any precautions, which might be a potential source of disease transfer to young age groups, which means that screening of these patients should be strongly suggested.\n\nThis study has multiple strengths. First, that data was routinely maintained program data, recorded in both hard and soft forms at PMDT sites. Second, data was double-entered and validated to ensure quality21. Third, all RR-TB patients included in study to obtain the precise results. Lastly, the study was conducted in accordance to guidelines of Strengthening the Reporting of Observational Studies in Epidemiology (STROBE)22.\n\nThe limitations of this study was that we couldn’t access patients directly as the data were collected from previous routinely recorded data; most of those patients who were referred out for treatment, particularly those from outside Pakistan, were reported as pretreatment loss to follow-up because we couldn’t find any record of their treatment. However, they might be undergoing treatment.\n\nThe results of this study indicate important implications for policy makers. A strong strategy is needed to strengthen the out-of-country referral system. A strong channel should be made between Xpert sites and PMDT sites for registration of patients and coordination training should be given to persons involved in this process. I.D cards should be made mandatory to fill patient fields in the Xpert register at time of registration to provide accurate details for tracing purpose. Data from both PMDT and Xpert sites should be routinely reviewed to ascertain patient registration status and the timely tracing of patients. Patient proper education and awareness at the time of referral and enrollment for MTB/RIF assay at Xpert site. Community awareness interventions should be initiated to improve knowledge about TB, in particular RR-TB, and to counter stigma against this disease in society.\n\n\nConclusion\n\nThe high proportion of pre-treatment loss to follow-up among detected patients with RR-TB in Baluchistan needs immediate strategies for establishment of linkages between Xpert and PMDT sites for the timely management of patients to prevent the spread of DR-TB infection.\n\n\nData availability\n\nRaw data associated with this study are available on OSF. Also included is a description of abbreviations used in the dataset. DOI: https://doi.org/10.17605/OSF.IO/9UP8715.\n\nData are available under the terms of the Creative Commons Zero “No rights reserved” data waiver (CC0 1.0 Public domain dedication).", "appendix": "Grant information\n\nThis research was conducted through the Structured Operational Research and Training Initiative (SORT IT), a global partnership led by the Special Programme for Research and Training in Tropical Diseases at the World Health Organization (WHO/TDR). The training model is based on a course developed jointly by the International Union against Tuberculosis and Lung Disease (The Union, Paris, France) and Médecins Sans Frontières (MSF, Geneva, Switzerland). The specific SORT IT programme that resulted in this publication was implemented by the National Tuberculosis Control Programme of Pakistan, through the support of the Global Fund to Fight AIDS, Tuberculosis and Malaria (The Global Fund, Geneva, Switzerland). The publication fee was covered by the Special Programme for Research and Training in Tropical Diseases at the World Health Organization (WHO/TDR).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe acknowledge the Provincial TB control Program Balochistan, Mr Muhammad Umar and Mr Abdul Samad microbiologists of the PRL Balochistan through supporting the research by extracting the program data and resources.\n\n\nReferences\n\nGlobal TB Report. 2018. Reference Source\n\nMacPherson P, Houben RM, Glynn JR, et al.: Pre-treatment loss to follow-up in tuberculosis patients in low- and lower-middle-income countries and high-burden countries: a systematic review and meta-analysis. Bull World Health Organ. 2014; 92(2): 126–38. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWali A, Kumar AMV, Hinderaker SG, et al.: Pre-treatment loss to follow-up among smear-positive TB patients in tertiary hospitals, Quetta, Pakistan. Public Health Action. 2017; 7(1): 21–5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAcosta CD, Rusovich V, Harries AD, et al.: Selection of a moxifloxacin dose that suppresses drug resistance in Mycobacterium tuberculosis, by use of an in vitro. pharmacodynamic infection model and mathematical modeling. International Journal of Tuberculosis and Lung Disease. 2014; 2.\n\nAt first global ministerial meeting on TB, MSF and Stop TB Partnership give governments deadline to dramatically increase access to testing and treatment - World ReliefWeb. [cited 2018 Mar 30]. Reference Source\n\nUsaid: Mdr-Tb Country Profile. 2016; 1. Reference Source\n\nHossain ST, Isaakidis P, Sagili KD, et al.: The Multi-Drug Resistant Tuberculosis Diagnosis and Treatment Cascade in Bangladesh. PLoS One. 2015; 10(6): e0129155. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShewade HD, Kokane AM, Singh AR, et al.: Treatment Initiation among Patients with Multidrug Resistant Tuberculosis in Bhopal District, India. J Tuberc Res. 2017; 5: 237–42. Publisher Full Text\n\nShewade HD, Shringarpure KS, Parmar M, et al.: Delay and attrition before treatment initiation among MDR-TB patients in five districts of Gujarat, India. Public Health Action. 2018; 8(2): 59–65. PubMed Abstract \| Free Full Text\n\nHoang TT, Nguyen NV, Dinh SN, et al.: Challenges in detection and treatment of multidrug resistant tuberculosis patients in Vietnam. BMC Public Health. 2015; 15(1): 980. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCox H, Dickson-hall L, Ndjeka N, et al.: Delays and loss to follow-up before treatment of drug-resistant tuberculosis following implementation of Xpert MTB/RIF in South Africa: A retrospective cohort study. PLoS Med. 2017; 14(2): e1002238. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBalochistan \| province, Pakistan \| Britannica.com. [cited 2018 Mar 31]. Reference Source\n\nDIVISION , DISTRICT / CENSUS DISTRICT BALOCHISTAN PROVINCE. 1. Reference Source\n\nUnits A, Pakhtunkhwa K: PROVINCE WISE PROVISIONAL RESULTS OF CENSUS - 2017 ADMINISTRATIVE UNITS POPULATION 2017 POPULATION 1998 KHYBER PAKHTUNKHWA PUNJAB residing with the local population. 2017; 1–18. Reference Source\n\nKurd S: Pre-Treatment Loss to Follow-up among Patients with Rifampicin-Resistant Tuberculosis in Baluchistan, Pakistan, 2012-17: A Retrospective Cohort Study. OSF. Web. 2018. http://www.doi.org/10.17605/OSF.IO/9UP87\n\nRami K, Thakor N, Patel A: Awareness and knowledge about tuberculosis in patient of tuberculosis at GMERS Medical College and Hospital Dharpur, Patan, Gujarat. Int J Med Sci Public Health. 2015; 4(7): 906–909. Publisher Full Text\n\nDesalu OO, Adeoti AO, Fadeyi A, et al.: Awareness of the Warning Signs, Risk Factors, and Treatment for Tuberculosis among Urban Nigerians.Tuberc Res Treat. 2013; 2013: 369717. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nThapa B, Prasad BM, Chadha SS, et al.: Serial survey shows community intervention may contribute to increase in knowledge of Tuberculosis in 30 districts of India. BMC Public Health. 2016; 16(1): 1155. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCremers AL, de Laat MM, Kapata N, et al.: Assessing the consequences of stigma for tuberculosis patients in urban Zambia. PLoS One. 2015; 10(3): e0119861. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCourtwright A, Turner AN: Tuberculosis and Stigmatization: pathways and interventions. Public Health Rep. 2010; 125 Suppl 4: 34–42. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShewade HD, Nair D, Klinton JS, et al.: Low pre-diagnosis attrition but high pre-treatment attrition among patients with MDR-TB: An operational research from Chennai, India. J Epidemiol Glob Health. 2017; 7(4): 227–33. PubMed Abstract \| Publisher Full Text\n\nvon Elm E, Altman DG, Egger M, et al.: The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet. 2007; 370(9596): 1453–7. PubMed Abstract \| Publisher Full Text" }	[ { "id": "44665", "date": "04 Mar 2019", "name": "Hemant Deepak Shewade", "expertise": [ "Reviewer Expertise MDR-TB", "TB", "TB-DM", "Primary health care", "ACF for TB" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nOverall there is lack of clarity of what the study population is. Are they all RR-patients irrespective of their address? Of them, how was the outcome ascertained? A clear operational definition of “not initiated on treatment” is missing. Death has been excluded which should not be done. It should be included under pre-treatment loss to follow-up or reported separately. Detailed comments below:\nFirst two lines of abstract:\nPlease check the English.\n\nMay not be relevant for this paper.\n\nAbstract – Background:\n\"implementation of Xpert MTB/RIF assay, which is a rapid molecular based test and more sensitive than conventional microscopy which detects MTB even present in small limit of 136 MTB/ml of sputum\" - may not be relevant for this paper.\n\nAbstract – Background:\nWhy is this study not clear? Is it because there is no information in Pakistan on this topic?\n\nAbstract – Methods:\nAvoid using the term “retrospective” (please follow STROBE guidelines).\n\nAbstract – Methods:\nHow can patients enrolled into PMDT sites (I assume this means starting MDR-TB treatment) be included in the denominator? This is the outcome of interest and is included in the numerator.\n\nAbstract - Results:\n84 out of 396? Of 84, 6 were started on treatment. So the remaining 78, were they not considered pre-treatment LTFU? Making it 78+78?\n\nIntroduction - first line:\n“continues”, not “continue”.\n\nIntroduction – paragraph 2:\n13% RR-TB are missed from care - is this a repetition? Wasn’t the point already made before?\n\n“Missed from care” is not clear. Do you mean after diagnosis? The reasons mentioned below also include reasons for pre-diagnosis LTFU.\n\nIntroduction - paragraph 2:\nClarity of operational definitions: 1) Not initiated on treatment - within the district? In a province or anywhere in the country? 2) Not initiated on treatment - within how many days of diagnosis?\n\nIntroduction - paragraph 3:\nWHO recommends (not “recommend”).\n\nIntroduction - paragraph 3:\n\"Xpert MTB/RIF assay should be used rather than conventional microscopy as the initial diagnostic test in presumptive TB cases (PTC),\" - is this relevant for this study?\n\nIntroduction - paragraph 3:\n\n\"In 2014; about 3243 cases of RR-TB were detected in Pakistan, while 2662 were enrolled for treatment6.\" - why share 2014 data here when 2017 data has been shared before?\n\nIntroduction - last line:\n\"that could be investigated thoroughly.\" - may be removed.\n\nStudy design:\nSame comment as in the abstract (may mention cohort study involving secondary data).\n\nStudy settings:\nAre there three Xpert sites for 33 districts in Balochistan? If yes, please stress this point. Is this sufficient or are these too few? Please make this judgement call in the setting (can bring this later in the discussion).\n\n2 million (Quetta population) appears to be more than 1.2 million (province population).\n\nSo there were three PMDT sites. Were these also the Xpert sites? Of these three sites, only two were functional.\n\nPlease mention, to which population did these sites cater to? Are they for the whole of the province (Balochistan)?\n\nStudy population:\nFor RR-TB patients detected in Xpert sites with an address out of the province, were they included or excluded?\n\nThe way I would frame study population is as follows: “RR-TB patients detected at the two Xpert sites in Balochistan during 2012-17 and belonging to Balochistan were included in the study.”\n\nStudy population should not include the enrolled patients.\n\nAll diagnosed patients (RR-TB) from the Xpert sites should be included. Enrolment at PMDT is your outcome of interest.\n\n\"RR-TB patients referred out for enrolment at other than the study PMDT sites were also included.\" - this should not be mentioned under study population.\n\nDeath is one of the reasons for pre-treatment loss to follow up and therefore should be considered as pre-treatment LTFU. First of all, this statement again (that deaths were excluded) should not be mentioned here in study population. It should be mentioned under outcome ascertainment.\n\nSources of data and data collection:\nThe authors need to mention the operational definitions here (may be a separate paragraph after this paragraph).\n\nClarity of operational definitions: 1) Not initiated on treatment - within the district? In a province or anywhere in the country? 2) Not initiated on treatment - within how many days of diagnosis?\n\nDeaths should be considered as pre-treatment loss to follow up. In other words, if patients do not get initiated on treatment, we should also document how many of them died before treatment initiation.\n\n“Out of district or within district” - should this not be “province”? There is no description of district before. These sites cater to Balochistan. Also there are 33 districts in Balochistan.\n\nAlso why were the dates not collected (diagnosis and treatment)? This way the authors could have calculated the pre-treatment delay among patients initiated on treatment\n\nData confidentiality:\nThis paragraph is not required.\n\nEthics:\nMention the name of the ethics committee providing approval (approval number and date).\n\nMention whether administrative approval was obtained.\n\nPut this paragraph after analysis.\n\n“validated” may not be the right word to use.\n\nStatistical analysis:\nAs this is a cohort study, please mention (should be clarified earlier) what the duration was of follow up for each patient before he/she was declared as pre-treatment loss to follow up.\n\nMay then summarize the association between factors and pre-treatment LTFU using relative risk (cumulative incidence ratio if follow period was consistent for each patient).\n\nResults:\nAs the study population lacks clarity I am not clear how to interpret this.\n\nThere were 396 patients (I assume you included all RR-TB patients irrespective of their addresses). Of them how did you decide how many started treatment? Where did they start? Was it anywhere in these three sites? Was it anywhere in the country?\n\nRegarding the 84 that were referred out: Are these 84 out of 396? Of 84, 6 were started on treatment. So the remaining 78, were they not considered pre-treatment LTFU? Making it 78+78?\n\nFor “factors associated”, please present a table with RR and aRR (for risk factor analysis).\n\nDiscussion - third line:\nJust say Balochistan, as you included all sites in Balochistan.\n\nDiscussion:\nAll discussions around risk factors based on the aRR and their 95% Cis.\n\nAnother point for consideration - Results/Discussion: There are three Xpert sites and three PMDT sites. If both are located in the same hospital, then if the patient belongs to the population covered under the PMDT site and is diagnosed in the Xpert facility in the PMDT site, then treatment initiation should not be a problem. If the patient belongs to a population covered under another PMDT site, then the patient has to be referred for treatment to other sites and here loss to follow up is common. Should this be considered as one of the factors (for analysis) for pre-treatment LTFU? However, the larger question is the operational definition of study population and operational definition of outcomes (treatment initiation).\n\nIf the patient belongs to the population covered under the PMDT site and is diagnosed in the Xpert facility in the PMDT site, then treatment initiation should not be a problem. This should be included as one of the factors (for analysis) for pre-treatment LTFU.\n\nIs the work clearly and accurately presented and does it cite the current literature? No\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] }, { "id": "45259", "date": "26 Apr 2019", "name": "Amer Hayat Khan", "expertise": [ "Reviewer Expertise Research in Infections & Nephrology" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nTable 1: Socio-demographic… “Address” - what does it mean and what is its impact on the findings?\n\nTable 1: Xpert results (RR+VE)…I think it is Inclusion criteria, I will suggest to omit this from the table.\n\nTable 2: PTLF and P-Value…For the “Age less than 15 years” group, PTLF is 00, so how is the P-value 0.24?\n\nTable 2: It is confusable, Urban and Rural P-values for both, while Address P-values only once? Uniformity must be brought to the article and its format.\n\nDo the researchers want to prove that such defaulter is due to distance? There are no further reasons?\n\nThe discussion needs to be improved.\n\nThe conclusion seems like a general statement, it should be based on the study findings.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/7-1905
https://f1000research.com/articles/7-1901/v1	06 Dec 18	{ "type": "Software Tool Article", "title": "ProbeSpec: batch specificity testing and visualization of oligonucleotide probe sets implemented in ARB", "authors": [ "Tim Kahlke", "Paavo Jumppanen", "Ralf Westram", "Guy C.G. Abell", "Levente Bodrossy", "Paavo Jumppanen", "Ralf Westram", "Guy C.G. Abell", "Levente Bodrossy" ], "abstract": "High-throughput molecular methods such as quantitative polymerase chain reaction (qPCR) and environmental microarrays are cost-effective methods for semi-quantitative assessment of bacterial community structure and the identification of specific target organisms. Both techniques rely on short nucleotide sequences, so-called oligonucleotide probes, which require high specificity to the organisms in question to avoid cross-hybridization with non-target taxa. However, designing oligonucleotide probes for novel taxa or marker genes that show sufficient phylogenetic sensitivity and specificity is often time- and labor-intensive, as each probe has to be in-silico tested for its specificity and sensitivity. Here we present ProbeSpec, to our knowledge the first batch sensitivity and specificity estimation and visualization tool for oligonucleotide probes integrated into the widely used ARB software. Using ProbeSpec’s interactive “mismatch threshold” and “clade marked threshold” we were able to reduce the development time of highly specific probes for a recently published environmental oligonucleotide microarray from several months to one week.", "keywords": [ "Probe design", "qPCR", "microarray", "Bioinformatics", "molecular ecology", "microbiology" ], "content": "Introduction\n\nThe analysis of the microbial community structure and abundance based on universal conserved marker genes has become a powerful tool for many disciplines in life science with a specific focus on next-generation sequencing technologies1,2. In addition to these qualitative methods technologies such as environmental microarrays and quantitative polymerase chain reaction (qPCR) offer cost-effective and highly reproducible techniques for semi-quantitative estimation of microbial communities. Genetic markers commonly used for microarrays and qPCR are ribosomal RNA (rRNA) genes, e.g. 16S, for bacterial communities3,4, as well as functional genes that determine microbial community structure with regards to specific metabolic functions5,6. Both technologies rely on taxon-specific short nucleotide sequences of the marker gene of interest, so-called oligonucleotide probes (OPs). In qPCR experiments OPs act as the primer to initiate the amplification reaction whereas in microarrays the probe is spotted onto a glass slide and the complementary sequence is hybridized with it.\n\nA major challenge in using both techniques for novel organisms and marker genes, however, is the development of OPs with appropriate levels of taxonomic specificity and sensitivity: especially functional genes show highly variable levels of conservation, not only between sequences of different taxa but also between sequences of closely related organisms. Thus, depending on the experiment, the functional marker and the organisms of interest, hundreds or even thousands of OPs with varying levels of conservation have to be designed and subsequently in-silico tested for their phylogenetic specificity and sensitivity. A major bottleneck for this process is the lack of software tools that enable researchers to test multiple potential OPs for their phylogenetic specificity at once.\n\nHere we present ProbeSpec7, a user-friendly, interactive probe specificity and sensitivity assessment tool for OPs with batch analysis support. ProbeSpec’s functionality is incorporated into the widely used ARB software8 which is freely available for non-commercial use (detailed copyright information can be found here and in the license agreement included in each tarball). To our knowledge, ProbeSpec is the only batch probe specificity assessment tool which provides interactive manipulation of specificity and sensitivity thresholds.\n\n\nMethods\n\nProbeSpec is implemented in ARB’s PROBE_DESIGN class utilizing its prefix tree database server. ProbeSpec’s functionality is implemented in the classes ArbProbe and ArbProbeCollection (abstraction of OP sequences and import/export functionality), ArbProbeMatchWeighting (providing weighting matrices for position specific nucleotide substitutions), ArbMatchResult, ArbMatchResultSets and ArbMatchResultsManager (abstraction of OP to PT-Server sequences with given weighting matrices and maximum number of mismatches) and ArbStringCache (providing string to disk caching of match string results).\n\nProbe specificity calculations in ProbeSpec are based on the initial mismatch penalties given by a 4×4 substitution matrix for all possible nucleotide substitutions. Additionally, each mismatch penalty is weighted based on the position of a mismatch in the probe: mismatches at the ends of an OP are less likely to affect the binding of complementary sequences than mismatches in the center of a probe. Positional weights are calculated as follows: for a mismatch at position p in a given OP sequence of length l a weight W is calculated with\n\nW=eSP2(1)\n\nwhere\n\nS=−In(10)w(2)\n\nand\n\nP=2*p−ll−b(3)\n\nThe weight distribution given by (1) follows a bell curve penalizing mismatches at either end of the OP sequence less than mismatches in the center of the sequence. The user defined parameter w in equation (2) controls the spread of the weight distribution; user defined parameter b in equation (3) controls the midpoint and therefore enables the user to increase positional weights on either side of the OP sequence. For default parameters of w=1 and b=0, positional weights range from a minimum 0.1 for mismatches at the first and last nucleotide in the sequence to a maximum of 1 for mismatches at the center.\n\nFor user interaction with ProbeSpec ARB’s general user interface was extended with four new dialog windows: (i) a Probe Collection dialog, (ii) a Probe match with specificity, (iii) Match display control dialog and (iv) a Tree Marker settings dialog (Figure 1).\n\n(A) Probe Collection dialog. (B) Tree marking settings dialog. (C) Match Display Control; (D) Match Display Control. Coloured vertical bars on the left of the main window represent (partially) matching probes.\n\nThe Probe match with specificity is the main entry point of ProbeSpec. It displays all loaded probes which can be edited, imported and exported through the Probe Collection dialog. Additionally, the Probe Collection Dialog allows the user to change the default settings for substitution penalties and positional weight parameters.\n\nThe main GUI of ARB was extended to graphically represent the probe matching results: each probe is represented by a colored vertical bar indicating a match of the OP to the specific phylogenetic group. Incomplete cover of a phylogenetic group is represented by transparency of a bar: the fewer members of a group that are covered by a given probe the higher the transparency of a bar is.\n\nThe dialogs Match Display Control and Tree marking settings enable interactive adjustment of probe match parameters such as mismatch threshold, group marked and group partially marked threshold.\n\nARB and the included ProbeSpec functionality can be run on any common PC, laptop or workstation. However, we recommend system specifications of at least 4GB of RAM and a dual-core processor to run ProbeSpec.\n\n\nUse case\n\nUsing ProbeSpec we were able to test the specificity of 345 OP sequences against an ARB database of 20,314 bacterial and archaeal ammonia mono-oxygenase sequences on a Ubuntu Virtual Machine with 4 GB of RAM and one processor allocated in less than 30 minutes. In comparison: sequential specificity testing without ProbeSpec for a recent publication9 on the same data set took several days\n\nFor any probe development, ProbeSpec requires a phylogeny of target sequences and organisms that the OPs should match to as well as a list of potential OPs.\n\nFor an introduction to sequence analysis using ARB, please refer to the main ARB documentation at http://www.arb-home.de/documentation.html. For evaluation purposes a sub-set of the data published in Krausfeldt et al. (2017) can be found on Zenodo10. To set up ARB select the provided nitrifyers_2017_04_for_paper.arb database file on start of ARB. To be able to run ProbeSpec a PTServer has to be created from the database via the Probes tab and the PT_Server Admin option in the PT Server Admin widget. Select the loaded database and click Build server. After completion close the progress bar and the PT Server Admin widget.\n\nBefore running a batch specificity test, a probe collection, i.e., a list of probes to be tested, has to be created using the Probe Collection window where probes can be added to and removed from a collection: Open the Probe Match with Specificity window via the Probes-tab in ARB (Figure 1A) and select Edit (Figure 1B) to open the Probe Collection window (Figure 1C). To open the provided test data set use the load button and select the provided amoA70mers.xpc probe collection. Additionally, the sequence of new probes can be entered into the Target String text field. To add new OPs to the collection press Add. Probe collections can also be in this dialog.\n\nThe Probe Collection window can be used to define the specificity measures used by ProbeSpec to identify matching probes. This includes the definition of specific mismatch penalty values as well as the values for bias b and weight w (see subsection Operations in the Methods section for details).\n\nAfter creation of a probe collection and configuration of the match parameters the Probe Collection window can be closed and the specificity search can be started by clicking the Match button (Figure 1B). A status dialog will appear and show the progress of the search.\n\nThe final match results are shown in the ARBs main window: each matching probe is represented by a coloured bar next to the group/clade the probe matches with the given thresholds (Figure 1A). The visualization can be configured using the two dialogs Match Display Control (Figure 1B) and Tree Display settings (Figure 1C), the latter of which can be accessed via the Marker Display Settings button on the Match Display Control widget.\n\n\nConclusion\n\nHere we present ProbeSpec, to our knowledge, the first tool for batch specificity testing of OP sequences implemented in ARB. ProbeSpec offers significant time saving for projects developing and testing large oligonucleotide probe datasets for use in technologies such as qPCR and environmental microarrays.\n\n\nData availability\n\nFor test and validation purposes, a sub-set of the data published in Krausfeldt et al. (2017) can be found at Zenodo, DOI: http://doi.org/10.5281/zenodo.148295810. The dataset includes a phylogeny of archaeal and bacterial amoA sequences (nitrifyers_2017_04_for_paper.arb) as well as a sub-set of 185 OPs used to create the environmental microarray.\n\n\nSoftware availability\n\nProbeSpec is included in the production version of ARB, available at: http://download.arb-home.de/special/manual-builds/.\n\nArchived version of the production version directory: http://doi.org/10.5281/zenodo.14829477.\n\nLicense: ARB License.", "appendix": "Grant information\n\nThis work was supported by the Environmental Genomics grant from CSIRO Oceans & Atmosphere (R-02412).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nBrown MV, van de Kamp J, Ostrowski M, et al.: Systematic, continental scale temporal monitoring of marine pelagic microbiota by the Australian Marine Microbial Biodiversity Initiative. Sci Data. 2018; 5: 180130. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBork P, Bowler C, de Vargas C, et al.: Tara Oceans. Tara Oceans studies plankton at planetary scale. Introduction. Science. 2015; 348(6237): 873. PubMed Abstract \| Publisher Full Text\n\nLazarevic V, Gaïa N, Girard M, et al.: Decontamination of 16S rRNA gene amplicon sequence datasets based on bacterial load assessment by qPCR. BMC Microbiol. 2016; 16: 73. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFigueroa IA, Barnum TP, Somasekhar PY, et al.: Metagenomics-guided analysis of microbial chemolithoautotrophic phosphite oxidation yields evidence of a seventh natural CO2 fixation pathway. Proc Natl Acad Sci U S A. 2018; 115(1): E92–E101. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAbell GC, Robert SS, Frampton DM, et al.: High-throughput analysis of ammonia oxidiser community composition via a novel, amoA-based functional gene array. PLoS One. 2012; 7(12): e51542. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLee YJ, van Nostrand JD, Tu Q, et al.: The PathoChip, a functional gene array for assessing pathogenic properties of diverse microbial communities. ISME J. 2013; 7(10): 1974–1984. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKahlke T, Jumppanen P, Westram R, et al.: ARB tarballs 19.10.2018 (Version r17491). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1482947\n\nLudwig W, Strunk O, Westram R, et al.: ARB: a software environment for sequence data. Nucleic Acids Res. 2004; 32(4): 1363–1371. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKrausfeldt LE, Tang X, van de Kamp J, et al.: Spatial and temporal variability in the nitrogen cyclers of hypereutrophic Lake Taihu. FEMS Microbiol Ecol. 2017; 93(4): fix024. PubMed Abstract \| Publisher Full Text\n\nKahlke T, Jumppanen P, Westram R, et al.: ProbeSpec validation data [Data set]. Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1482958" }	[ { "id": "41599", "date": "22 Jan 2019", "name": "Michael Dondrup", "expertise": [ "Reviewer Expertise Bioinformatic", "genomics", "computational biology" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this manuscript, the authors present a software for testing the specificity of oligonucleotide probe-sets that is integrated into the sequence analysis software ARB with a focus on microbial ecology. The software allows for batch-scoring the specificity of probe-sets using a bell-shaped scoring function based on the distance from the center of the probe, with several user-definable parameters.\n\nThe application domain of the new functionality is timely because high-throughput methods that rely on oligonucleotide probes are becoming more and more common. The article is overall well written, but lacking with respect to the description of the methodology, state of the art, and depiction of software development and code.\n\nFor the sake of a comprehensive evaluation, I have tested the ARB software using the latest development build 6.1.rev17491and the author-provided dataset in a virtual machine under Linux Mint 9.\n\nMajor Concerns:\n\nThe authors present a new scoring function that introduces bell-shaped weights of mismatches depending on the distance to center base of the probe. It is unclear what the motivation for using this function is, and what underlying assumptions are that it is based on, or if has been used in the literature before. I assume it is trying to compensate for some hybridization effects, but these might also vary between technologies. Other publications have discussed thermodynamic parameters instead, see e.g. Kabilov et al.1. It should also be made more clear, if and how the user can specify parameters, adjust them to the assay or how to use a flat scoring function instead. The authors claim that their software is the first tool for batch testing, which might be true. However, with respect to the new scoring functions, authors should still compare their estimates to other existing tools. Searching for other tools I found probeCheck2, which is an online-tool also using ARB as a back-end, but allows display of max 10 results at a time, and does not allow to use custom databases, but the outcomes could be comparable for the same probe set. As ProbeSpec is part of large software package and is only obtainable as a component it is hard to evaluate the contribution by source code. It should be clearly pointed out which files of the ARB distribution contain the source code of ProbeSpec or possibly provide a separate (e.g. git) repository that contains the sources. The functional test completes properly, but I noticed that the probe set is loaded and match parameters could be set under Edit which is a bit confusing (substitution matrix, shape parameters). It should be better explained in the UI that these are matching parameters.\n\nMinor concerns:\nS=−In(10) in Equation (2), should it be ln(10) for logarithm (L vs. I) Even though it is an aspect of the software which cannot be changed easily, I would like to question the use of ARB as the software platform. The appearance of the user interface is archaic, while this might be dismissed as cosmetics, worse so the build system is arcane too. While I was finally able to compile the latest ARB revision including ProbeSpec from source under Linux Mint, this will be prohibitive for most users, mostly due to undocumented/not up to data dependencies (e.g. boost library never mentioned anywhere) including combined with the lack of a (e.g. autoconf generated) configure script. Therefore, for most users the options are restricted running pre-compiled binaries to either Debian, Linux Mint or a very old Ubuntu (10) release under a virtual machine, something not very timely in the age of Docker and containers. This is a pity because common package managers contain the latest stable ARB release 6.0.6, dating back to August 2016, but not containing ProbeCheck. I would therefore recommend to the authors to: contact the maintainers of ARB to get a new stable release out supporting popular OS', contribute to a complete and updated dependency list, or provide a containerized version of the whole system.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Partly\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? No\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Partly", "responses": [] }, { "id": "46526", "date": "14 May 2019", "name": "Jizhong Zhou", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors present a novel tool, ProbeSpec, to test specificity of probes against sequence databases. The tool is presented with enough detail and deserves indexing if the authors successfully address the following questions:\nThe authors assert that ProbeSepc’s performance is superior than other similar tools in the “Use case” section. Only one recent publication was mentioned.\nAre there any additional publications recently with similar settings? Can the author provided more specific instructions for the reviewer to find where Krausfeldt mentioned their testing “took several days”? What is the most important improvement ProbeSpec has done to outperform other tools? Can the authors provide a breakdown of the running time of a typical ProbeSpec run to prove it?\n\nIn “Probe specificity matching”, the authors proposed a new scheme for positional weights.\nThe formula is very similar to a formula on page 35 of the supplement material from \"Spatial and temporal variability in the nitrogen cyclers of hypereutrophic Lake Taihu1\" (can be found here). Why use w=1 as default, which is different from w=3 in the above reference. Eq. (3). It is strange to use b!=0, in which case the weighting function will not be symmetric. What is the rationale to assign different weight to the starting and ending bp? If not necessary, what is the point to introduce parameter b? It is a little confusing to use both p and P, while there are plenty of other letters available.\n\nReference:\nKrausfeldt LE, Tang X, van de Kamp J, et al.: Spatial and temporal variability in the nitrogen cyclers of hypereutrophic Lake Taihu.1\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Partly", "responses": [] } ]	1	https://f1000research.com/articles/7-1901
https://f1000research.com/articles/7-1899/v1	06 Dec 18	{ "type": "Research Article", "title": "The influence of storage condition on nitrite, nitrate and vitamin C levels in vegetables", "authors": [ "Henni Cintya", "Jansen Silalahi", "Effendy De Lux Putra", "Rikson Siburian", "Henni Cintya", "Effendy De Lux Putra", "Rikson Siburian" ], "abstract": "Vegetables are the main sources of nitrate and nitrite in food. The presence of nitrate and nitrite at a high level may cause a negative impact on health, because nitrite and nitrate when reduced to nitrite, may react with alkylamine to form carcinogenic nitrosamine. The influence of temperature and time of storage on nitrite, nitrate, and vitamin C contents in vegetables were investigated in this study. The vegetables were sweet mustard, bokchoy, spinach and lettuce obtained from a local market. Samples were stored at ±25oC and ±5oC. Analysis of nitrite, nitrate, and vitamin C was conducted in fresh samples, after storage for 24 and 48 hours. Nitrite was analyzed by spectrophotometry at 540 nm. Nitrate reduced into nitrite with Zn in acidic conditions and then analyzed as nitrite. Vitamin C was analyzed by titration with 2.6-dichlorophenolindophenol. During storage, nitrite and nitrate increased, while vitamin C decreased. Nitrite and nitrate content in fresh samples were 15.22 and 22.46 mg/kg (sweet mustard), 12.57 and 6.55 mg/kg (bokchoy), 20.26 and 90.60 mg/kg (spinach), 18.77 and 32.68 mg/kg (lettuce), respectively. Vitamin C content in fresh samples was 101.15 mg/100g (mustard), 92.17 mg/100g (bokchoy), 88.95 mg/100g (spinach), 40.03 mg/100g (lettuce). After storage for 48 hours at ±25oC, nitrite and nitrate increased 44.97% and 53.19% (mustard), 46.18% and 62.59% (bokchoy), 43.86% and 16.48% (spinach), and 41.05% and 47.09% (lettuce), respectively. Vitamin C decreased 67.57% (mustard), 24.68% (bokchoy), 81.25% (spinach), and 79.74% (lettuce). Storage at ±5oC, showed that nitrite and nitrate increased 27.54% and 35.08% (mustard), 13.75% and 43.51% (bokchoy), 19.59% and 10.60% (spinach), 19.85% and 25.16% (lettuce), respectively. Vitamin C decreased 30.88% (mustard), 6.05% (bokchoy), 60.92% (spinach), and 74.94% (lettuce). During storage, nitrite and nitrate increased more significantly at ±25oC than ±5oC while vitamin levels C decreased and were more effective at 25oC than 5oC.", "keywords": [ "Nitrite", "Nitrate", "Vitamin C", "Storage Condition." ], "content": "Introduction\n\nVegetables are a major source of nitrite and nitrate intake from food. Nitrite and nitrate are also used as preservatives and coloring agents in processed meats1–5. Nitrate and nitrite contents in vegetables vary widely from 1 to 10.000 mg/kg, and this is affected by many factors, including environmental factors such as storage condition, processing procedure, temperature and agricultural practices6–9.\n\nNitrate can be reduced into nitrite by enzyme nitrate reductase and other reducing agents, including vitamin C, which is also contained in vegetables. Nitrite may react with alkylamines to form carcinogenic nitrosamines10,11. Therefore, the intake permitted (Acceptable Daily Intake = ADI) by the Food and Agriculture Organization of the United Nations/World Health Organization is 220 mg of nitrate and 8 mg of nitrite per day for adults weighing an average of 60 kg12,13. Previous studies reported that the longer the storage, the higher nitrite and nitrate contents. These effects are more influential at room temperature than at refrigeration. But the effect of temperature and storage condition on vitamin C have not yet been reported to the best of our knowledge. The aim of this study was to investigate the effect of storage condition on nitrite, nitrate and vitamin C contents in vegetables.\n\n\nMethods\n\nChemicals used were analysis grade products from Merck KGaA (Germany): N-(1-naphthyl) ethylenediamine dihydrochloride (NED), sodium nitrite, sulfanilic acid, glacial acetic acid, hydrochloric acid, antipyrine, ferrous sulfate, zinc powder, sodium nitrite, ascorbic acid, metaphosphoric acid, and 2.6-dichlorophenolindophenol.\n\nThe vegetables analyzed in this study were sweet mustard (Brassica rapa chinensis), bokchoy (Brassica rapa L.), spinach (Amaranthus tricolor L), and lettuce (Lactuca sativa L). These vegetables were obtained from a local market in Medan, Indonesia. Samples were stored for 0, 24, until 48 hours at room temperature (±25°C) and in a refrigerator (±5°C).\n\nIn total, 4 ml of standard solution of nitrite (C=10.0 μg/ml) was transferred into 50 ml volumetric flask, added 2.5 ml sulfanilic acid solution and shaken. After 5 min, 2.5 ml NED reagent was added and made to volume with distilled water and homogenized (C=0.8 μg/ml). Absorbance was measured at wave length of 400–800 nm. Then, absorbance and wave length was plotted to construct absorbance curve. Wave length of maximum absorbance was determined from the absorbance curve10.\n\nIn total, 4 ml of standard solution of nitrite (C=10.0 μg/ml) was transferred into volumetric flask of 50 ml, to which 2.5 ml of sulfanilic acid and stirred. After 5 min, 2.5 ml NED reagent was added and distilled water was added to make 50 ml. Absorbance was measured at wave-length of maximum absorbance obtained from absorbance curve (540 nm), and stability of absorbance was determined by observing absorbance at every minute for 1 hr. The absorbance was found to be relatively stable within 6 min in 7–12 min10.\n\nStandard solution of nitrite (C=10.0 μg/ml) of different volume (0.5, 1, 2, 3, 4 dan 5 ml) were transferred into separated volumetric flasks of 25 ml, then 2.5 ml sulfanilic acid reagent added and stirred to homogenize. After 5 min, 2.5 ml NED reagent was added, then distilled water was added to make volume of 25 ml and homogenized. The series of concentration of prepared solutions were of 0.1 μg/ml, 0.2 μg/ml, 0.4 μg/ml, 0.8 μg/ml, 1.0 μg/ml. Absorbance of each solution was measured at wave-length of 540 nm within 7 min. Calibration curve was made by plotting absorbance versus concentration of each solution. From the graph obtained, then linearity of regression equation and correlation coefficient were calculated (Y=aX+b)10.\n\nAbout 10 g ground sample, using blender, was transferred into a glass beaker. Distilled water was added to about 150 ml, heated in a waterbath (80°C) and shaken for 5 minute then cooled and filtered. The supernatant was transferred into a test tube, then 2.5 ml sulfanilic acid reagent was added and stirred. After 5 minutes, 2.5 ml reagent NED was added. Nitrite was identified using sulfanilic acid and NED solution, and the appearance of a violet color indicated the presence of nitrite. Nitrate was identified by adding several drops ferrous sulfate solution and then slowly adding a few drops of concentrated sulfuric acid. The formation of chocolate ring indicates the presence of nitrate10.\n\nNitrite. Determination of nitrite was carried out with procedure previously described10. Around ten (10) gram grounded sample transferred into 250 ml beaker glass to which hot distilled water (± 80ºC) was added about 150 ml. This mixture was homogenized by stirring and heated on waterbath for 15 minute while stirring. Allowed to cool and then transferred quantitatively into 250 ml volumetric flask, distilled water added to volume, then filtered. Ten (10 ml) of filtrate transferred into a volumetric flask of 50 ml, then 2.5 ml sulfanilic acid reagent was added and stirred. After 5 minutes, 2.5 ml reagent NED was added, then distilled water added to make 50 ml, and then homogenized. Absorbance was measured at wavelength of 540 nm after period of 7 to 12 minutes time.\n\nNitrate. In total, 10 g grounded sample transferred into 250 ml beaker glass to which hot distilled water (± 80ºC) was added to about 150 ml. This mixture was homogenized by stirring and heated on waterbath for 15 minute while stirring. Allowed to cool and then transferred quantitatively into 250 ml volumetric flask, distilled water added to volume, then filtered. 10 ml of filtrate transferred into a volumetric flask of 50 ml, then 0.1 g Zn powder and 1 ml HCl 1 N added and allowed to stand for 10 minutes to reduce nitrate to nitrite, then 2.5 ml sulfanilic acid reagent was added and stirred. After 5 minutes, 2.5 ml reagent NED was added, then distilled water added to make 50 ml, and then homogenized. Absorbance was measured at wavelength of 540 nm after period of 7 to 12 minutes time.\n\nNitrite concentration from reduction of nitrate into nitrite was calculated:\n\nNitrite concentration from nitrate reduction = Total nitrite content - Initial nitrite concentration in samples\n\nNitrate content was calculated:\n\nNitrite concentration from nitrate reduction = Concentration of nitrate × MoleculeWeigtnitrateMoleculeWeigtnitrite\n\n10 gram grounded sample using blender. About 0.5 ml of sample solution in a test tube was neutralized to a pH of 6–8 with NH4OH 1 N, three drops of 3% FeCl3 was added – a purple color indicates the presence of vitamin C.\n\n10 g grounded sample using blender was transferred into 100 ml volumetric flask, then acetic metaphosphoric acid 3% added to make 100 ml, then homogenized and filtered. Two (2) ml of filtrate was transferred into an erlenmeyer, and then added 5 ml of acetic metaphosphoric acid, then titrated with 2.6-dichlorophenol indophenol solution 0.025% until pink steadily11. The levels of vitamin C was calculated.\n\nVitamin C (mg/g) = (Vt−Vb)×Equivalence×VLVp×Bs\n\nVb = The volume of blank (ml); VL = The volume of volumetric flask (100 ml); Vp = The volume of pipetted sample solution(ml); Bs = Sample weight (g)\n\n\nResults\n\nIt is found that samples contain nitrite indicated by the appearance of violet color to prove that all samples contained nitrite. The reaction with antipyrine in dilute hydrochloric resulted in the formation of green color to prove the present of nitrite. Nitrate in samples was identified using ferrous sulfate and concentrated sulfuric acid produced brown ring10.\n\nIdentification using FeCl3 3% reagent generated violet color to prove that all samples contained vitamin C11.\n\nThe absorbance curve of the nitrite derivative solution (10 μg/ml) is presented in Figure 1. From Figure 1 it is shown that the maximum absorption was at 540 nm, which is similar to the value previously reported2,7, which was used to determine the analysis of nitrite and nitrate in samples.\n\nWorking time for nitrite and nitrate analysis was determined to know the period of time within which the absorbance of solution still remains stable. Absorbance of nitrite derivative with Griess reagent presented in Figure 2. Figure 2 shows that absorbance was stable within minute 7 to minute 12 then used in the analysis procedure2,7.\n\nCalibration curve made by plotting absorbance versus concentration of each solution, then linearity of regression equation was determined. The calibration curve presented in Figure 3. Regression equation obtained is Y= 0.58064X + 0.0015 with coefficient correlation (r) of 0.99977(where r > 0.999). Figure 3 shows that the correlation coefficient was high (r=0.999) indicated linearity between concentration and absorbance2,7.\n\nThe levels of nitrite, nitrate, and vitamin C during storage at ±25°C and ±5°C can be seen in Table 1 and Table 2.\n\nNote: data is the mean of six replicates\n\nNote: data is the mean of six replicates\n\nNitrite and nitrate levels in fresh samples were 15.22 and 22.46 mg/kg (sweet mustard), 12.57 and 6.55 mg/kg (bokchoy), 20.26 and 90.60 mg/kg (spinach), and 18.77 and 32.68 mg/kg (lettuce), respectively. Vitamin C levels in fresh samples were 101.15 mg/100g (sweet mustard), 92.17 mg/100g (bokchoy), 88.95 mg/100g (spinach), and 40.03 mg/100g (lettuce).\n\nFrom Table 1 can be seen that the levels of nitrite and nitrate also increased with storage time. After storage for 48 hours at ±25°C, nitrite and nitrate levels increased 44.97% and 53.19% (sweet mustard), 46.18% and 62.59% (bokchoy), 43.86% and 16.48% (spinach), and 41.05% and 47.09% (lettuce), respectively. While, vitamin C decreased 67.57% (sweet mustard), 24.68% (bokchoy), 81.25% (spinach), and 79.74% (lettuce).\n\nTable 2 shows that storage at ±5°C, nitrite and nitrate levels increased 27.54% and 35.08% (sweet mustard), 13.75% and 43.51% (bokcoy), 19.59% and 10.60% (spinach), 19.85% and 25.16% (lettuce), respectively. Vitamin C levels decreased 30.88% (sweet mustard), 6.05% (bokcoy), 60.92% (spinach), 74.94% (lettuce), respectively.\n\n\nDiscussion\n\nFrom Table 1 and Table 2 it can been seen that nitrite levels are generally relatively low in fresh vegetables compared to nitrate levels, except in bokchoy. This value is similar with those reported by researchers that nitrate is usually higher than nitrite12. The nitrate in plant also changes with age of the plant. The differences in nitrate content may be due to the fertilization, harvesting time, and storage time8–12.\n\nThe results indicate that storage temperature and time affect nitrite, nitrate, and vitamin C levels in vegetables. The longer the storage time the higher nitrite and nitrate levels and the lower vitamin C levels. These effects are more influential at room temperature than at refrigeration, as has previously been reported3,12.\n\nTable 1 and Table 2, suggest that the vitamin C and other antioxidant content in vegetables may reduce nitrate into nitrite, and then nitrite may react with amine compounds, especially secondary amines to form a carcinogenic nitrosamine. On the other hand, ascorbic acid available in fresh vegetables may prevent the formation of nitrosamine8–12.\n\n\nConclusion\n\nStorage condition affects nitrite, nitrate and vitamin C content in vegetables. The higher the temperature and the longer the time of storage, the higher nitrite and nitrate levels,and the lower vitamin C levels. This effect is more influential at 25°C than at 5°C.\n\n\nData availability\n\nF1000Research: Dataset 1. Raw data including working time and calibration curve data; nitrite, nitrate and vitamin C levels 25°C and 5°C and 0, 24 and 48 hours storage for 6 replicates for mustard, bokchoy, spinach and lettuce., https://doi.org/10.5256/f1000research.16853.d22722414", "appendix": "Grant information\n\nThis study was supported by DP2M DIKTI (Directorate of Higher Education) Ministry of Research Technology and High Education, Indonesia through “Hibah PMDSU” Research Grant 2017.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe gratefully thank to DP2M DIKTI (Directorate of Higher Education) Ministry of Research Technology and High Education, Indonesia through “Hibah PMDSU” Research Grant 2017 for financial support in this study.\n\n\nReferences\n\nAlexander P, Handawa P, Charles UT: Determination of Nitrate and Nitrite Contents of Some Edible Vegetables in Guyuk Local Government Area of Adamawa State, Nigeria. American Chemical Science Journal. 2016; 13(3): 1–7. Publisher Full Text\n\nHill M: Nitrate and Nitrite in Food and Water. Camridge: Woodhead Publishing Limited. 1996; 98. Reference Source\n\nSilalahi J, Fattah A, Ginting N, et al.: The Effect Storage Condition on Nitrite and Nitrate Content in Lettuce. Int J Pharmtech Res. 2016; 9(8): 422–427. Reference Source\n\nRaczuk J, Wanda W, Katarzyna G: Nitrates and nitrites in selected vegetables purchased at supermarkets in Siedlce, Poland. Rocz Panstw Zakl Hig. 2014; 65(1): 15–20. PubMed Abstract\n\nWHO: Nitrate and Nitrite. JECFA Food Additives. 2009: Series 50. Reference Source\n\nLeszczynska R, Filipak A, Cieslik E, et al.: Effects of Some Processing Methods on Nitrate and Nitrite Changes in Cruciferous Vegetables. J Food Compos Anal. 2009; 22(4): 315–321. Publisher Full Text\n\nAmr A, Hadidi N: Effect of Cultivar and Harvest Date on Nitrate (NO3) and Nitrite (NO2) Content of Selected Vegetables Grown Under Open Field and Greenhouse Conditions in Jordan J Food Compos Anal. 2001; 14(1): 59–67. Publisher Full Text\n\nWalters CL: Nitrate and Nitrite in Food. In: Hill, M. Nitrates and Nitrites in Foods and Water. Woodhead Publishing Limited. 2000; 97. Reference Source\n\nAutherhoff H, Kovar KA: Identifizierung Von Arzeistoffen. Translated by Sugiarso, Identification of Drug. ITB. 2002; 94.\n\nCintya H, Silalahi J, Putra EDL: Analysis of Nitrate and Nitrite in Vegetables in Medan City. Der Pharma Chemica. 2016; 8(24): 47–52. Reference Source\n\nDitjen POM: Indonesian Pharmacopoeia 4thEdition. Jakarta: Department of Health RI. 1995; 1133, 1135, 1164, 1168, 1215-1216.\n\nChung JC, Chou SS, Hwang DF: Changes in nitrate and nitrite content of four vegetables during storage at refrigerated and ambient temperatures. Food Addit Contam. 2004; 21(4): 317–322. PubMed Abstract \| Publisher Full Text\n\nQazi MA, Rizzatti F, Piknova B, et al.: Effect of storage levels of nitric oxide derivatives in blood components [version 1; referees: 5 approved]. F1000Res. 2012; 1: 35. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCintya H, Silalahi J, De Lux Putra E, et al.: Dataset 1 in: The influence of storage condition on nitrite, nitrate and vitamin C levels in vegetables. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16853.d227224" }	[ { "id": "43053", "date": "21 Jan 2019", "name": "Marco Iammarino", "expertise": [ "Reviewer Expertise Food Safety", "Food Science and Technology", "Analytical Chemistry", "Analytical methods validation" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this paper, the authors have reported about the influence of storage condition (5°C and 25°C) on nitrite, nitrate and ascorbic acid in vegetables.\nThe topic has already been investigated, so, the originality is lacking.\nThe references section is incomplete. Many papers dealing about this topic were not included.\n\nThe results obtained in this study cannot be considered as reliable. Indeed, the analytical techniques adopted were not validated or compared to reference parameters. Moreover, they seem obsolete. This weakness is confirmed by evaluating the results obtained for both nitrite (many authors do not report quantifiable concentrations of nitrite in spinach and lettuce) and (specially) nitrate, since the NO3- concentrations in spinach and lettuce reported in the bibliography are well higher.\n\nThe authors have not specified the number of samples analyzed and their characteristics (origin, type, etc.). Consequently the statistical analysis is not possible and the significance of these results is not high.\nIn view of these criticisms, I regret to not recommend this paper for indexing.", "responses": [ { "c_id": "4410", "date": "21 Feb 2019", "name": "Rikson Siburian", "role": "Author Response", "response": "Dear Referee,Marco Iammarino, National Reference Center for the Detection of Radioactivity in Feed and Foodstuff, Istituto Zooprofilattico Sperimentale della Puglia e della Basilicata, Foggia, Italy Thank you so much for your comments:In this paper, the authors have reported about the influence of storage condition (5°C and 25°C) on nitrite, nitrate and ascorbic acid in vegetables.The topic has already been investigated, so, the originality is lacking. The references section is incomplete. Many papers dealing about this topic were not included. The results obtained in this study cannot be considered as reliable. Indeed, the analytical techniques adopted were not validated or compared to reference parameters. Moreover, they seem obsolete. This weakness is confirmed by evaluating the results obtained for both nitrite (many authors do not report quantifiable concentrations of nitrite in spinach and lettuce) and (specially) nitrate, since the NO3- concentrations in spinach and lettuce reported in the bibliography are well higher. The authors have not specified the number of samples analyzed and their characteristics (origin, type, etc.). Consequently the statistical analysis is not possible and the significance of these results is not high. We would like to respond below as follows:1. We have revised and added the references according with previous study.2. We have made the validation in this research accordingly as the reviewer required. After that, we have compared the result of the previous study with my research. In this research, validation was done, namely recovery. In the previous article before revision, this was not included in the article. According to the referee’s comments, we have corrected and added validation in this study. The recovery shows good accuracy. The recovery percentage meets the requirement between 80%-120%.3. We have specified the number of samples analyzed and their characteristics in the methods. About the statistical analysis, its possible because we have analyzed and have significant difference p < 0.05.For detail and further revision, we also sent our revised manuscript to F1000 editors.Hopefully, it may be clear now.Best regards,Rikson Siburian" } ] }, { "id": "43056", "date": "26 Mar 2019", "name": "Barbora Piknova", "expertise": [ "Reviewer Expertise nitric oxide metabolic pathway" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nVegetables and processed meat are the main dietary sources of inorganic nitrate. Nitrate itself has no know direct physiological effects. However, the mammalian body can convert nitrate into nitrite and nitric oxide (NO) in a two-step process either by mammalian nitrate and nitrite reductases or by the help of the oral and gut microbiome. It had been well documented over the past decade that NO and nitrite have a wide array of beneficial effects on cardiovascular health, such as lowering blood pressure, and improves the ability to exercise by decreasing the oxygen cost of exercise and/or by improving the blood flow into contracting muscle - for a review about the effect of nitrate and nitrite on exercise performance see Jones et al., 20181.\nI believe these 2 points needs to be addressed in the present manuscript before being accepted for indexing:\n\n1. The manuscript by Cintya et al. contains useful information about the amounts of nitrite/nitrate and vitamin C in various vegetables and on the effect of storage on the amount of these ions and vitamin C. However, it seems that the general message the authors try to convey in their abstract and introduction is that nitrite is a carcinogen and should be avoided. This idea had long been proven wrong and the message should be modified. Nitrate and nitrite are an integral part of a healthy diet and their importance for cardiovascular health is not questionable (Bryan et al., 20122 and Lundberg and Weitzberg, 20133); evidence for nitrate/nitrite-caused cancer is not convincing (Eichholzer and Gutzwiller, 19984 and Forman et al., 19855). As some case studies show, nitrite poisoning is usually a result of sodium nitrite overdose when used as meat-preserving salt (Jones et al., 20181) or when sodium nitrite is used instead of table salt (sodium chloride) by mistake (Lee et al., 20176).\n\n2. While the finding that storage affects levels of nitrate/nitrite is important, there is no mention about the reasons. Even if the reasons may be as trivial as loss of water during storage, it still should be mentioned.", "responses": [] } ]	1	https://f1000research.com/articles/7-1899
https://f1000research.com/articles/7-1645/v1	16 Oct 18	{ "type": "Research Article", "title": "Relationship between Candida albicans and Streptococcus mutans in early childhood caries, evaluated by quantitative PCR", "authors": [ "Endang W. Bachtiar", "Boy M. Bachtiar", "Endang W. Bachtiar" ], "abstract": "Background: The aim of this study was to analyze the synergistic relationship between Candida albicans and Streptococcus mutans in children with early childhood caries (ECC) experience. Methods: Dental plaque and unstimulated saliva samples were taken from 30 subjects aged 3-5 years old, half with (n=15, dmft > 4) and half without (n=15) ECC. The abundance of C. albicans and S. mutans and relative to total bacteria load were quantify by real-time PCR (qPCR). This method was also employed to investigate the mRNA expression of glycosyltransferase (gtfB) gene in dental plaque. Student’s t-test and Pearson’s correlation were used to perform statistical analysis. Results: Within the ECC group, the quantity of both microorganisms were higher in the saliva than in dental plaque. The ratio of C. albicans to total bacteria was higher in saliva than in plaque samples (p < 0.05). We observed the opposite for S. mutans (p < 0.05). The different value of C. albicans and S. mutans in saliva was positively correlated, and negatively correlated in dental plaque. Transcription level of S. mutans gtfB showed a positive correlation with C. albicans concentration in dental plaque. Conclusion: C. albicans has a positive correlation with cariogenic traits of S. mutans in ECC-related biofilm of young children.", "keywords": [ "Early childhood caries", "C.albicans/ S. mutans", "Saliva", "Dental plaque", "qPCR", "Indonesian" ], "content": "Introduction\n\nEarly childhood caries (ECC) remain the most common childhood oral health problem, globally1, and Streptococcus mutans has been known for its important role in ECC development2,3. However, in recent years, Candida albicans has frequently been linked with its synergistic relationship with S. mutans in dental plaque recovered from children with ECC4,5. Consequently, many studies using different methods have been conducted to identify, quantify, and explored the relationship of this fungus with S. mutans6–8. However, a controversial report does exist, where C. albicans tends to decrease the cariogenic traits of S. mutans in in vitro dual-species biofilm9. Therefore, the main purpose of this study was to validate the synergistic relationships between C. albicans and S. mutans, when growing in caries-related biofilm. For this reason, we requited preschool children with ECC experience, and we used qPCR since it is practice and reliable as a quantitative molecular tool of clinical oral samples10. The fungus and bacterium concentrations in saliva sample were used as control, and we compared the outcomes with those subjects noted as children with free caries (FC).\n\n\nMethods\n\nOral samples were collected from 30 requited preschool children (male and female, 3–5 years old), in two different location located near (about 30 km) to Jakarta, the capital city of Indonesia. The diagnostic of ECC referred to the criteria provided by the American Academy of Pediatric Dentistry, as previously reported11. The preschool children were recruited to get 15 subjects for each group. Thus, in this study, children were categorized into two different groups; children without any history of caries, including white-spot lesions, (caries free; CF group) and ECC group with decay-missing-filled teeth (dmft) index >4. To be included as subjects in this study, the children were required to be free of symptomatic oral candidiasis, have the absence of any medication therapy during the one month before this study, and have not worn any intraoral appliances. Before oral samples collection, written permission (informed consent) for children to participate was obtained from parents or guardians, according to the guidance provided by the Ethics Committee of Faculty of Dentistry, Universitas Indonesia.\n\nSamples from supragingival plaque, obtained from the selected teeth deciduous (either molar or incisive) were isolated with sterile cotton rolls and pure cotton buds. Samples from the ECC group was obtained by gathering carious biofilm around the affected enamel12, including dentin, as the fungus does not invade carious human dentin13. For the FC group, samples were obtained from enamel in clinically sound gingival areas. In each group, samples collected from molar or incisor, upper or lower teeth were not separated. Therefore, the obtained plaques were pooled to give a single sample for each subject and put in a microcentrifuge tube containing 1 ml PBS (pH 7.4). Unstimulated saliva was collected from all subjects on the same occasion and immediately after the plaque collection, by spitting the saliva into sterile Falcon plastic tubes. The minimum volume of collected saliva was 0.5 ml.\n\nSaliva and plaque samples were immediately cold-transported to the laboratory. For saliva samples, after centrifugation, the sediments were washed three times with 0.5 ml sterile milli Q water between each centrifugation and kept in -80°C until use. Similarly, plaque samples from caries-free or those with ECC were cold-transported to the laboratory and processed as mentioned above, then stored at -80°C until use.\n\nBacterial/fungal DNA was obtained by centrifugation each sample in the microtube, using Trizol reagent (Sigma-Aldrich, Dorset, UK). Extracted DNA sample was kept at -20°C after determining the concentration and the quality by Qubit assay reagents (Invitrogen, Carlsbad, CA). The genomic DNA samples were dissolved in Tris-EDTA (TE) buffer and kept in -20°C freeze until used. Further, the DNA samples were quantified through a qPCR reaction with universal primers for the 16S rRNA genes and C. albicans/S. mutans-specific primers as shown in Table 1. For PCR-quantification, each sample was run in triplicate on an ABI StepOnePlus Real-Time PCR System with SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA, USA) according to the manufacturer’s protocol.\n\nThe PCR conditions were run in a final reaction volume of 10 µl, composed of 50 ng of sample DNA and 1 µM of each primer (Table 1), with thermal cycling condition consisted of a 10 min initial denaturation at 95°C, followed by 40 cycles of 95°C for 15 s and 65°C for 60 s. The cycle threshold (Ct) were determined automatically by the instrument, and a dissociation curve of the amplified fragment set as follow; 95°C for 15 s, 60°C for 60 s, and 95°C for 15 s.\n\nEstimating the amount of genomic DNA of both microorganisms tested was determined by constructing standard curves with r2 values for both organisms tested as well as total bacteria (Figure 1A–C). To do this, we used a 10-fold serial dilution of extracted fungal and bacterial genomic DNA from overnight cultured of C. albicans ATCC 10231, S. mutans Xc, and Escherichia coli JM 107, respectively. The number (CFU/ml) assessed by plating culture dilutions on sabouraud agar, tryptone-yeast extract cysteine with sucrose and bacitracin (TYCSB) agar and Luria Bertani (LB) broth for C. albicans16, S. mutans17, and E. coli18, respectively. The same strains were used as positive control for qPCR. Therefore, quantification of C. albicans and S. mutans from plaques and saliva achieved by plotting the Ct values against the log of the respective standard curve. In this study, the ratio of C. albicans or S. mutans in the microbial community, in each sample, was determined as each microorganism proportion to total bacteria.\n\nStandard curves of total bacteria (A), C. albicans (B) and S. mutans (C). Also shown are melt curve profiles and melting temperatures for total bacteria (D), C. albicans (E), and S. mutans (F).\n\nFor both C. albicans and S. mutans, the detection limit by qPCR method was determined according to the limit of quantification (LOQ), obtained by the highest dilution of the template of the standard curve. When the Ct value of samples was higher than the LOQ, they would be considered positive, but their melting curve profile should be the same as those of the standards included when running the qPCR.\n\nRNA isolation, purification, and reverse transcription of cDNA were performed similarly those in the previous study19. Platinum SYBR Green qPCR SuperMix-UDG (Invitrogen Life Technologies, Carlsbad, CA, USA), passive reference (ROX, Invitrogen), and S. mutans gtfB primers (Table 1), as well as 1 µg of cDNA, were used to quantify the cDNA, and non-transcribed RNA samples were used as control for genomic DNA contamination. The qPCR reaction was run on a similar machine as stated above with cycling conditions consisted of a 10 min initial denaturation at 95°C followed by 40 PCR cycles of 15 s at 95°C, and 60°C for 1 min. The formula of fold change 2-ΔΔCt was used to calculate S. mutans gtfB gene expression that was normalized to the 16S rRNA, a well-established housekeeping gene20, and gtfB expressed in dental plaque of FC group was set to be the control.\n\nThe variables for quantification, proportion, and the mean quantitative gene expression were assessed using Student’s t-test, while Pearson’s correlation two-tailed test was used to depict a linear association. Microsoft Excel software was used to perform statistical analysis, and a p-value < 0.05 was considered significant.\n\n\nResults\n\nStandard curves were used to determine the corresponding number of microorganism tested while melting peaks were used to assess the specificity of the amplicon using saliva and plaque samples (Figure 1D–F).\n\nIn general, this study showed that in all saliva and plaque samples, from either FC or ECC children, both C. albicans and S. mutans were present. The quantification (log DNA copies) and proportion (% to total bacteria) of C. albicans and S. mutans in saliva, as well as plaque samples, are presented in Figure 2. Comparatively, it was observed, in either sample tested, a significantly higher number of both microorganisms in ECC children was found more than in those of the FC children (p< 0.05). However, in either group, the quantitative level of C. albicans, in the saliva sample was found to be significantly lower than those of S. mutans (p< 0.05). When comparing plaque and saliva samples within ECC children, we observed that the load (log DNA copies) of either microorganism in plaque was significantly lower than that in saliva (p< 0.05, Figure 2A).\n\nMean and standard error of absolute numbers (A) and ratios (B) of C. albicans and S. mutans detected within the same saliva (Sa) and dental plaque (Dp) samples. CF, caries-free; ECC, early childhood caries. *p < 0.05.\n\nFurthermore, we compared the proportion of C. albicans and S. mutans DNA relative to total bacterial DNA in saliva and dental plaque samples, in FC and ECC children. Within ECC children, there was a significantly higher proportion ratio (Ca/Tb) of C. albicans in saliva samples (35.5%), than that in plaque samples (13.5%) (p < 0.05). For FC children the ratio was not statistically different (saliva, 8% and plaque, 5.3%) (Figure 2B). Similar analysis was carried out for the S. mutans proportion ratio (Sm/Tb). The result showed a different trend, with a significantly higher proportion of S. mutans in plaque (99%) than that in the saliva (62%) of ECC children (Figure 2B). The result also showed the proportion of this bacterium to total bacteria in saliva and plaque samples showed a significant difference between ECC and FC children (p < 0.05). Interestingly, there was a trend within dental plaque sample in ECC children, S. mutans DNA increased most with increased of C. albicans’ DNA (Figure 2B).\n\nWe further evaluated the possible linear relationship of C. albicans and S. mutans load or their proportion and ECC experience in the subjects. Pearson correlations coefficient analysis revealed that the association between the numbers of these two microorganisms in saliva was moderate positively significant (r = 0.1, p < 0.05). On the contrary, a weak negative not substantial (r = 0.03, p > 0.05) between the decreasing number of C. albicans and the quantity of S. mutans in plaque samples observed in the ECC group (Figure 3A and B).\n\nCorrelation between C. albican and S. mutans loads in saliva (A) and in dental plaque (B), in the same subjects. Each circle depicts the value of C. albicans (Y-axis) and S. mutans (X-axis) in log CFU/ml for each subject.\n\nTo confirm all the above results, we selected the gtfB gene, which has been reported to be mostly involved in the synergistic relationship between C. albicans and S. mutans in biofilm development7, and compared its expression in each dental plaque of children tested. The qPCR result showed that level of mRNA gtfB was induced approximately 4.5-fold in ECC-derived dental plaque samples, and it was a significant difference compared to transcription level of gtfB mRNA in dental plaque sampled from children with FC (p< 0.05) (Figure 4A). Also, gtfB transcription levels and the amount of C. albicans and S. mutans (CFU/ml) in dental plaque of children with ECC showed moderate (r = 0.2) and strong (r = 0.9) positive correlations, respectively, which was statistically significant (p< 0.05, Figure 3B and C).\n\n(A) The fold regulation of gtfB cDNA that was normalized to the amount of cDNA of 16S rRNA. (B, C) the correlation between gtfB transcription rate and the amount of C. albicans and S. mutans, respectively.\n\n\nDiscussion\n\nNumerous studies7,8,21 have supported the idea that ECC may be better understood by focusing on the effect of the functional relationship between species within consortia instead of individual pathogens. This study focused on the relationship between C. albicans and S. mutans, as these oral microorganisms are frequently detected in the plaque of children with ECC12,22. Since ECC may indicate a vast proliferation of the cariogenic microorganism, we sought to evaluate the extent to which the amount of C. albicans might correlate to the ECC experience. To do this, we used qPCR. This method enabled the quantification of the targeted microorganisms’ genomic DNA from oral samples23. Thus, we examined the similarities and differences by comparing the amount and ratio to total bacteria of each microorganism in saliva and dental plaque samples, and verified their correlation with the occurrence of ECC. In general, we observed that the oral cavities of preschool children in this study, with or without ECC, are colonized by yeast (C. albicans) and bacteria, as represented by S. mutans. Overall, the fungus presence was always simultaneously detected with S. mutans, although children with ECC had a higher amount of C. albicans and S. mutans in their oral cavity, compared to those children with FC. As expected, in addition to saliva, these oral microbiotas commonly exist together in an ECC-related biofilm.\n\nOur observation is in accordance with results from other studies, which found that in addition to S. mutans as a specific cariogenic bacterium24,25, C. albicans can be part of the dental lesion26,27. Moreover, previous studies showed strong synergism when C. albicans and S. mutans co-existed in biofilm, suggesting that this co-existence enhanced their virulence7,8,28. However, our data showed, the fungus was detected at lower levels in dental plaque, compared to S. mutans. This support the previous in vitro study29, which found that the presence of C. albicans might favor the extensive colonization of S. mutans in dental biofilm.\n\nThe causes to generate site specificity bacteria proportion are believed to include local sucrose concentration in the oral cavity30,31. In addition to sucrose32, many factors may link to the presence of C. albicans in children oral cavity. These include infection at birth, baby’s feeding bottles, infected pacifiers, and carious teeth12. We speculate that a high cariogenic diet might influence the interaction between C. albicans and S. mutans in these children tested. In turns, it becomes critical for ECC, since the presence of sucrose in the children oral cavity may lead to the ability of this species to grow within structured microbial biofilm33,34. Further studies are therefore necessary.\n\nThe presence of yeast and bacteria is one of the local factors that contributes to the etiology of ECC7. To obtain an overall insight into the impact of simultaneous participation of C. albicans and S. mutans when detecting together in each sample tested, we compared the percent proportion of C. albicans or S. mutans relative to total bacteria. As expected, the proportion of C. albicans was higher in saliva than in carious plaque among children with ECC. Other studies have reported this phenomenon, where Candida species were frequently isolated more, qualitatively and quantitatively, from saliva than from dental plaque35 and subgingival samples36. During in saliva, this fungus might act as a bridge for oral bacteria to adhere to a mucosal surface, a mechanism that may protect this bacterium from being removed by salivary flow and swallowing21. On the other side, biofilm formation is vital for C. albicans to survive as a pathogen, which involves attachment, colonization, and development of structural biofilm integrity composed of yeast and hypha37,38. Our data illustrate that co-adhesion between C. albicans and S. mutans in cariogenic biofilm is one mechanism by which the fungus, in yeast form, survives in the oral cavity39,40. Although the fungus morphology was not observed in this study, it has been reported by other studied that hypha morphology is not crucial for the C. albicans–S. mutans relationship when they grow in multispecies dental biofilms41–43. Therefore, in addition to the morphology, both the number and proportion of salivary C. albicans influence the fungal–bacterial relationships, which further increases the risk of caries. Additionally, the flushing effect of saliva, as part of innate defense mechanism, might contribute to decrease Candida adherence to the oral surface44, including tooth surface. Thus, the quality and quantity of saliva have an essential role in maintaining fungus behavior, as commensal or pathogen45.\n\nWe observed that the proportion of S. mutans was higher in dental plaque than in saliva, and there was a tendency for the percentage of C. albicans to be lower in carious plaque, where the proportion of S. mutans increased. This observation indicates that S. mutans has an active role in orchestrating the development of cariogenic biofilms46,47. This species has an essential part in attenuating the virulence of the fungus48 by interfering with the fungus transition, from yeast to hypha form when these oral microflorae interact and grow in biofilm49,50. This result further supported by the data of correlation analysis, in which the fungus-bacterium concentration in dental plaque sample showed a negative association, although a positive correlation was found in saliva sediment. This suggests that the ECC rate may not be connected to the quantity of C. albicans involves.\n\nOne of the mediators for the synergistic relationship between C. albicans and S. mutans is the streptococcal GtfB enzyme7,32. Our finding indicates that cariogenic biofilm developed in ECC children accompanied by the increased transcription level of gtfB mRNA and enhance of S. mutans growth in dental plaque derived from children with ECC. In a clinical situation, this observation is relevant, since GtfB is the enzyme that synthesizes glucan polymers from sucrose6. Thus, result of this study suggests that a high sucrose concentration, which is critical for the development of dental caries, might exist in the children oral cavity. Future studies are thus recommended.\n\nCollectively, this study indicates that although the proportion of C. albicans was less in ECC-associated biofilm, it may support facilitation of a fungal-bacterial synergistic relationship. Yeast cells could be used by S. mutans to promote fitness and the bacterial survival, as shown by enhanced transcription level of gtfB, which reflects that more extracellular polysaccharides were produced to promote the fungal-bacterium relationship in caries-related biofilm7.\n\nThe results of the present study cannot explain the reason for such an association. However, at least this study provides information on ECC experience among preschool children. The presence of C. albicans in dental plaque and saliva could be merely an indicator of oral health conditions and the high carbohydrate intake among the young children selected in this study, which might confer a survival advantage for C. albicans and it is favorable for ECC development. More studies are needed.\n\nThere is some limitation in this study. First, the primary disadvantage of qPCR used in this study is its inability to separate and quantify the viable from nonviable cells. This technique may result in false positives or an overly high estimation, as all DNA extracted from life or dead C. albicans, or S. mutans cell will be amplified. Since the number of viable cells is especially significant for diagnosing and monitoring disease, adding cell viability information to qPCR-based diagnostics should be considered. Second, the number of children involved in this study was small (15 subjects per group) because of difficulty in sample collection, primarily to obtain the plaque on the dentin surface.\n\n\nConclusions\n\nThis study shows that C. albicans contributes to increasing concentration of S. mutans by inducing the expression of gtfB mRNA in ECC-related biofilm. Therefore, results from this study would be useful as a starting point to consider C. albicans, as a potential target in prevention programs to reduce the high rates of ECC in individuals or groups of young children.\n\n\nData availability\n\nDataset 1. The raw data associated with this study. Data are grouped according to the figure with which they are associated. DOI: htpps://doi.org/10.5256/f1000research.16275.d22130451.", "appendix": "Grant information\n\nThis study was supported by a grant from the Ministry of Research, Technology, and Higher Education, The Republic of Indonesia 2017.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nWe thank Anissa, Vivi, and Asti for the laboratory work at Oral Science Research Centre, Faculty of Dentistry, Universitas Indonesia. We also thank for Arif, Tika, Inge, and Hanli for collecting oral samples.\n\n\nReferences\n\nPetersen PE, Bourgeois D, Ogawa H, et al.: The global burden of oral diseases and risks to oral health. Bull World Health Organ. 2005; 83(9): 661–669. PubMed Abstract \| Free Full Text\n\nTanzer JM, Baranowski LK, Rogers JD, et al.: Oral colonization and cariogenicity of Streptococcus gordonii in specific pathogen-free TAN:SPFOM(OM)BR rats consuming starch or sucrose diets. Arch Oral Biol. 2001; 46(4): 323–333. PubMed Abstract \| Publisher Full Text\n\nKoga T, Oho T, Shimazaki Y, et al.: Immunization against dental caries. Vaccine. 2002; 20(16): 2027–2044. PubMed Abstract \| Publisher Full Text\n\nRaja M, Hannan A, Ali K: Association of oral candidal carriage with dental caries in children. Caries Res. 2010; 44(3): 272–276. PubMed Abstract \| Publisher Full Text\n\nEllepola K, Liu Y, Cao T, et al.: Bacterial GtfB Augments Candida albicans Accumulation in Cross-Kingdom Biofilms. J Dent Res. 2017; 96(10): 1129–1135. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKuramitsu HK: Virulence properties of oral bacteria: impact of molecular biology. Curr Issues Mol Biol. 2001; 3(2): 35–36. PubMed Abstract\n\nFalsetta ML, Klein MI, Colonne PM, et al.: Symbiotic relationship between Streptococcus mutans and Candida albicans synergizes virulence of plaque biofilms in vivo. Infect Immun. 2014; 82(5): 1968–1981. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKoo H, Bowen WH: Candida albicans and Streptococcus mutans: a potential synergistic alliance to cause virulent tooth decay in children. Future Microbiol. 2014; 9(12): 1295–1297. PubMed Abstract \| Publisher Full Text\n\nWillems HM, Kos K, Jabra-Rizk MA, et al.: Candida albicans in oral biofilms could prevent caries. Pathog Dis. 2016; 74(5): pii: ftw039. PubMed Abstract \| Publisher Full Text\n\nLyons SR, Griffen AL, Leys EJ: Quantitative real-time PCR for Porphyromonas gingivalis and total bacteria. J Clin Microbiol. 2000; 38(6): 2362–2365. PubMed Abstract \| Free Full Text\n\nDrury TF, Horowitz AM, Ismail AI, et al.: Diagnosing and reporting early childhood caries for research purposes. A report of a workshop sponsored by the National Institute of Dental and Craniofacial Research, the Health Resources and Services Administration, and the Health Care Financing Administration. J Public Health Dent. 1999; 59(3): 192–197. PubMed Abstract \| Publisher Full Text\n\nde Carvalho FG, Silva DS, Hebling J, et al.: Presence of mutans streptococci and Candida spp. in dental plaque/dentine of carious teeth and early childhood caries. Arch Oral Biol. 2006; 51(11): 1024–1028. PubMed Abstract \| Publisher Full Text\n\nMaijala M, Rautemaa R, Järvensivu A, et al.: Candida albicans does not invade carious human dentine. Oral Dis. 2007; 13(3): 279–284. PubMed Abstract \| Publisher Full Text\n\nFeldman M, Ginsburg I, Al-Quntar A, et al.: Thiazolidinedione-8 Alters Symbiotic Relationship in C. albicans-S. mutans Dual Species Biofilm. Front Microbiol. 2016; 7: 140. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMaeda H, Fujimoto C, Haruki Y, et al.: Quantitative real-time PCR using TaqMan and SYBR Green for Actinobacillus actinomycetemcomitans, Porphyromonas gingivalis, Prevotella intermedia, tetQ gene and total bacteria. FEMS Immunol Med Microbiol. 2003; 39(1): 81–86. PubMed Abstract \| Publisher Full Text\n\nBachtiar EW, Bachtiar BM, Jarosz LM, et al.: AI-2 of Aggregatibacter actinomycetemcomitans inhibits Candida albicans biofilm formation. Front Cell Infect Microbiol. 2014; 4: 94. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVan Palenstein Helderman WH, Ijsseldijk M, Huis in 't Veld JH: A selective medium for the two major subgroups of the bacterium Streptococcus mutans isolated from human dental plaque and saliva. Arch Oral Biol. 1983; 28(7): 599–603. PubMed Abstract \| Publisher Full Text\n\nBachtiar BM, Coloe PJ, Fry BN: Knockout mutagenesis of the kpsE gene of Campylobacter jejuni 81116 and its involvement in bacterium-host interactions. FEMS Immunol Med Microbiol. 2007; 49(1): 149–154. PubMed Abstract \| Publisher Full Text\n\nBachtiar BM, Bachtiar EW: Proinflammatory MG-63 cells response infection with Enterococcus faecalis cps2 evaluated by the expression of TLR-2, IL-1β, and iNOS mRNA. BMC Res Notes. 2017; 10(1): 401. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStipp RN, Gonçalves RB, Höfling JF, et al.: Transcriptional analysis of gtfB, gtfC, and gbpB and their putative response regulators in several isolates of Streptococcus mutans. Oral Microbiol Immunol. 2008; 23(6): 466–473. PubMed Abstract \| Publisher Full Text\n\nKrom BP, Kidwai S, Ten Cate JM: Candida and other fungal species: forgotten players of healthy oral microbiota. J Dent Res. 2014; 93(5): 445–451. PubMed Abstract \| Publisher Full Text\n\nQiu R, Li W, Lin Y, et al.: Genotypic diversity and cariogenicity of Candida albicans from children with early childhood caries and caries-free children. BMC oral health. 2015; 15(1): 144. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKirakodu SS, Govindaswami M, Novak MJ, et al.: Optimizing qPCR for the Quantification of Periodontal Pathogens in a Complex Plaque Biofilm. Open Dent J. 2008; 2: 49–55. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBowen WH, Koo H: Biology of Streptococcus mutans-derived glucosyltransferases: role in extracellular matrix formation of cariogenic biofilms. Caries Res. 2011; 45(1): 69–86. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKrzyściak W, Jurczak A, Kościelniak D, et al.: The virulence of Streptococcus mutans and the ability to form biofilms. Eur J Clin Microbiol Infect Dis. 2014; 33(4): 499–515. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nXiao J, Moon Y, Li L, et al.: Candida albicans Carriage in Children with Severe Early Childhood Caries (S-ECC) and Maternal Relatedness. PLoS One. 2016; 11(10): e0164242. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYang XQ, Zhang Q, Lu LY, et al.: Genotypic distribution of Candida albicans in dental biofilm of Chinese children associated with severe early childhood caries. Arch Oral Biol. 2012; 57(8): 1048–1053. PubMed Abstract \| Publisher Full Text\n\nSztajer H, Szafranski SP, Tomasch J, et al.: Cross-feeding and interkingdom communication in dual-species biofilms of Streptococcus mutans and Candida albicans. ISME J. 2014; 8(11): 2256–2271. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBarbieri DdS'AV VV, Fraiz FC, Lavoranti OJ, et al.: Analysis of the in vitro adherence of Streprococcus mutans and Candida albicans. Braz J Microbiol. 2007; 38(4): 624–631. Publisher Full Text\n\nDawes C, MacPherson LM: The distribution of saliva and sucrose around the mouth during the use of chewing gum and the implications for the site-specificity of caries and calculus deposition. J Dent Res. 1993; 72(5): 852–857. PubMed Abstract \| Publisher Full Text\n\nHallett KB, O'Rourke PK: Early childhood caries and infant feeding practice. Community Dent Health. 2002; 19(4): 237–242. PubMed Abstract\n\nGregoire S, Xiao J, Silva BB, et al.: Role of glucosyltransferase B in interactions of Candida albicans with Streptococcus mutans and with an experimental pellicle on hydroxyapatite surfaces. Appl Environ Microbiol. 2011; 77(18): 6357–6367. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nUppuluri P, Pierce CG, López-Ribot JL: Candida albicans biofilm formation and its clinical consequences. Future Microbiol. 2009; 4(10): 1235–1237. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMetwalli KH, Khan SA, Krom BP, et al.: Streptococcus mutans, Candida albicans, and the human mouth: a sticky situation. PLoS Pathog. 2013; 9(10): e1003616. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSalam MA, Senpuku H, Nomura Y, et al.: Isolation of opportunistic pathogens in dental plaque, saliva and tonsil samples from elderly. Jpn J Infect Dis. 2001; 54(5): 193–195. PubMed Abstract\n\nJabra-Rizk MA, Ferreira SM, Sabet M, et al.: Recovery of Candida dubliniensis and other yeasts from human immunodeficiency virus-associated periodontal lesions. J Clin Microbiol. 2001; 39(12): 4520–4522. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFinkel JS, Mitchell AP: Genetic control of Candida albicans biofilm development. Nat Rev Microbiol. 2011; 9(2): 109–118. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRamage G, Saville SP, Thomas DP, et al.: Candida biofilms: an update. Eukaryot Cell. 2005; 4(4): 633–638. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNikawa H, Egusa H, Makihira S, et al.: Alteration of the coadherence of Candida albicans with oral bacteria by dietary sugars. Oral Microbiol Immunol. 2001; 16(5): 279–283. PubMed Abstract \| Publisher Full Text\n\nThein ZM, Samaranayake YH, Samaranayake LP: Effect of oral bacteria on growth and survival of Candida albicans biofilms. Arch Oral Biol. 2006; 51(8): 672–680. PubMed Abstract \| Publisher Full Text\n\nHwang G, Liu Y, Kim D, et al.: Candida albicans mannans mediate Streptococcus mutans exoenzyme GtfB binding to modulate cross-kingdom biofilm development in vivo. PLoS Pathog. 2017; 13(6): e1006407. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPereira-Cenci T, Deng DM, Kraneveld EA, et al.: The effect of Streptococcus mutans and Candida glabrata on Candida albicans biofilms formed on different surfaces. Arch Oral Biol. 2008; 53(8): 755–764. PubMed Abstract \| Publisher Full Text\n\nJoyner PM, Liu J, Zhang Z, et al.: Mutanobactin A from the human oral pathogen Streptococcus mutans is a cross-kingdom regulator of the yeast-mycelium transition. Org Biomol Chem. 2010; 8(24): 5486–5489. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRamage G, Vandewalle K, Wickes BL, et al.: Characteristics of biofilm formation by Candida albicans. Rev Iberoam Micol. 2001; 18(4): 163–170. PubMed Abstract\n\nEpstein JB, Truelove EL, Izutzu KT: Oral candidiasis: pathogenesis and host defense. Rev Infect Dis. 1984; 6(1): 96–106. PubMed Abstract \| Publisher Full Text\n\nKanasi E, Johansson I, Lu SC, et al.: Microbial risk markers for childhood caries in pediatricians' offices. J Dent Res. 2010; 89(4): 378–383. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTanner AC, Kent RL Jr, Holgerson PL, et al.: Microbiota of severe early childhood caries before and after therapy. J Dent Res. 2011; 90(11): 1298–1305. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBarbosa JO, Rossoni RD, Vilela SF, et al.: Streptococcus mutans Can Modulate Biofilm Formation and Attenuate the Virulence of Candida albicans. PLoS One. 2016; 11(3): e0150457. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJarosz LM, Deng DM, van der Mei HC, et al.: Streptococcus mutans competence-stimulating peptide inhibits Candida albicans hypha formation. Eukaryot Cell. 2009; 8(11): 1658–1664. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVilchez R, Lemme A, Ballhausen B, et al.: Streptococcus mutans inhibits Candida albicans hyphal formation by the fatty acid signaling molecule trans-2-decenoic acid (SDSF). Chembiochem. 2010; 11(11): 1552–1562. PubMed Abstract \| Publisher Full Text\n\nBachtiar EW, Bachtiar BM: Dataset 1 in: Candida albicans and Streptococcus mutans relationship in early childhood caries, evaluated by quantitative real time-PCR. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16275.d221304" }	[ { "id": "39491", "date": "31 Oct 2018", "name": "Zamirah Zainal-Abidin", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nMethods:\nThe authors did not mention how the examiners were calibrated, in order to reduce the inter- and intra-examiner variability when examining the preschool children, before grouping them into the early childhood caries and caries free groups.\n\nThe authors may have been able to collect more information on the oral hygiene habits or dietary intake of the subjects, before taking the oral samples.\n\nResults:\nIn the caption for Figure 2, the symbol p<0.05 should be indicated clearly. For Figure 2A, it is to show that there is a statistically-significant difference between the C. albicans in saliva and dental plaque in the ECC group, and a statistically-significant difference between the S. mutans in saliva and dental plaque in the ECC group. The use of top square bracket ( ⎴ ) with an asterisk would be better to indicate which parameters are statistically compared.\n\nFor Figure 4A, the use of top square bracket ( ⎴ ) with an asterisk would be better to indicate the parameters which are statistically compared.\n\nDiscussion:\nIn paragraphs 3 and 7, the authors speculated that a high cariogenic/high sucrose concentration diet might influence the interaction between C. albicans and S. mutans in the tested subjects. The authors may have been able to draw a better suggestion if the information on the oral hygiene habits or diet is obtained during the recruitment of the subjects (See my comment in Methods (2)).\n\nIn paragraphs 3, 7 and 9, the authors could have suggested what further studies are necessary to address the suggested theories/findings.\n\nGeneral comments:\nThe paper has a few grammar and language issues, in the Abstract and Methods sections.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4204", "date": "08 Nov 2018", "name": "Endang Bachtiar", "role": "Author Response", "response": "Dear editor and reviewer,I am pleased to resubmit for publication the revised version of Relationship between Candida albicans and Streptococcus mutans in early childhood caries, evaluated by quantitative PCR.We appreciate all the insightful comments provided by the reviewer. Based on the reviewer’s guidance, includes a number of positive changes, we endeavoured to improve the fit of the paper with the journal.We hope that these revisions improve the paper such that the editor and reviewer now deem it worthy of publication in F1000 Research. Thank you for taking the time to help us improve the paper. Sincerely,Authors : 1. Endang W Bachtiar 2. Boy M Bachtiar Author’s response to reviewer commentsComment provided by Reviewer #1 Methods:Reviewer’s comment:The authors did not mention how the examiners were calibrated, in order to reduce the inter- and intra-examiner variability when examining the preschool children, before grouping them into the early childhood caries and caries-free groups. Author’s response: We thank the reviewer for this correction. The problem has been fixed.As suggested by the reviewer, we have added sentence within material and method as follow: “two weeks prior to collect to clinical samples, the examiners were calibrated and trained by providing with the manual describing study protocol and guidance regarding the examination of preschool children. Therefore, only those trained-examiners evaluated the preschool children”. Reviewer’s comment:The authors may have been able to collect more information on the oral hygiene habits or dietary intake of the subjects, before taking the oral samples. Author’s response: In this study, we used a questionnaire regarding the parent’s/ guardian’s perception concerning children’s general and oral health, the risk factor for caries, dietary intake, and access to dental care. We noticed, in general, most of the responders believed that dental decay is a natural phenomenon. However, considering the complexity of the issue studied in this study, it is not the authors’ intent to include oral hygiene habits or dietary intake of the subjects (preschool children). Therefore, the related-data were not included in the current study. Reviewer’s comment: In the caption for Figure 2, the symbol p<0.05 should be indicated clearly. For Figure 2A, it is to show that there is a statistically significant difference between the C. albicans in saliva and dental plaque in the ECC group, and a statistically significant difference between the S. mutans in saliva and dental plaque in the ECC group. The use of top square bracket ( ⎴ ) with an asterisk would be better to indicate which parameters are statistically compared. For Figure 4A, the use of top square bracket ( ⎴ ) with an asterisk would be better to indicate the parameters which are statistically compared.Author’s response:According to the suggestions provided by Reviewer#1, we have corrected the old figures. The top square bracket has been added in Fig. 2A to indicate a statistically significant difference as suggested. For the Fig. 4A, we have added the top square bracket with an asterisk to indicate a significant difference between the parameters compared. Thank you. Discussion:Reviewer’s comment:In paragraphs 3 and 7, the authors speculated that a high cariogenic/high sucrose concentration diet might influence the interaction between C. albicans and S. mutans in the tested subjects. The authors may have been able to draw a better suggestion if the information on the oral hygiene habits or diet is obtained during the recruitment of the subjects (See my comment in Method (2)).Author’s response:As mentioned above (response to the reviewer’s comment in methods (2); it is not the authors’ intent to include the oral hygiene habits or diet of the preschool children in this study. This is because the information regarding oral hygiene habit or diet, provided by the accompanying guardian are difficult to be understood. Therefore, it is hard to interpret oral hygiene habit or diet data. Reviewer’s comment: In paragraphs 3, 7 and 9, the authors could have suggested what further studies are necessary to address the suggested theories/findings. Author’s responses: According to reviewer suggestion, we have added a better suggestion at the end of paragraph-3, as follow; “Further studies to obtain data regarding cariogenic diet are therefore necessary”. Additionally, we have added a sentence at the end of paragraph 7, as follow: “More studies regarding the involvement of gtfB expression, and how it relates to oral health conditions as well as cariogenic intake in Indonesian preschool children, are needed. At the end of paragraph 9 (conclusion), as suggested by the reviewer, we added sentences; “Moreover, since the information regarding oral hygiene habit or diet, provided by the accompanying guardian are difficult to be understood, future studies involving preschool children may wish to involve examiners who are more experience in dealing with such parents/guardians when planning studies”. Thank you. Dear Editor,We are pleased to inform you that; 1/ for all the revision, as suggested by the reviewer, were showed in yellow-highlighting words, 2/ we revised the figure 2 and 4 to address the reviewer’s comments. Although the Fig. 3 is not included in reviewer’s comment, we revised the font, from the original (Arial) to new version (Times New Roman). Thus, all founds depicted in figures are similar. Thank you." } ] }, { "id": "39489", "date": "06 Nov 2018", "name": "Shahida Mohd-Said", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper reports an interesting and relevant study on ECC in children. The methods and contents of the papers are current, relevant and well-structured. However, some improvements could benefit the presentation including checking on grammar and typos, some revisions to sentence construction, highlights on the significance of this study, how the findings can be important to update of knowledge and current management of ECC, and comparison of data from Indonesia with other populations to perhaps illustrate novelty and significance to current available data.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4203", "date": "08 Nov 2018", "name": "Endang Bachtiar", "role": "Author Response", "response": "Dear reviewer,I am pleased to resubmit for publication the revised version of Relationship between Candida albicans and Streptococcus mutans in early childhood caries, evaluated by quantitative PCR.We appreciate all the insightful comments provided by the reviewer. Based on the reviewer’s guidance, includes a number of positive changes, we endeavoured to improve the fit of the paper with the journal.We hope that these revisions improve the paper such that the editor and reviewer now deem it worthy of publication in F1000 Research. Thank you for taking the time to help us improve the paper. Sincerely,Authors: Endang W Bachtiar and Boy M BachtiarResponses to reviewer #1:We have fixed all grammar, typos, and some revisions to sentence construction. Furthermore all responses to the suggestions have been added in the manuscript in the yellow highlight sentences." } ] } ]	1	https://f1000research.com/articles/7-1645
https://f1000research.com/articles/7-1898/v1	05 Dec 18	{ "type": "Research Article", "title": "The effect of osseodensification and different thread designs on the dental implant primary stability", "authors": [ "Abdullah Saleh Almutairi", "Maher Abdullatif Walid", "Mohamed Ahmed Alkhodary", "Abdullah Saleh Almutairi", "Maher Abdullatif Walid" ], "abstract": "Background: It is difficult to achieve good primary stability of dental implants in soft bone, such as that in the posterior maxillae. Osseodensification (OD) burs, working in a non-subtractive fashion, condense the implant osteotomy bone in lateral direction and increase in the bone implant contact. Also, dental implants with deeper threads, and decreased thread pitch can increase initial bone implant anchorage. Methods: This study utilized 48 custom-made machined surface dental implants that were 13 mm long, with a major diameter of 4.5 mm and a minor diameter of 3.5 mm, a thread pitch of 1 mm, a thread depth of 0.5 mm, and a 4 mm long cutting flute at the apex. The implants were divided into 4 groups, each group was made of 12 implants with a different thread design; V-shaped, trapezoid, buttress, and reverse buttress. The implants were inserted in 4-mm thick cancellous bone slices obtained from the head of Cow femur bone. The ostoetomies were prepared by conventional drilling and by OD drilling. Each inserted implant was then tested for primary stability using the Periotest. The Periotest values (PTVs) for the implant stability were tabulated and analyzed using a chi square test at significance level p< 0.05. Results: The results of this this study revealed no statistically significant difference between the Periotest readings for the implants in each category placed in either the OD or the regular osteotomies. However, it has been found that the implants placed in regular drilling ostoetomies had a significantly better primary stability than the implants placed in OD osteotomies. Conclusions: It was concluded that OD is not necessary in situations where there is bone of good quality and quantity.", "keywords": [ "Implant primary stability", "osseodensification", "implant thread designs", "Periotest." ], "content": "Introduction\n\nThe primary stability of dental implants depends on the quantity and quality of the available bone, the implant macro- and micro-design, the implant surface features, and the surgical technique used for creating the osteotomy. Conventionally, ostoetomies are created with bone-removing drills, and the last drill had a smaller diameter than that of the implant to ensure primary stability. However, this technique is only effective up to certain limits in soft bone, such as that in the posterior maxillae1.\n\nOsseodensification (OD) burs, working in a non-subtractive fashion, condense the implant osteotomy soft bone in lateral direction, leading to a greater bone volume and density, an increase in the bone implant contact, with subsequent increase in insertion torque levels, and reduction in micromotion2–6. However, it has been claimed that OD increases the implant bone bed density, but does not improve implant primary stability7.\n\nAnother maneuver to increase implant primary stability in a poor bone quality situation, is to use an implant with deeper threads, and decreased thread pitch, to increase initial bone implant anchorage. This principle can be applied to different dental implant thread designs; V-shaped, buttress, reverse-buttress, and trapezoid. However, each thread design is thought to give a varying degree of apical and lateral compression to the surrounding bone, which will produce a certain amount of osteocompression and primary stability8–13.\n\nClinically, the dental implant primary stability can be evaluated using several techniques, such as the amount of torque needed during insertion, or after insertion using the resonance frequency analysis technology implemented in the Osstell device, or the mechanical percussion principle used in the Periotest14–20.\n\nAlthough the OD may minimize the use of other more invasive techniques, such as ridge splitting21, sinus lifting22, and onlay bone grafting23, there has been a claim that the OD surgical technique may not be effective in improving the primary stability of dental implants7, and that the dental implant macro-design is not crucial for the implant primary stability so long as the surrounding bone is of good quantity and quality24. The aim of this study was to test the effect of both variables, the OD surgical technique and the implant macro-design, on the dental implant primary stability, using custom made dental implants with four different thread shapes, with the same thread pitch and depth, placed in two different types of ostoetomies prepared by the conventional and the OD technique, and evaluated using the Periotest.\n\n\nMethods\n\nThis study utilized 48 custom made machined surface dental implants that were 13 mm long, with a major diameter of 4.5 mm and a minor diameter of 3.5 mm. The implants were divided into four groups, each group made up of 12 implants with a different thread design: V-shaped, trapezoid, buttress, and reverse buttress (Figure 1). All the groups had the same thread pitch of 1 mm, a thread depth of 0.5 mm, and a 4 mm long cutting flute at the apex of the implants.\n\nThe thread shapes, from left to right, are V-shaped, buttress, reverse buttress, and trapezoid.\n\nThe head of a Cow femur was used as the bone model25, to reveal its cancellous bone core, it was sliced in 4 cm thick slices in which the implants were inserted. As shown in Figure 2, all the implants were inserted into osteotomies prepared by drills coupled to hydrated hand piece of a dental drill unit (Osseoset 200, Nobel Biocare) to a full length of 13 mm. In total, six implants of each group were inserted into osteotomies prepared by conventional cutting drills (cutting mode, clockwise rotation 1100 RPM), starting with drill size 1.5 mm, 2 mm, 2.4–2.8 mm, 2.8–3.3mm, 3.2–3.6mm and 3.8–4.2mm. The other six were inserted into osteotomies prepared by OD burs (Densah Burs) (OD mode, contra-clockwise rotation 1100 RPM), that started first with conventional drilling with 1.5 mm and 2 mm cutting drills (cutting mode), then OD burs (Figure 3) of size 2.5 mm (DENSAH Bur-G2 VS2228), 3.0 mm (DENSAH Bur-G2 VT2535), and 3.5 mm (DENSAH Bur-G2 VS3238) were used.\n\nEach inserted implant was tested for primary stability using the Periotest, with the tip of the Periotest retractable pin was applied to the same position in the implant abutment (Figure 4). The Periotest values interpretation, as described by the manufacturer, state that values between -8 and zero indicate good primary stability, and values above that range indicate insufficient integration between the implant and the surrounding bone.\n\nThe Periotest values (PTVs) for the implant stability were tabulated and analyzed using statistical package for social science (SPSS version 20 for windows). Comparisons between the study groups were carried out using the chi square test at significance level p< 0.05\n\n\nResults\n\nThis work assessed the effect of OD versus regular drilling on dental implant primary stability. To rule out the effect of the implant design, the most commonly used thread shapes were tested in both ostoetomies. The Periotest was used to read each implant primary stability. The results of this this study revealed no statistically significant difference between the Periotest readings for the implants in each category placed in either the OD or the regular osteotomies (Table 1).\n\nWhen all the implants placed in regular drilling ostoetomies were compared to all the implants placed in OD ostoetomies, statistical analysis of the Periotest readings for primary stability has shown that the implants placed in regular drilling ostoetomies were significantly more stable than the implants placed in OD osteotomies (Table 2).\n\n\nDiscussion\n\nThe aim of this study was to evaluate the effect of dental implant osteotomy preparation, using the OD technique, on the dental implant primary stability compared to conventional drilling. Since the dental implants may present many variables, such as the thread designs and surface treatments, which may affect the primary stability, this study utilized custom-made implants, with machined surface to avoid the effect of surface treatments on primary stability, and different thread patterns having the same thread pitch and depth, to detect which thread design will provide better primary stability in conjunction with OD.\n\nSince the implants used in this study were custom-made, it was difficult to use the resonance frequency analysis to evaluate the primary stability of the implants as the Osstell device requires the use of a smart peg which was difficult to custom make; however, the Periotest was used. Javed et al.14 have stated that the both the Osstell and the Periotest can be used to measure the dental implant primary stability, although the Periotest readings are less sensitive. Andresen et al.16 have also approved the use of Periotest once the clinicians consider its limitations and the difficulty in results interpretation. A different perspective was given by Noguerol et al.17 who stated that the Periotest mechanical testing would definitely give a better evaluation for implant stability than any radiographic study. Furthermore, Oh et al.18,19 found that the Periotest was comparable and as reliable as the Osstell. However, Aparicio et al.20 emphasized that it is important to consider several readings of either device over a long period of time in order to be able to evaluate the implant stability.\n\nIn the presence of soft bone, under-sizing the implant osteotomy is thought to give a better implant primary stability; however, Jimbo et al.1 stated that this technique is efficient when the implant bed is decreased by 10% of its diameter and that any further decrease did not improve the primary stability. Huwais and Meyer2 introduced the bone compaction technique through the OD drilling, and claimed that it increased the insertion torque, bone-to-implant contact, and accordingly resulted in greater primary stability compared to conventional drilling and to Summers15’ osteotome technique. This hypothesis has been confirmed by the work of Lahens et al.3 who reported a significantly higher bone-to-implant contact for OD, and Lopez et al.4 who tested the OD technique in vivo and reported its significant success over conventional drilling mechanically using the pull-out testing and microscopically using the histomorphometry. Additionally, Trisi et al.5 have shown that OD allows the use of wider implants diameters in narrow edentulous ridges, with consequent increase in bone volume. This increase in bone was later shown to reach 30% of the original ridge dimensions by Podaropoulos6.\n\nHowever, the results of this study did not find any statistically significant difference between the effects of OD and conventional drilling on the dental implant primary stability with any of the different thread designs used. This came in agreement with the findings of Wang et al.7, who reported that OD increased the apparent density of the peri-implant bone, but did not significantly improve the bone-implant contact, or the primary stability, and that OD created high strains at the bone implant interface, with damage to the bone trabeculae leading to extended periods of resorption and delayed secondary stability.\n\nIn accordance with the findings of Abuhussein et al.8, the implants used in this study had deep threads with a decreased thread pitch to ensure bone anchorage, and based on the conclusion of Chong et al.11, being without self-tapping properties, the threads were thought to provide higher primary stability than self-tapping threads. However, none of the tested thread shapes has shown any superiority in achieving better primary stability.\n\nNotably, the results of this study showed that there was no statistically significant difference between the OD and the regular drilling techniques, nor between the different thread designs used based on the Periotest values recorded for the implant primary stability. Considering the bone model used, Alkhodary et al.25 have stated that the elastic modulus of the cancellous bone of the cow femur head was comparable to that of the cancellous bone of the human mandible, which is in turn more compact than the bone in the posterior maxillae, and according to Chong et al.11 and Summers15, optimal bone quality and quantity can mask any difference in the implant different designs. This suggestion has been further potentiated by Bischof et al.24 who studied the factors affecting the dental implants primary stability, and reported that it is not the diameter or length of the implant, nor the implant thread deepening that affect primary stability, rather than the bone type in the mandible or the maxilla.\n\nAccordingly, it can be concluded that OD is not useful in compact bone, and might have a different effect in soft bone, and that the effects of different thread designs are more noticed in cancellous, rather than compact bone as shown by Kong et al.10 who reported different stress distribution patterns by the different thread shapes at the cancellous bone-implant interface. This came in agreement with the second finding of this study, where all implants placed in regular osteotomies had a significantly better primary stability than all the implants placed in OD osteotomies. This can be explained by the fact that soft bone has wider marrow spaces between the bone trabeculae, allowing for bone compaction, rather than the compact bone which the OD would lead to lateral compression that exceeds the viscoelastic limit of the thick and dense bone trabeculae, with subsequent damage and a weaker bone implant interface. Also, it is recommended to examine the effect of OD in an in vivo situation, where the bone available is soft, and where the monitoring process by the Periotest, or the resonance frequency analysis can be conducted several times to detect the effect of OD on both the primary and secondary stability of the dental implants.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.", "appendix": "Grant information\n\nThe authors gratefully acknowledge Qassim University, represented by the Deanship of Scientific Research, on the material support for this research under the number (dent-2018-1-14-S-3570) during the academic year 1440 AH/2018 AD.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nJimbo R, Tovar N, Anchieta RB, et al.: The combined effects of undersized drilling and implant macrogeometry on bone healing around dental implants: an experimental study. Int J Oral Maxillofac Surg. 2014; 43(10): 1269–1275. PubMed Abstract \| Publisher Full Text\n\nHuwais S, Meyer EG: A Novel Osseous Densification Approach in Implant Osteotomy Preparation to Increase Biomechanical Primary Stability, Bone Mineral Density, and Bone-to-Implant Contact. Int J Oral Maxillofac Implants. 2017; 32(1): 27–36. PubMed Abstract \| Publisher Full Text\n\nLahens B, Neiva R, Tovar N, et al.: Biomechanical and histologic basis of osseodensification drilling for endosteal implant placement in low density bone. An experimental study in sheep. J Mech Behav Biomed Mater. 2016; 63: 56–65. PubMed Abstract \| Publisher Full Text\n\nLopez CD, Alifarag AM, Torroni A, et al.: Osseodensification for enhancement of spinal surgical hardware fixation. J Mech Behav Biomed Mater. 2017; 69: 275–281. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTrisi P, Berardini M, Falco A, et al.: New Osseodensification Implant Site Preparation Method to Increase Bone Density in Low-Density Bone: In Vivo Evaluation in Sheep. Implant Dent. 2016; 25(1): 24–31. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPodaropoulos L: Increasing the Stability of Dental Implants: The Concept of Osseodensification. Balk J Dent Med. 2017; 21(3): 133–140. Publisher Full Text\n\nWang LY, Wu Y, Perez KC, et al.: Effects of Condensation on Peri-implant Bone Density and Remodeling. J Dent Res. 2017; 96(4): 413–420. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAbuhussein H, Pagni G, Rebaudi A, et al.: The effect of thread pattern upon implant osseointegration. Clin Oral Implants Res. 2010; 21(2): 129–136. PubMed Abstract \| Publisher Full Text\n\nPark JH, Lim YJ, Kim MJ, et al.: The effect of various thread designs on the initial stability of taper implants. J Adv Prosthodont. 2009; 1(1): 19–25. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKong L, Liu B, Li D, et al.: Comparative study of 12 thread shapes of dental implant designs: a three-dimensional finite element analysis. World Journal of Modelling and Simulation. 2006; 2(2): 134–140. Reference Source\n\nChong L, Khocht A, Suzuki JB, et al.: Effect of implant design on initial stability of tapered implants. J Oral Implantol. 2009; 35(3): 130–135. PubMed Abstract \| Publisher Full Text\n\nLan TH, Du JK, Pan CY, et al.: Biomechanical analysis of alveolar bone stress around implants with different thread designs and pitches in the mandibular molar area. Clin Oral Investig. 2012; 16(2): 363–369. PubMed Abstract \| Publisher Full Text\n\nWu SW, Lee CC, Fu PY, et al.: The effects of flute shape and thread profile on the insertion torque and primary stability of dental implants. Med Eng Phys. 2012; 34(7): 797–805. PubMed Abstract \| Publisher Full Text\n\nJaved F, Ahmed HB, Crespi R, et al.: Role of primary stability for successful osseointegration of dental implants: Factors of influence and evaluation. Interv Med Appl Sci. 2013; 5(4): 162–167. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSummers RB: A new concept in maxillary implant surgery: the osteotome technique. Compendium. 1994; 15(2): 152, 154–6, 158 passim; quiz 162. PubMed Abstract\n\nAndresen M, Mackie I, Worthington H: The Periotest in traumatology. Part I. Does it have the properties necessary for use as a clinical device and can the measurements be interpreted? Dent Traumatol. 2003; 19(4): 214–217. PubMed Abstract \| Publisher Full Text\n\nNoguerol B, Muñoz R, Mesa F, et al.: Early implant failure. Prognostic capacity of Periotest: retrospective study of a large sample. Clin Oral Implants Res. 2006; 17(4): 459–464. PubMed Abstract \| Publisher Full Text\n\nOh JS, Kim SG, Lim SC, et al.: A comparative study of two noninvasive techniques to evaluate implant stability: Periotest and Osstell Mentor. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2009; 107(4): 513–518. PubMed Abstract \| Publisher Full Text\n\nOh JS, Kim SG: Clinical study of the relationship between implant stability measurements using Periotest and Osstell mentor and bone quality assessment. Oral Surg Oral Med Oral Pathol Oral Radiol. 2012; 113(3): e35–e40. PubMed Abstract \| Publisher Full Text\n\nAparicio C, Lang NP, Rangert B: Validity and clinical significance of biomechanical testing of implant/bone interface. Clin Oral Implants Res. 2006; 17 Suppl 2: 2–7. PubMed Abstract \| Publisher Full Text\n\nWaechter J, Leite FR, Nascimento GG, et al.: The split crest technique and dental implants: a systematic review and meta-analysis. Int J Oral Maxillofac Surg. 2017; 46(1): 116–128. PubMed Abstract \| Publisher Full Text\n\nEsposito M, Grusovin MG, Rees J, et al.: Effectiveness of sinus lift procedures for dental implant rehabilitation: a Cochrane systematic review. Eur J Oral Implantol. 2010; 3(1): 7–26. PubMed Abstract\n\nChiapasco M, Casentini P, Zaniboni M: Bone augmentation procedures in implant dentistry. Int J Oral Maxillofac Implants. 2009; 24 Suppl: 237–259. PubMed Abstract\n\nBischof M, Nedir R, Szmukler-Moncler S, et al.: Implant stability measurement of delayed and immediately loaded implants during healing. Clin Oral Implants Res. 2004; 15(5): 529–39. PubMed Abstract \| Publisher Full Text\n\nAlkhodary MA, Abdelraheim AE, Elsantawy AE, et al.: The development of a composite bone model for training on placement of dental implants. Int J Health Sci (Qassim). 2015; 9(2): 153–161. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "43178", "date": "23 Jan 2019", "name": "Khalid Almas", "expertise": [ "Reviewer Expertise Periodontics and implant Dentistry" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an interesting in vitro study. Osseodensification technique is getting popularity due to improved local bone quality for implant placement with the technique.\n\nThe pilot work is interesting as far as the unified approach is concerned with 13 mm long implants and customized surface configurations. It is suggested that in future following variations should be considered to test the effect of osseodensification.\n1. Implants length less than 13 mm (8-11 mm) 2. Variable width. 3. Different thread pitch and depth. 4. Implant surface characteristics should be with SLA surface or any other non-machined surface. 5. The immediate stability of implants may not reflect the normal healing and stability physical equilibrium. As in real time clinical implant surgery, the initial stability and after 1-2 weeks stability may vary.\n\nThe discussion part of the manuscript should include variables mentioned above and future studies may include those points.\n\nFurther, radiographic or any other imaging technique should be used to see immediate bone-implant contact in both regular and OD technique.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "45551", "date": "29 Mar 2019", "name": "Jun Lee", "expertise": [ "Reviewer Expertise This paper appears to be a comparison of densah bur and conventional drill system according to implant design in animal models. Overall", "the content is simple but well proven. Only in this experiment", "it is considered that the animal model is designed properly" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper appears to be a comparison of densah bur and conventional drill system according to implant design in animal models. Overall, the content is simple but well proven. Only in this experiment, it is considered that the animal model is designed properly.\nI mention a few minor revisions:\n\nAbstract is 4-mm thick cancellous bone slices, but it is written as 4-cm in the text. In the experimental design, four implants are placed in one Cow femur slice, and I wonder how much distance there is between the implants (critical distance between implants). Indeed, the important consequences of this experiment would be to provide an experimental condition that does not interfere with drilling or implant placement to yield reliable results\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "47949", "date": "14 May 2019", "name": "Zachary Evans", "expertise": [ "Reviewer Expertise Implant surface technology", "digital strategies for implant planning" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a focused manuscript attempting to evaluate the effects of thread design on implant stability with an OD drill protocol in trabecular bone. It is a surprise that the OD protocol did not improve initial implant stability. Therefore, this is clinically relevant as it may influence providers to use the OD protocol for ridge expansion or sinus lifting, but not for densification. Studies like this are giving us information that may influence our clinical decisions in this regard. It is a simple, but very useful study that I hope will lead to a larger clinical study.\n\nTwo small comments to consider:\nConsider the use of the animal model terminology \"bovine\" vs. \"Cow\". Consider adding a comment on clinical relevance to the discussion - for example, what clinical scenarios would be appropriate or not for the OD protocol.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1898
https://f1000research.com/articles/7-1897/v1	05 Dec 18	{ "type": "Software Tool Article", "title": "An accessible GenePattern notebook for the copy number variation analysis of Illumina Infinium DNA methylation arrays", "authors": [ "Clarence K. Mah", "Jill P. Mesirov", "Lukas Chavez", "Jill P. Mesirov" ], "abstract": "Illumina Infinium DNA methylation arrays are a cost-effective technology to measure DNA methylation at CpG sites genome-wide and across cohorts of normal and cancer samples. While copy number alterations are commonly inferred from array-CGH, SNP arrays, or whole-genome DNA sequencing, Illumina Infinium DNA methylation arrays have been shown to detect copy number alterations at comparable sensitivity. Here we present an accessible, interactive GenePattern notebook for the analysis of copy number variation using Illumina Infinium DNA methylation arrays. The notebook provides a graphical user interface to a workflow using the R/Bioconductor packages minfi and conumee. The environment allows analysis to be performed without the installation of the R software environment, the packages and dependencies, and without the need to write or manipulate code.", "keywords": [ "Illumina Infinium methylation arrays", "DNA methylation", "copy number variation", "pre-processing", "interactive", "visualization", "GenePattern Notebook", "Jupyter Notebook", "open-source", "conumee", "minfi", "R/Bioconductor" ], "content": "Introduction\n\nAlthough Illumina Infinium DNA methylation arrays, including the 450k and EPIC (“850k”) BeadChips, have been designed for detecting genome-wide DNA methylation, the resulting data can also be used to analyze copy number profiles (Feber et al., 2014). This feature allows the simultaneous analysis of DNA methylation and copy number variation (CNV) and reduces the quantity of material needed to perform both analyses. We have implemented an Illumina Infinium DNA methylation array-based CNV analysis workflow as an accessible, interactive GenePattern notebook, which integrates background information, workflow instructions, a graphical user interface, source code, and the results in a single electronic notebook document (Mah, 2018). Leveraging the popular GenePattern Notebook environment (Reich et al., 2017), the notebook enables the sharing of reproducible analyses and results.\n\nThe workflow is initiated by a single step and performs two main analyses: loading and preprocessing the data, and copy number analysis (Figure 1). Multiple samples can be analyzed in parallel. The preprocessing step utilizes the minfi R package to load and process Illumina Infinium DNA methylation array data and to perform data normalization (Aryee, 2014). Copy number analysis is performed using the conumee R package, which compares each sample to a set of user-provided normal reference samples (Hovestadt & Zapatka, 2015). This analysis outputs a set of copy number plots for the entire genome, individual chromosomes, and for user defined gene loci of interest. Copy number profiles are described as segments along the genome and can be exported as text files for visualization with tools such as the Integrated Genome Viewer (Robinson et al., 2011) and for further analysis.\n\nThe flowchart shows the main inputs and outputs necessary for the copy number variation analysis.\n\n\nMethods\n\nThe entire workflow is implemented as a GenePattern notebook, which can be accessed at the GenePattern Notebook Repository (http://www.genepattern-notebook.org/) and run there by the user. Data preprocessing and CNV analysis steps are implemented as a GenePattern module (Reich et al., 2006) and utilized by the MethylationCNVAnalysis notebook.\n\nTo begin the analysis, two sets of data are required: the query sample data for which the copy number profiles are to be analyzed and appropriate control sample data used to establish baseline copy number profiles for comparison (Figure 2). The input data for this notebook (query and control samples) are raw IDAT files generated by the microarray scanner, representing two different color channels prior to normalization. As described in the minfi documentation, IDAT files are the most complete data types, because they include measurements on control probes, which are necessary for assessing bisulfite conversion efficiency and for normalizing technical variability.\n\nThe “MethylationCNVAnalysis” module is presented as an input form using the GenePattern Notebook graphical user interface. The user links or uploads input files and selects analysis parameters before pressing “Run” to execute the workflow.\n\nTo load the Illumina Infinium methylation array data into the notebook, the IDAT files must be combined into a single archive (.zip or .gz formats). The archive can be organized either as a flat archive where all IDAT files are packed without subfolders, or as an archive in the standard folder structure as presented in the Illumina demo dataset. The IDAT archive can be selected and loaded through the graphical user interface of the GenePattern notebook. Both 450k array and EPIC array types are compatible as long as all samples in a single archive are of the same array type. If the query samples or control samples are of different array types, only the common set of probes between 450k and EPIC array types are evaluated across all samples.\n\nFor each sample, the data is normalized with respect to background and positive control probes on the arrays according to the implementation in Illumina’s proprietary GenomeStudio software. Upon loading the data, the notebook generates a quality control report containing two plots for identifying poor quality samples. The first plot shows the log2 median intensity of the methylated versus unmethylated channels (Figure 3A). Poor-quality samples tend to have lower median intensities and separate from the good quality samples. The second plot shows the DNA methylation levels (Beta values) of all probes on the array and for all samples as a density plot in which we expect to see a bimodal distribution with peaks at zero (no methylation) and one (100% methylation) (Figure 3B).\n\n(A) Median intensity plot of query & control samples. Log median intensity of the methylated channel is along the x-axis and log median intensity of unmethylated channel is along the y-axis. Bad-quality samples fall under the threshold and are colored red. There is no bad quality sample in this plot. (B) DNA methylation (Beta-value) density plot of query & control samples. A density plot showing the distribution of beta values across each sample. Beta values should be bimodal and peak around 0 and 1.0.\n\nControl samples should be free of CNVs and have a similar methylation profile as the samples of interest. The best practice is to use control samples of the corresponding normal tissue type. If control samples are included in the query sample dataset, no separate data needs to be loaded. Instead, the control samples can be specified by providing the sample names in the CNV analysis step. Otherwise, the control data will be loaded as a separate archive of IDAT files.\n\nAs outlined in the conumee documentation, the copy number analysis is performed as follows: each query sample is normalized to the control samples by multiple linear regression yielding the linear combination of control samples that most closely fits the intensities of the query sample. Next, the log2 ratio of probe intensities of the query sample versus the combination of control samples are calculated. Probes are then combined within predefined genomic bins. Intensity values are shifted to minimize the median absolute deviation of all bins to zero to determine the copy-number neutral state. The genome is segmented into regions of the same copy number state using the circular binary segmentation algorithm (Seshan & Olshen, 2018).\n\nGenomic loci of genes to be highlighted in the CNV plots are retrieved from the hg19 Ensembl database using the BiomaRt R package (Durinck, 2005; Durinck, 2009). The notebook also offers an option to exclude regions from analysis, such as highly polymorphic regions that would yield inaccurate copy number calls. In addition, X and Y chromosomes can be excluded to avoid misleading results in case no appropriate control data is available.\n\nTo run the MethylationCNVAnalysis notebook, the user must have a GenePattern account that can be created on the GenePattern Notebook website (http://genepattern-notebook.org). After logging in, the notebook can be found in the “Community” section of the “Public Notebooks” page. The notebook can then be run from the GenePattern Notebook site, with no additional software installations needed.\n\n\nUse case\n\nThe use case presented by the notebook evaluates the copy number profile of a glioblastoma tumor analyzed by an Illumina Infinium 450k DNA methylation array. This sample has been classified as an IDH wild-type midline glioblastoma according to the methylation-based classifier described by Capper et al. (2018). Recurrent chromosomal alterations of this tumor type are gain of chromosome 7 with or without EGFR amplification (>80%), loss of 9p21 (CDKN2A/B; >50%) and chromosome 10 loss (>70%). Amplifications of the PDGFRA oncogene are enriched in this class (present in 20–30% of cases) (Capper, 2018).\n\nWe used the 450k methylation profiles of 119 normal brain tissue samples as the corresponding control data (Capper, 2018). By inspecting the generated CNV plots, we can visually identify significant copy number loss of CDKN2A/B relative to normal brain tissues (Figure 4). Additionally, several copy number changes that are associated with glioblastoma stand out, notably MET amplification and loss of RB1.\n\n(A) Copy number plot of the entire genome in the example glioblastoma sample. A plot of all chromosomes across the genome. Intensity values of each bin are plotted as colored dots, green indicating above normal copy number, red indicating below normal copy number, and grey indicating close to normal copy number. Blue lines indicate the median intensity of each bin. Specified genes to be highlighted are annotated. (B) Copy number plot of common cancer genes in the example glioblastoma sample. An overview of the genomic loci of common cancer genes are shown in more detail. Copy number values are visualized as described in Figure 4A.\n\n\nConclusion\n\nThe GenePattern notebook MethylationCNVAnalysis, hosted in the GenePattern Notebook Repository, processes Illumina Infinium DNA methylation array data and generates CNV segments and plots. Different designs of Illumina Infinium DNA methylation arrays have been produced by the manufacturer including the 450k and EPIC arrays. Importantly, different batches of these designs can contain a variable set of probes. As a result, the GenePattern notebook requires all query samples to be of the same array design. Similarly, all control samples have to be of the same array design, which can be different from the query samples. If the query samples and the control samples are of different array designs, only the common set of probes between the array designs are evaluated for the CNV analysis. As described above, the choice of control samples is crucial for the resulting copy number profiles. The control samples should be free of CNVs and have a similar methylation profile as the samples of interest. Provided that query and corresponding control samples are available, the MethylationCNVAnalysis notebook in the GenePattern Notebook Repository allows the CNV analysis to be performed without the installation of software and without the need to write or manipulate code.\n\n\nData availability\n\nThe notebook includes links to the data for running the use case described above. The raw data can be found in GEO Series GSE90496: https://identifiers.org/geo/GSE90496.\n\n\nSoftware availability\n\nGenePattern Notebook is available from: http://genepattern-notebook.org/.\n\nA public preview of the notebook is available from: https://notebook.genepattern.org/services/sharing/notebooks/136/preview/\n\nGenePattern Notebook source code is available from: https://github.com/genepattern/methylation_cnv_analysis_notebook.\n\nArchived source code at time of publication: https://doi.org/10.5281/zenodo.1419319 (Mah, 2018).\n\nLicense: BSD 3-Clause.", "appendix": "Grant information\n\nThis study was funded by the National institutes of Health (grant numbers U24CA194107, U01CA184898, U41HG007517 and R01CA109467 to J.P.M.).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nThe authors thank members of the Mesirov Lab, including Edwin Juarez, Ted Liefeld and the GenePattern Development team for assisting with software development. The authors also thank Briana Prager for testing the notebook on independent datasets and Stephen Mack for providing additional datasets for testing. We thank Konstantin Okonechnikov and Volker Hovestadt for providing feedback on the analysis.\n\n\nReferences\n\nAryee MJ, Jaffe AE, Corrada-Bravo H, et al.: Minfi: a flexible and comprehensive Bioconductor package for the analysis of Infinium DNA methylation microarrays. Bioinformatics. 2014; 30(10): 1363–1369. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCancer Genome Atlas Research Network: Comprehensive genomic characterization defines human glioblastoma genes and core pathways. Nature. 2008; 455(7216): 1061–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCapper D, Jones DTW, Sill M, et al.: DNA methylation-based classification of central nervous system tumours. Nature. 2018; 555(7697): 469–474. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDurinck S, Moreau Y, Kasprzyk A, et al.: BioMart and Bioconductor: a powerful link between biological databases and microarray data analysis. Bioinformatics. 2005; 21(16): 3439–3440. PubMed Abstract \| Publisher Full Text\n\nDurinck S, Spellman PT, Birney E, et al.: Mapping identifiers for the integration of genomic datasets with the R/Bioconductor package biomaRt. Nat Protoc. 2009; 4(8): 1184–1191. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFeber A, Guilhamon P, Lechner M, et al.: Using high-density DNA methylation arrays to profile copy number alterations. Genome Biol. 2014; 15(2): R30. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHovestadt V, Zapatka M: conumee: Enhanced copy-number variation analysis using Illumina 450k methylation arrays. R package version 0.99, 4. 2015. Reference Source\n\nMah C: genepattern/methylation_cnv_analysis_notebook v1.0.1 (Version v1.0.1). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1419327\n\nReich M, Liefeld T, Gould J, et al.: GenePattern 2.0. Nat Genet. 2006; 38(5): 500–1. PubMed Abstract \| Publisher Full Text\n\nReich M, Tabor T, Liefeld T, et al.: The GenePattern Notebook Environment. Cell Syst. 2017; 5(2): 149–151.e1. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRobinson JT, Thorvaldsdóttir H, Winckler W, et al.: Integrative genomics viewer. Nat Biotechnol. 2011; 29(1): 24–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSeshan VE, Olshen A: DNAcopy: DNA copy number data analysis. R package version 1.54.0. 2018. Reference Source" }	[ { "id": "45332", "date": "15 Apr 2019", "name": "Robert Ivanek", "expertise": [ "Reviewer Expertise bioinformatics" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article by CK Mah et al. describes a new GenePattern Notebook \"MethylationCNVAnalysis\" which allows users of GenePattern platform to run copy number analysis (CNV) of Illumina Infinium DNA methylation arrays. The tool provides graphical user interface to the CNV analysis based on Bioconductor packages `minfi` and `conumee`. Such analysis can be with this tool performed even without programming experience or without functional R installation.\n\nCompared to analysis on the R command line, the user is asked to set only few parameters required for the analysis: set of test and control samples, gene list for detailed view, a black-list of regions excluded from the analysis and a coice to in-/exclude sex chromosomes. In my opinion it would be helpful to create a section with \"Advanced settings\" and allow also specification of few other parameters: `bin_minprobes` and `bin_minsize` for function `CNV.create_anno` from `conumee` package and parameters for segmentation of log2ratio (functions from `DNAcopy` package). In both cases the authors of `conumee` package warn, that they optimized parameters for 450k arrays and another array type might require further optimization (see http://bioconductor.org/packages/release/bioc/vignettes/conumee/inst/doc/conumee.html). Also in respect to general need for higher reproducibility of analyses, the tool should export short summary of software, their versions and parameters used for analysis.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [] }, { "id": "47252", "date": "02 May 2019", "name": "Aris Floratos", "expertise": [ "Reviewer Expertise Computational biology", "bioinformatics" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article describes a GenePattern notebook for inferring CNVs using methylation profiling data generated by Illumina Infinium arrays. The notebook leverages the well-established Bioconductor packages minfi and conumee to implement an analysis workflow that comprises quality control, CNV calling, and results visualization. Functionality is made available through a web browser interface and requires no software installation/configuration, making it an attractive option for users with limited informatics expertise.\nSome thoughts about possible improvements:\nGiven that the entire workflow can be somewhat time consuming (the notebook documentation indicates that processing a single sample takes about 2 minutes), it would be useful if there was an option to run the QC step as a stand-alone computation, not combined with the CNV calling. As things stand right now, a significant amount of time can be spent waiting for the analysis to complete, only to realize that some control samples are of low quality and, thus, need to be removed and the analysis be rerun. When running the notebook without specifying values in the “genes to highlight” or “ignore regions” parameter boxes, the run fails with “getopt” error messages. It is not clear why these parameters are mandatory (e.g., in the case of \"genes to highlight\", it is conceivable that one may want to only inspect genome-wide patterns of aberration, without focusing on specific genes); but if so, it would be helpful to state clearly in the documentation section.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1897
https://f1000research.com/articles/6-946/v1	20 Jun 17	{ "type": "Research Article", "title": "Mapping of microRNAs related to cervical cancer in Latin American human genomic variants", "authors": [ "Milena Guerrero Flórez", "Olivia Alexandra Guerrero Gómez", "Jaqueline Mena Huertas", "María Clara Yépez Chamorro", "Olivia Alexandra Guerrero Gómez", "Jaqueline Mena Huertas", "María Clara Yépez Chamorro" ], "abstract": "Background: MicroRNAs are related to human cancers, including cervical cancer (CC), which is mainly caused by human papillomavirus (HPV) infection. In 2012, approximately 70000 cases and 28000 deaths from this cancer were registered in Latin America according to GLOBOCAN reports. The most frequent genotype worldwide is HPV-16. The main molecular mechanism of HPV in CC is related to integration of viral DNA into the hosts’ genome. However, the different variants in the human genome can result in different integration mechanisms, specifically involving microRNAs (miRNAs). Methods: miRNA sequences associated with CC and four human genome variants from Latin American populations were obtained from miRBase and the 1000 Genomes Browser, respectively. HPV integration sites near cell cycle regulatory genes were identified. miRNAs were mapped on human genomic variants. miRSNPs (single nucleotide polymorphisms in miRNAs) were identified in the miRNA sequences located at HPV integration sites on the human genomic Latin American variants. Results: Two hundred seventy-two miRNAs associated with CC were identified in 139 reports from different geographic locations. By mapping with the Blast-Like Alignment Tool (BLAT), 2028 binding sites were identified from these miRNAs on the human genome (version GRCh38/hg38); 42 miRNAs were located on unique integration sites; and miR-5095, miR-548c-5p and miR-548d-5p were involved with multiple genes related to the cell cycle. Thirty-seven miRNAs were mapped on the human Latin American genomic variants (PUR, MXL, CLM and PEL), but only miR-11-3p, miR-31-3p, miR-107, miR-133a-3p, miR-133a-5p, miR-133b, miR-215-5p, miR-491-3p, miR-548d-5p and miR-944 were conserved. Conclusions: 10 miRNAs were conserved in the four human genome variants, and in the remaining 27 miRNAs, substitutions, deletions or insertions were observed in the nucleotide sequences. This variability can imply differentiated mechanisms towards each genomic variant in human populations, relative to specific genomic patterns and geographic features. These findings may be decisive in determining susceptibility to the development of CC. Further identification of cellular genes and signalling pathways involved in CC progression could lead to the development of new therapeutic strategies based on miRNAs.", "keywords": [ "cervical cancer", "HPV", "HPV integration sites", "microRNAs", "miRNAs", "secondary structure", "human genome variants", "bioinformatics tools" ], "content": "Introduction\n\nCervical cancer (CC) is the second most common malignancy in women worldwide. According to GLOBOCAN reports, approximately 530,000 women are diagnosed with CC and 265,672 die from it each year1. Infection by human papillomavirus (HPV) has been recognized as the major risk factor in this pathology2,3, but the virus presence is not the main cause for the development of this cancer4,5. Viral DNA integration into the host cell genome is considered a conducive factor for cervical intraepithelial neoplasia (CIN) to develop into CC5–7.\n\nNumerous microRNAs (miRNAs) have been identified in proximity to HPV integration sites8,9. miRNAs are a class of small (18 to 26 nucleotides length), noncoding, evolutionarily conserved RNAs that are processed from longer transcripts known as pre-miRNAs (60 to 100 nucleotides in length)10. They are located on regions known as fragile sites and distributed in intergenic, intronic and exonic segments of the human genome involved in cancer11,12. Functionally, they regulate post-transcriptional expression levels of up to 60% of total protein-encoding genes by binding their seed sequences (2–8 nucleotides length). The 5'-UTR end of the miRNA seed sequence is complementary to the 3'-UTR end of the target mRNAs13. This recognition event can affect the expression of important regulatory genes. Deregulation of genes such as tumour suppressor genes and oncogenes can lead to cancer development, including CC14–16.\n\nHuman genome variants generate different patterns of miRNA deregulation17, which can contribute to cancer development susceptibility, treatment efficacy and patient prognosis18–20. 99% of the human genome is genetically identical, and the remaining 1% is responsible for all human diversity. miRNAs represent a major part of this genetic variation21. miRSNPs (single nucleotide polymorphisms in miRNAs) are human polymorphisms at or near predicted miRNA target sites22. The occurrence of miRSNPs can influence miRNA functionality on all levels, including transcription, maturation, and mRNA target binding.\n\nKnowledge on miRNAs related to CC development in human genome variants from Latin American populations is scarce. Thus, in this study, we mapped miRNAs associated with CC in human genome variants obtained from Colombia, Mexico, Peru and Puerto Rico. Complete genomes were included in this study. Additionally, the relationships between HPV integration sites, genes close to these sites, mapping profiles and mutation patterns for each of the miRNAs were estimated for each of the genome sequences. The objective of this research was to analyse how genetic variation of CC-associated miRNAs identified in previously reported HPV integration sites affects cell cycle regulatory genes in human genomic variants from Latin America.\n\n\nMethods\n\nTwo hundred and seventy-two miRNAs associated with CC were selected as described in the systematic review published by Guerrero & Guerrero23. With the information contained in miRBase24–26, miRNAMap27 and miRNAstart, features, such as length, chromosomal and genomic location of pre-miRNAs and mature miRNAs, were analysed. The mature miRNA reference sequences were obtained in FASTA format from the miRBase database (Dataset 128).\n\nFour human genome sequences were obtained from randomly selected female participants in the 1000 Genomes Project from Latin American populations22,29. Their codes were CLM (from Medellin in Colombia), MXL (from Los Angeles and of Mexican ancestry in the USA), PEL (from Lima in Peru) and PUR (from Puerto Rico). The control sequence was a variant that is phylogenetically distant to Latin American variants and identified with the code BEB (from Bangladesh and of Bengali ancestry). Access codes were obtained from the 1000 Genomes Project resources21,30. This information is summarized in Table 1.\n\nViral insertion sites and nearby genes on the human genome were identified with the UCSC Genome Bioinformatics search engine31,32. To establish possible functional relationships with the development of CC, functional information on the associated functions of these human genes was obtained from UniProt33,34.\n\nAccording to Xia et al.35, the mature miRNA sequences are located in regions with pre-miRNA secondary structure complementarity (3' and 5'). In total, 445 miRNA sequences were analysed. The Blast-Like Alignment Tool (BLAT) available on the UCSC Genome Bioinformatics website was used for mapping the miRNAs associated with the full human genome with the following parameters: (a) genome, human; (b) assembly, Dec. 2013 (GRCh38/hg38); (c) query type, DNA; (d) sort output, query; and (e) score and output, hyperlinks. A matrix of chromosomal location data was built with Microsoft Excel 2013 (‘Matrix of data’ in Dataset 236). From this matrix, the miRNAs over HPV integration sites were manually identified.\n\nTo identify miRNA mutations in the four Latin American human genome variants, the available tools, including ideogram view, subjects and exon navigator, in the NCBI 1000 Genomes Browser (Phase 3, version 3.7) were used. The code for each female genetic variant selection (Colombia, Mexico, Peru, Puerto Rico and Bangladesh) was inserted and the sequence of each miRNA identified in viral integration sites was introduced and the mapped nucleotide positions were selected. Using WebLogo 337, logos were created to view the nucleotide differences. The bioinformatics workflow is summarized in Figure 1.\n\n\nResults\n\nA total of 44 publications were identified between 1987 and 2015 related to HPV integration sites in the human genome. The most frequent types of HPV associated with CC were HPV-16 and HPV-18. Details of these articles are outlined in Supplementary File 1. Five hundred and seventy-eight integration sites for 8 types of HPV associated with different histological cervical conditions were identified, of which 63.84% were HPV-16 (Figure 2 and ‘HPV integration sites’ in Dataset 236).\n\nHPV-16 and HPV-18 have integration sites on all human chromosomes. HPV-16 has more integration sites on chromosomes 2, 1, 3, 6, 9, 5, 8 and 4, while HPV-18 has more on chromosomes 2, 1, 8, 12, 5, 10, 4, 6 and 9. Some less frequently oncogenic HPV types have integration sites on specific chromosomes, such as HPV-45 on 2, 1, 3, 9, 4, 7 and 13; HPV-33 on 9, 13, 5, 6, 8, 11, 16, 18 and X; HPV-58 on 4, 12 and 18; HPV-31 on 2 and 17; HPV-67 on 4 and 13; and HPV-68 on chromosome 18. Chromosomes 1 and 2 displayed a higher number of viral insertion sites (41 and 45, respectively), while chromosomes 13 and 18 displayed insertion sites for 5 different HPV genotypes. The chromosomal loci with the highest numbers of HPV integration sites are presented in Table 2.\n\nInformation on the associated functions of genes located near HPV integration sites obtained from UniProt showed that 86.1% of the genes located in close proximity were involved in apoptosis, cell adhesion, cell differentiation, ion transport and metabolic processes. Fifty-four genes were involved in direct regulation of the cell cycle. Twenty-six of these were tumour suppressor genes, 8 were oncogenes, 8 were proto-oncogenes and 12 did not have a determined functionality in the development of this neoplasia (Figure 3).\n\nThe 2028 miRNA binding sites associated with CC in the human genome were identified from BLAT mapping using previously identified miRNAs23, including 432 sites previously reported in miRBase (‘Results of mapping with BLAT’ in Dataset 236). These sites were located on both DNA strands (52.97% on the positive strand and 47.03% on the negative strand). 1881 binding sites were fully complementary (100% sequence identity) to miRNA sequences, while 1, 24, and 122 binding sites had 96.2%, 95.7% and 95.5% sequence identity, respectively.\n\nmiR-5095 was mapped onto 853 binding sites on 23 chromosomes. Four hundred and twenty-four mature miRNAs sequences (98.15%) mapped to one, two, three and even ten different binding sites. miR-522-5p and miR-523-5p binding sites mapped only a single chromosome (Chr. 19). Table 3 shows the chromosomal location and number of binding sites for each specific miRNA associated with CC.\n\nThe distribution of the 2028 binding sites was not homogeneous along the human genome. 41% of the total binding sites were identified on chromosomes 1, 19, 5, 2, 3, 14, 7 and X. Although the number of miRNA binding sites correlated with the size of each chromosome, some short chromosomes, such as 19 and X, had more miRNA binding sites when compared to other larger chromosomes (Table 4).\n\nCHR= Chromosome.\n\n14.89% (302) of binding sites grouped into the following 19 specific chromosomal locations: (1) 19q13.42 (51 sites/14 miRNAs), (2) 14q32.31 (34 sites/16 miRNAs), (3) 13q31.3 (16 sites/11 miRNAs), (4) 14q32.2 (16 sites/9 miRNAs), (5) 4q25 (16 sites/7 miRNAs), (6) 20q13.33 (15 sites/7 miRNAs), (7) 16p13.3 (15 sites/4 miRNAs), (8) Xq26.2 (14 sites/8 miRNAs), (9) 7q22.1 (14 sites/6 miRNAs) and (10) 1p31.3 (14 sites/6 miRNAs). The remaining 9 chromosomal locations contained between 10 and 13 binding sites (Supplementary File 2). 92% (1865/2028) of the binding sites were distributed into 250 groups along the human genome; the remaining 8% (163/2028) of binding sites for various miRNAs including miR-5095 were distributed along the human genome without being distributed into any groups.\n\nEach group contains between 2 and 7 miRNA binding sites, although some groups contain between 8 and 16 (Figure 4). The majority of the groups are located on chromosomes 1, 2, 3, 5, 10 and 11. The biggest groups are located on chromosome 19, with 51 binding sites for 25 miRNAs involved in CC development.\n\n58.8% of miRNA binding sites associated with CC (1194 binding sites) are located in intergenic regions, 39.65% (804 binding sites) in intronic regions, 1.28% (26 binding sites) in exonic regions and 0.19% (4 binding sites) between intronic and exonic regions (mixed miRNAs). Figure 5 shows the variation in the number of intergenic, exonic and intronic miRNAs associated with CC.\n\nThirty-eight integration sites were found for six types of oncogenic HPV (HPV-16, -18, -33, -45, -58 and -68) in miRNA binding sites and cell cycle regulatory genes associated with CC (Table 5). The largest number of HPV integration sites was found for miR-5095 (33 sites), followed by miR-548c-5p (11 sites) and miR-548d-5p (11 sites) (Table 5). In 14 integration sites, no miRNA binding sites were detected. The highest number of miRNA binding sites was found in chromosome regions 18q11.2 and 19p13.12 (Supplementary File 2).\n\n1The information in brackets shows the number of miRNA binding sites, and whether the miRNAs are located on the positive sense or negative sense DNA strand.\n\n2The information in brackets shows and whether the cell cycle regulatory genes are located on the positive sense or negative sense DNA\n\n3 Cl: Classification of cellular genes; ST: tumor suppressors; OG: Oncogenes; PO: Proto-oncogenes.\n\nNinety-six possible interactions were identified between 37 mature miRNAs associated with CC and 42 cell cycle regulatory genes located in proximity to the viral insertion sites. The network of interactions is presented in Figure 6. 35.42% of the interactions involved miR-5095, 12.5% involved miR-548c-5p and 12.5% miR-548d-5p.\n\nRectangles of various colors represent the cell cycle regulatory genes, and color depend on their classification (ST - , OG - , POG - e IND - ). The arrows represent the interactions between miRNAs and genes involved in cell cycle regulation; each arrow's color depends on the DNA chain where miRNAs and cell cycle regulatory genes are located.\n\n38.1% of genes identified in HPV integration sites have binding sites for a single miRNA, and 61.9% have binding sites for more than two miRNAs. Table 6 displays genes with more than five miRNA binding sites.\n\nA gene may have binding sites for both regions of complementarity (3' and 5') of a miRNA38. In this study, we found that the TTC39C gene has binding sites for miR-133a-3p and miR-133a-5p and MAP3K1 has binding sites for miR-449b-3p and miR-449b-5p, though some mature sequences from one miRNA also showed binding sites to different genes (Figure 6). As an example, the miR-548c-3p mature chain has binding sites in the HAUS4 gene as well as in the MAP3K1, CDCA8, BCL2, ID4, cMYC, RAD51B, TSC2, ZBTB7C, FBXW7, CHEK2 and CDC7 genes (Figure 6).\n\n26.31% (10/42) of the miRNAs analysed (miR-11-3p, miR-31-3p, miR-107, miR-133a-3p, miR-133a-5p, miR-133b, miR-215-5p, miR-491-3p, miR-548d-5p and miR-944) were identical across the Latin American human genome variants, and 73.69% showed a genetic mutation (substitution or deletion of nucleotides) (Figure 7, Panels A and B).\n\nA) Number of miRNAs and nucleotide substitutions found in each human genomic variant; B) Number of miRNAs with between 1 and 7 nucleotide substitutions; C) Number of miRNAs with nucleotide substitutions in one, two or three genomic variants in the Latin American human genome, and D) Types of nucleotide substitutions in the miRNA sequences associated with CC in the selected human genome variants.\n\nWhen mapping the sequences of these miRNAs to the selected Latin American human genome variants (Supplementary File 3), 88 miRSNPs related to miRNAs or miRNA binding sites were identified on the Latin American variants compared with 33 on the reference variant. Twenty-one miRSNPs were located in the miRNA seed sequences of Latin American variants compared with 3 located in the reference variant. The most representative mapping results are shown in Table 6.\n\nTypes of nucleotide substitutions in the miRNA sequences associated with CC in the selected human genome variants showed that there were more frequent transversions than transitions and that the most frequent nucleotide substitutions were G→U (16.9%), followed by A→C (15.7%), C→A (15.7%) and G→A (10.8%) (Figure 7).\n\nBetween one and 18 nucleotide deletions were detected in miR-27a-3p, miR-31-5p, miR-103a-3p, miR-191-3p, miR-215-3p and miR-574. The sequences of miR-28, miR-152, miR-548c-5p, miR-572 and miR-5095 only mapped to reference sequences (version GRCh38/hg38), but not to any of the Latin American human genomic variants. miR-152 did not map to the PUR variant (Table 6).\n\nTable 7 displays the nucleotide variations from human genome variants obtained from Colombia, Mexico, Peru and Puerto Rico and Bangladesh, which was the control variant.\n\nMore data is available in Supplementary File 3.\n\n1HG: Human genome; CLM: variant of Medellin, Colombia; MXL: Los Angeles with Mexican ancestry; PEL: Lima, Peru; PUR: of Puerto Rico; BEB: Bengali, Bangladesh.\n\n2The size of each letter indicates the enrichment of each nucleotide in Latin American variants of the human genome, displayed through WebLogo.\n\n\nDiscussion\n\nAccording to the literature, approximately 570 integration sites have been identified for eight oncogenic HPV types associated with CC (Figure 2). HPV integration into cellular DNA and consequent deregulation of genes is considered a crucial step in cancer progression. Genotype HPV-16 is the most studied for its relationship with CC, as it is responsible for 70% of cases worldwide39. This could be a consequence of the greater proportion of integration sites reported for this genotype. In contrast, low risk genotypes, such as HPV-45, -66 and -93 reported in Colombia, are frequent in CC40–44.\n\nHPV integration into the host genome occurs in regions well-known as fragile sites, breakpoints or transcriptionally active regions45. This integration induces functional alterations of cellular genes in close proximity12,46–48. According to our results, the 8q24.21 chromosome region is the most affected by HPV integration. If we take into account that proto-oncogenes such as the MYC gene are located here49 (as displayed in Figure 3) and that MYC represents a family of genes overexpressed in several tumours including CC49–51, inhibition of MYC expression can induce cancer cell destruction50. In this context, the MYC gene could be both a tumour biomarker and potential treatment target for several tumours51 (Table 2).\n\nChromosomes 1, 14, 19 and X contain significantly more mature miRNAs than others, and chromosome 18 contains fewer miRNAs. The 19q13.4 chromosome region contains the largest group of human miRNAs (known as the group of miRNAs on chromosome 19 \"C19 MC\"), with alterations in several that have been previously reported in cancer52. Studies have reported associations between chromosome 1 and malignant transformation in cancers, including CC53.\n\nThe 578 integration sites identified in eight HPV types associated with CC were located in cell cycle regulatory genes, including the tumour suppressor genes TP73, P3H2, TP63, NBN, PTEN, BRCA1, and TPX2; the oncogenes EIF4E, CDCA8, MDM2, and PVT1; and the proto-oncogenes SRC, MYC, MCM5, CXCL8, and BCL2. Their deregulation could explain the progression of CC (Figure 3).\n\nIn 2011, Reshmi et al. used BLAT to determine the exact location of four miRNA binding sites associated with CC using bioinformatics programmes and computational tools54. To the best of our knowledge, this study is the first to use BLAT to identify miRNA binding sites in proximity to HPV integration sites involved in CC progression. In this study, 2028 binding sites from 272 CC-associated miRNAs were identified.\n\nIdentification of the target mRNAs of these miRNAs is considered a key step in their structural and functional analysis to establish possible interactions and consequently, cellular processes that may be altered in CC progression55–57. miRNAs located in the two strands of cellular DNA (5’ and 3’ strands) demonstrate their ability to interact in both orientations with the two strands of DNA and form triple helix structures to enhance RNA stability58,59.\n\nEach CC-associated miRNA showed a different number of binding sites in the human genome (Table 3, Supplementary File 2), and in the human genomic variants17,21,60,61; miRNAs were distributed throughout the genomes in both intronic or exonic regions13. In this study, CC-associated miRNAs were distributed in the karyosome, with chromosomes 1, 19, 5, 2, 3, 14, 7 and X having the largest number of miRNA binding sites (Table 4). These results are consistent with those reported by Calin et al.12. Because some chromosomes have a greater number of miRNA binding sites, it provides evidence of a non-random distribution of miRNAs within the chromosomes.\n\nOur results showed a low number of exonic miRNAs. These exonic miRNAs are considered rare miRNAs62, which are important candidates for gaining a better comprehension of interaction networks between miRNAs and their CC-associated targets.\n\nThe miRNA binding sites are within a short distance of each other in the chromosome, indicating that they tend to cluster63–66. Altuvia et al. reported miRNAs in groups of two or three64. This coincides with our results on CC-associated miRNA binding sites, as we found that miRNAs are capable of forming groups of more than 6 miRNAs on both strands of human DNA (Figure 4). We identified an important group of 16 miRNAs that can form these clusters and are located on chromosome 14 region 14q32.31. They include hsa-miR-134, miR-299, miR-323a, miR-329, miR-376a, miR-376c, miR-379, miR-411, miR-485, miR-487a, miR-487b, miR-494, miR-495, miR-539, miR-654 and miR-5095 (Supplementary File 2). Understanding their individual and collective roles is important when studying the development of this neoplasia.\n\nmiR-5095 had the highest number of binding sites distributed throughout the human genome (Table 3), which is in accordance with previously reported data66–68 where approximately 900 binding sites were identified; they are probably related to the expression of many target mRNAs and biological processes. Based on its extensive genomic distribution and low specificity in CC, miR-5095 is a good candidate to be used as an indicator of genetic variability within the human population.\n\nTo identify the role of miRNAs, HPV integration sites located in cell cycle-controlling genes were analysed. Thirty-seven miRNAs were identified in HPV integration sites close to cell cycle-controlling genes (Table 5). Nambaru et al. and Schmitz et al. identified numerous miRNAs in the proximity of HPV integration sites and reported that approximately 65% of these were involved in cervical carcinogenesis8,9. Inactivation of tumour suppressor genes by viral integration increases genomic instability and leads to cervical malignant neoplasm progression69.\n\nThe multiple miRNA binding sites on a target may decrease the levels of mRNA translation and improve the specificity of gene regulation. For example, one miRNA can have multiple target genes and each individual mRNA can be regulated by numerous miRNAs13,70,71. Ninety-seven interactions were identified between miRNAs and cell cycle regulatory genes (Table 4–Table 5, Figure 4–Figure 6); miR-5095, -548c-5p and -548d-5p showed the highest number of interactions with these kinds of genes.\n\nIvashchenko et al. identified miR-5095 binding sites in the BRCA1 gene67. In this study, miR-5095 was also found to have binding sites in the BAK1, BARD1, CITED2, MDM5, SRC, PARD3B, PPP2CA, RHEB, SOX2 and XPO1 genes (Table 5 and Figure 6). Our findings provide a basis for searching for other interactions, gene targets, and CC-associated miRNAs.\n\nDuring miRNA biogenesis, each pre-miRNA produces two mature miRNAs, such as miRNA-5p and miRNA-3p72. Mature miRNA deregulation can have an important role in tumour development, suggesting the need to analyse each mature sequence (miRNA-5p and -3p). In this study, binding sites were analysed for both mature miRNA sequences (-5p and -3p) in several interactions (Figure 6). A mature miRNA sequence, such as miR-548c, demonstrated binding sites in different cellular genes. Thus, this miRNA could serve as candidate biomarker for CC prognosis and diagnosis.\n\nHan et al. characterized the two mature chains of miR-21 and their oncogenic roles in cervical cancer73. The regulation of the mature 5p and 3p chains from several miRNAs has been investigated in other cancers, including colorectal, gastric, breast, lung, kidney, and bladder36,72,74–77, suggesting the need to focus further studies on the two mature chains from the 272 miRNAs reported in this study.\n\nFigure 6 shows the complexity of the interactions between miRNAs and tumour suppressor genes, proto-oncogenes and oncogenes. The study of interaction networks between cell cycle genes and miRNAs involved in cancer is one of the most recent challenges in systems biology and is important for elucidating the control mechanisms for cancer biological process78–81.\n\nThe differences in miRNA expression profiles between normal and cancerous tissues have led to the identification of clinical biomarkers for the early detection of many diseases, including various cancers and their precursor stages79,82,83. Research on miRNAs associated with cancer has not taken into account the genetic variability in human populations, which influences the structure, expression and function of miRNAs in populations from different ethnic backgrounds. Studies on genetic variability are relevant to designing strategies for the diagnosis and prognosis of various diseases.\n\nmiR-11-3p, miR-31-3p, miR-107, miR-133a-3p, miR-133a-5p, miR-133b, miR-215-5p, miR-491-3p, miR-548d-5p and miR-944 were conserved in the four human genome variants. In the remaining 27 miRNAs, substitutions, deletions or insertions were observed in the nucleotide sequences, indicating that this variability can be decisive when determining susceptibility to the development of CC (Table 7 and Supplementary File 3).\n\nThere are numerous studies that analyse miRSNPs in different malignancies84–86, but there is no available data on the correlation of SNPs in CC-associated miRNAs located in HPV integration sites in Latin American human genomic variants.\n\nAccording to our results, the genomes from Latin America showed a lower miRSNP frequency compared to the control genome (BEB), although the Colombian (CLM) genome frequency was more similar to the BEB genome. Latin American populations have experienced migrations from European, Asian and African individuals87. Thus, our results could be a result of the specific interracial mixing of Colombian populations but also due to migration patterns during human settlement in Latin America.\n\nmiRSNPs can affect the structure and function of miRNAs by impacting interactions between miRNAs and their mRNA targets or interfering with the expression levels of individual miRNAs20–22,88,89. miRSNPs could cause the loss or gain of binding sites for the co-evolution of miRNAs and their target mRNA and even influence cell processes related to tumour progression, disease phenotypes or susceptibility to developing a specific disease.\n\nMore studies are needed to clarify the role, targets and transcriptional regulatory mechanisms of cellular events in which miRNA are involved, including differentiation, apoptosis, metabolism and carcinogenesis. The expression and deregulation of miRNAs in cancer as well as their role as biological markers in diagnosis and treatment of CC should be explored. Further identification of cellular genes and signalling pathways involved in CC progression could lead to the development of new therapeutic strategies based on miRNAs90,91. Additional biomarkers associated with apoptosis, necrosis and possible interactions with CRISPR complex sequences can be explored in order to develop therapeutic strategies in the future.\n\n\nData availability\n\nDataset 1. The mature miRNA reference sequences were obtained in FASTA format from the miRBase database. DOI, 10.5256/f1000research.10138.d16473228\n\nDataset 2. Matrix of data containing all the necessary components for the validation of data on CC-associated miRNAs in HPV integration sites in Latin American human genomic variants. DOI, 10.5256/f1000research.10138.d16473636", "appendix": "Author contributions\n\n\n\nMGF directed all the research and bioinformatics analysis, wrote the article and made the final edits. OAGG developed the methodology and bioinformatics analysis and edited the article. JMH co-advised the research, wrote the article and made the final edits, and MCYC wrote the article.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgments\n\nThe authors thank the recommendations and suggestions of cPhD. Guillermo Torres from Kiel University (Germany) to improve the bioinformatics approach in this research.\n\n\nSupplementary material\n\nSupplementary File 1. Articles that mention HPV integration sites, detailing the most frequent types of HPV associated with CC.\n\nClick here to access the data.\n\nSupplementary File 2. Diagram indicating the regions on all chromosomes with miRNA binding sites that are associated with cervical cancer.\n\nClick here to access the data.\n\nSupplementary File 3. miRNAs identified in HPV integration sites, displaying the nucleotide variations in the selected Latin American human genome variants and in the control variant.\n\nClick here to access the data.\n\n\nReferences\n\nFerlay J, Soerjomataram I, Dikshit R, et al.: Cancer incidence and mortality worldwide: Sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer. 2015; 136(5): E359–86. PubMed Abstract \| Publisher Full Text\n\nBernard HU, Calleja-Macias IE, Dunn ST: Genome variation of Human Papillomavirus types: Phylogenetic and medical implications. Int J Cancer. 2006; 118(5): 1071–6. PubMed Abstract \| Publisher Full Text\n\nBurd EM: Human Papillomavirus and Cervical Cancer. Clin Microbiol Rev. 2003; 16(1): 1–17. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRichardson H, Kelsall G, Tellier P: The Natural History of Type-specific Human Papillomavirus Infections in Female University Students. Cancer Epidemiol Biomarkers Prev. 2003; 12(6): 485–90. PubMed Abstract\n\nWoodman CB, Collins SI, Young LS: The natural history of cervical HPV infection: unresolved issues. Nat Rev Cancer. 2007; 7(1): 11–22. PubMed Abstract \| Publisher Full Text\n\nWentzensen N, Vinokurova S, von Knebel Doeberitz M: Systematic review of genomic integration sites of human papillomavirus genomes in epithelial dysplasia and invasive cancer of the female lower genital tract. Cancer Res. 2004; 64(11): 3878–84. PubMed Abstract \| Publisher Full Text\n\nPett M, Coleman N: Integration of high-risk human papillomavirus: a key event in cervical carcinogenesis? J Pathol. 2007; 212(2): 356–67. PubMed Abstract \| Publisher Full Text\n\nNambaru L, Meenakumari B, Swaminathan R, et al.: Prognostic significance of HPV physical status and integration sites in cervical cancer. Asian Pac J Cancer Prev. 2009; 10(3): 355–60. PubMed Abstract\n\nSchmitz M, Driesch C, Jansen L, et al.: Non-random integration of the HPV genome in cervical cancer. PLoS One. 2012; 7(6): e39632. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBartel DP: MicroRNAs: Genomics, Biogenesis, Mechanism, and Function. Cell. 2004; 116(2): 281–97. PubMed Abstract \| Publisher Full Text\n\nRodriguez A, Griffiths-Jones S, Ashurst JL, et al.: Identification of mammalian microRNA host genes and transcription units. Genome Res. 2004; 14(10A): 1902–10. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCalin GA, Sevignani C, Dumitru CD, et al.: Human microRNA genes are frequently located at fragile sites and genomic regions involved in cancers. Proc Natl Acad Sci U S A. 2004; 101(9): 2999–3004. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBartel DP: MicroRNA: Target Recognition and Regulatory Functions. Cell. 2009; 136(2): 215–33. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSharma G, Dua P, Agarwal SM: A Comprehensive Review of Dysregulated miRNAs Involved in Cervical Cancer. Curr Genomics. 2014; 15(4): 310–23. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAdams BD, Kasinski AL, Slack FJ: Aberrant regulation and function of MicroRNAs in Cancer. Curr Biol. 2014; 24(16): R762–76. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMelo SA, Esteller M: Dysregulation of microRNAs in cancer: Playing with fire. FEBS Lett. 2011; 585(13): 2087–99. PubMed Abstract \| Publisher Full Text\n\nCammaerts S, Strazisar M, De Rijk P, et al.: Genetic variants in microRNA genes: Impact on microRNA expression, function, and disease. Front Genet. Frontiers Media SA; 2015; 6: 186. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSudmant PH, Rausch T, Gardner EJ, et al.: An integrated map of structural variation in 2,504 human genomes. Nature. 2015; 526(7571): 75–81. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\n1000 Genomes Project Consortium, Abecasis GR, Altshuler D, et al.: A map of human genome variation from population-scale sequencing. Nature. 2010; 467(7319): 1061–73. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiu C, Rennie WA, Carmack CS, et al.: Effects of genetic variations on microRNA: Target interactions. Nucleic Acids Res. 2014; 42(15): 9543–52. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTorruella-Loran I, Laayouni H, Dobon B, et al.: MicroRNA Genetic Variation: From Population Analysis to Functional Implications of Three Allele Variants Associated with Cancer. Hum Mutat. 2016; 37(10): 1060–73. PubMed Abstract \| Publisher Full Text\n\nWu M, Jolicoeur N, Li Z, et al.: Genetic variations of microRNAs in human cancer and their effects on the expression of miRNAs. Carcinogenesis. 2008; 29(9): 1710–6. PubMed Abstract \| Publisher Full Text\n\nGuerrero A, Guerrero M: MicroRNAs asociados al Cáncer de Cuello Uterino y sus lesiones precursoras: Una revisión sistemática MicroRNAs associated with Cervical Cancer and its precursor lesions: A systematic Review. Rev Univ y Salud. 2016; 28(2): 1–26. Reference Source\n\nKozomara A, Griffiths-Jones S: miRBase: annotating high confidence microRNAs using deep sequencing data. Nucleic Acids Res. 2014; 42(Database issue): D68–73. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVan Peer G, Lefever S, Anckaert J, et al.: miRBase Tracker: keeping track of microRNA annotation changes. Database (Oxford). 2014; 2014: pii: bau080. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKozomara A, Griffiths-Jones S: miRBase: integrating microRNA annotation and deep-sequencing data. Nucleic Acids Res. 2011; 39(Database issue): D152–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHsu PW, Huang HD, Hsu SD, et al.: miRNAMap: genomic maps of microRNA genes and their target genes in mammalian genomes. Nucleic Acids Res. 2006; 34(Database issue): D135–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGuerrero Flórez M, Guerrero Gómez OA, Mena Huertas J, et al.: Dataset 1 in: Mapping of microRNAs related to cervical cancer in Latin American human genomic variants. F1000Research. 2017. Data Source\n\n1000 Genomes Project Consortium, Abecasis GR, Autonn A, et al.: An integrated map of genetic variation from 1,092 human genomes. Nature. 2012; 491(7422): 56–65. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nInternational HapMap Consortium: The International HapMap Project. Nature. 2003; 426(6968): 789–96. PubMed Abstract \| Publisher Full Text\n\nKarolchik D, Hinrichs AS, Kent WJ: The UCSC Genome Browser. Curr Protoc Bioinformatics. 2009; Chapter 1: Unit1.4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKarolchik D, Baertsch R, Diekhans M, et al.: The UCSC Genome Browser Database. Nucleic Acids Res. 2003; 31(1): 51–4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nApweiler R, Bairoch A, Wu CH, et al.: UniProt: the Universal Protein knowledgebase. Nucleic Acids Res. 2004; 32(Database issue): D115–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMagrane M; UniProt Consortium: UniProt Knowledgebase: a hub of integrated protein data. Database (Oxford). 2011; 2011: bar009. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nXia H, Li F, He T, et al.: Distribution of Mature MicroRNA on Its Precursor: A New Character for MicroRNA Prediction. Int J Inf Technol. 2005; 11(8). Reference Source\n\nGuerrero Flórez M, Guerrero Gómez OA, Mena Huertas J, et al.: Dataset 2 in: Mapping of microRNAs related to cervical cancer in Latin American human genomic variants. F1000Research. 2017. Data Source\n\nCrooks GE, Hon G, Chandonia JM, et al.: WebLogo: A sequence logo generator. Genome Res. 2004; 14(6): 1188–90. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKuo WT, Su MW, Lee YL, et al.: Bioinformatic Interrogation of 5p-arm and 3p-arm Specific miRNA Expression Using TCGA Datasets. J Clin Med. 2015; 4(9): 1798–814. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMuñoz N, Bravo LE: Epidemiology of cervical cancer in Colombia. Salud Publica Mex. 2014; 56(5): 431–9. PubMed Abstract \| Publisher Full Text\n\nAngulo A: Analisis bioinformatico de secuencias L1, E6, E7 de VPH de alto y bajo riesgo más frecuentes Latinoamerica. Universidad de Nariño; 2014.\n\nSanchez C, Suarez K, Yepez M, et al.: Infección por VPH en mujeres del municipio de Pasto, Colombia con resultados de citología normal. Rev Univ y Salud. 2013; 15(1): 7–21. Reference Source\n\nNicola SN: Tipificación del Virus del Papiloma Humano-VPH y su relación con características poblacionales y lesiones en Cáncer de Cuello Uterino en mujeres del Municipio de Pasto. Universidad de Nariño; 2014.\n\nBodelon C, Untereiner ME, Machiela MJ, et al.: Genomic characterization of viral integration sites in HPV-related cancers. Int J Cancer. 2016; 1(9): 1–11. PubMed Abstract \| Publisher Full Text\n\nSoto-De Leon SC, Camargo M, Sanchez R, et al.: Prevalence of infection with high-risk human papillomavirus in women in Colombia. Clin Microbiol Infect. 2009; 15(1): 100–2. PubMed Abstract \| Publisher Full Text\n\nKraus I, Driesch C, Vinokurova S, et al.: The majority of viral-cellular fusion transcripts in cervical carcinomas cotranscribe cellular sequences of known or predicted genes. Cancer Res. 2008; 68(7): 2514–22. PubMed Abstract \| Publisher Full Text\n\nThorland EC, Myers SL, Gostout BS, et al.: Common fragile sites are preferential targets for HPV16 integrations in cervical tumors. Oncogene. 2003; 22(8): 1225–37. PubMed Abstract \| Publisher Full Text\n\nDall KL, Scarpini CG, Roberts I, et al.: Characterization of naturally occurring HPV16 integration sites isolated from cervical keratinocytes under noncompetitive conditions. Cancer Res. 2008; 68(20): 8249–59. PubMed Abstract \| Publisher Full Text\n\nFerber MJ, Thorland EC, Brink AA, et al.: Preferential integration of human papillomavirus type 18 near the c-myc locus in cervical carcinoma. Oncogene. 2003; 22(46): 7233–42. PubMed Abstract \| Publisher Full Text\n\nHaws BT, Cui W, Persons DL, et al.: Clinical and Pathologic Correlation of Increased MYC Gene Copy Number in Diffuse Large B-Cell Lymphoma. Clin Lymphoma Myeloma Leuk. 2016; 16(12): 679–683. PubMed Abstract \| Publisher Full Text\n\nLee KS, Kwak Y, Nam KH, et al.: Favorable prognosis in colorectal cancer patients with co-expression of c-MYC and ß-catenin. BMC Cancer. 2016; 16(1): 730. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWolfer A, Wittner BS, Irimia D, et al.: MYC regulation of a “poor-prognosis” metastatic cancer cell state. Proc Natl Acad Sci U S A. 2010; 107(8): 3698–703. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRao PH, Arias-Pulido H, Lu XY, et al.: Chromosomal amplifications, 3q gain and deletions of 2q33-q37 are the frequent genetic changes in cervical carcinoma. BMC Cancer. 2004; 4(1): 5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWilting SM, Snijders PJ, Verlaat W, et al.: Altered microRNA expression associated with chromosomal changes contributes to cervical carcinogenesis. Oncogene. 2013; 32(1): 106–16. PubMed Abstract \| Publisher Full Text\n\nReshmi G, Chandra SS, Babu VJ, et al.: Identification and analysis of novel microRNAs from fragile sites of human cervical cancer: Computational and experimental approach. Genomics. 2011; 97(6): 333–40. PubMed Abstract \| Publisher Full Text\n\nPeter ME: Targeting of mRNAs by multiple miRNAs: the next step. Oncogene. 2010; 29(15): 2161–4. PubMed Abstract \| Publisher Full Text\n\nCarleton M, Cleary MA, Linsley PS: MicroRNAs and cell cycle regulation. Cell Cycle. 2007; 6(17): 2127–32. PubMed Abstract \| Publisher Full Text\n\nDevi KJ, Chakraborty S, Deb B, et al.: Computational identification and functional annotation of microRNAs and their targets from expressed sequence tags (ESTs) and genome survey sequences (GSSs) of coffee (Coffea arabica L.). Plant Gene. 2016; 6: 30–42. Publisher Full Text\n\nTrafton A: Shrinking tumors with an RNA triple-helix hydrogel glue. 2015; 1–3. Reference Source\n\nConde J, Oliva N, Atilano M, et al.: Self-assembled RNA-triple-helix hydrogel scaffold for microRNA modulation in the tumour microenvironment. Nat Mater. 2016; 15(3): 353–63. PubMed Abstract \| Publisher Full Text\n\nKertesz M, Iovino N, Unnerstall U, et al.: The role of site accessibility in microRNA target recognition. Nat Genet. 2007; 39(10): 1278–84. PubMed Abstract \| Publisher Full Text\n\nBulik-Sullivan B, Selitsky S, Sethupathy P: Prioritization of genetic variants in the microrna regulome as functional candidates in genome-wide association studies. Hum Mutat. 2013; 34(8): 1049–56. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSlezak-Prochazka I, Kluiver J, de Jong D, et al.: Cellular localization and processing of primary transcripts of exonic microRNAs. Wilusz CJ, editor. PLoS One. 2013; 8(9): e76647. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nConcepcion CP, Bonetti C, Ventura A: The microRNA-17-92 family of microRNA clusters in development and disease. Cancer J. 2012; 18(3): 262–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAltuvia Y, Landgraf P, Lithwick G, et al.: Clustering and conservation patterns of human microRNAs. Nucleic Acids Res. 2005; 33(8): 2697–706. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCai N, Wang YD, Zheng PS: The microRNA-302-367 cluster suppresses the proliferation of cervical carcinoma cells through the novel target AKT1. RNA. 2013; 19(1): 85–95. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIvashchenko A, Berillo O, Pyrkova A, et al.: The properties of binding sites of miR-619-5p, miR-5095, miR-5096, and miR-5585-3p in the mRNAs of human genes. Biomed Res Int. 2014; 2014: 720715. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIvashchenko A, Berillo O, Pyrkova A, et al.: The arrangements of the locations of miR-619, miR-5095, miR-5096 and miR-5585 binding sites in the human mRNAs. Recent Adv Biomed Chem Eng Mater Sci. 2014; 144–9. Reference Source\n\nIvashchenko A, Berillo O, Pyrkova A, et al.: The properties of binding sites of miR-619-5p, miR-5095, miR-5096, and miR-5585-3p in the mRNAs of human genes. Biomed Res Int. 2014; 2014: 720715. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchmitz M, Driesch C, Beer-Grondke K, et al.: Loss of gene function as a consequence of human papillomavirus DNA integration. Int J Cancer. 2012; 131(5): E593–602. PubMed Abstract \| Publisher Full Text\n\nDweep H, Sticht C, Gretz N: In-Silico Algorithms for the Screening of Possible microRNA Binding Sites and Their Interactions. Curr Genomics. 2013; 14(2): 127–36. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPalmero EI, de Campos SG, Campos M, et al.: Mechanisms and role of microRNA deregulation in cancer onset and progression. Genet Mol Biol. 2011; 34(3): 363–70. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChoo KB, Soon YL, Nguyen PN, et al.: MicroRNA-5p and -3p co-expression and cross-targeting in colon cancer cells. J Biomed Sci. 2014; 21(1): 95. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHan Y, Xu GX, Lu H, et al.: Dysregulation of miRNA-21 and their potential as biomarkers for the diagnosis of cervical cancer. Int J Clin Exp Pathol. 2015; 8(6): 7131–9. PubMed Abstract \| Free Full Text\n\nUchino K, Takeshita F, Takahashi RU, et al.: Therapeutic Effects of MicroRNA-582-5p and -3p on the Inhibition of Bladder Cancer Progression. Mol Ther. 2013; 21(3): 610–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMlcochova J, Faltejskova-Vychytilova P, Ferracin M, et al.: MicroRNA expression profiling identifies miR-31-5p/3p as associated with time to progression in wild-type RAS metastatic colorectal cancer treated with cetuximab. Oncotarget. 2015; 6(36): 38695–704. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMuti P, Sacconi A, Hossain A, et al.: Downregulation of microRNAs 145-3p and 145-5p is a long-term predictor of postmenopausal breast cancer risk: The ORDET prospective study. Cancer Epidemiol Biomarkers Prev. 2014; 23(11): 2471–81. PubMed Abstract \| Publisher Full Text\n\nLou C, Xiao M, Cheng S, et al.: MiR-485-3p and miR-485-5p suppress breast cancer cell metastasis by inhibiting PGC-1α expression. Cell Death Dis. 2016; 7(3): e2159. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWatanabe Y, Tomita M, Kanai A: Computational Methods for MicroRNA Target Prediction. Methods Enzymol. 2007; 427: 65–86. PubMed Abstract \| Publisher Full Text\n\nPritchard CC, Cheng HH, Tewari M: MicroRNA profiling: approaches and considerations. Nat Rev Genet. 2012; 13(5): 358–69. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWang N, Xu Z, Wang K, et al.: Construction and analysis of regulatory genetic networks in cervical cancer based on involved microRNAs, target genes, transcription factors and host genes. Oncol Lett. 2014; 7(4): 1279–83. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYin Y, Song M, Gu B, et al.: Systematic analysis of key miRNAs and related signaling pathways in colorectal tumorigenesis. Gene. 2016; 578(2): 177–84. PubMed Abstract \| Publisher Full Text\n\nHayes J, Peruzzi PP, Lawler S: MicroRNAs in cancer: Biomarkers, functions and therapy. Trends Mol Med. 2014; 20(8): 460–9. PubMed Abstract \| Publisher Full Text\n\nMa Q, Wan G, Wang S, et al.: Serum microRNA-205 as a novel biomarker for cervical cancer patients. Cancer Cell Int. 2014; 14: 81. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMu W, Zhang W: Bioinformatic Resources of microRNA Sequences, Gene Targets, and Genetic Variation. Front Genet. 2012; 3: 31. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMi Y, Wang L, Zong L, et al.: Genetic variants in microRNA target sites of 37 selected cancer-related genes and the risk of cervical cancer. PLoS One. 2014; 9(1):e86061. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHu Y, Yu CY, Wang JL, et al.: MicroRNA sequence polymorphisms and the risk of different types of cancer. Sci Rep. 2014; 4: 3648. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHomburger JR, Moreno-Estrada A, Gignoux CR, et al.: Genomic Insights into the Ancestry and Demographic History of South America. PLoS Genet. 2015; 11(12): e1005602. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBhartiya D, Scaria V: Genomic variations in non-coding RNAs: Structure, function and regulation. Genomics. 2016; 107(2–3): 59–68. PubMed Abstract \| Publisher Full Text\n\nRawlings-Goss RA, Campbell MC, Tishkoff SA: Global population-specific variation in miRNA associated with cancer risk and clinical biomarkers. BMC Med Genomics. 2014; 7(1): 53. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAhmad J, Hasnain SE, Siddiqui MA, et al.: MicroRNA in carcinogenesis & cancer diagnostics: a new paradigm. Indian J Med Res. 2013; 137(4): 680–94. PubMed Abstract \| Free Full Text\n\nLiu Z, Sall A, Yang D: MicroRNA: An emerging therapeutic target and intervention tool. Int J Mol Sci. 2008; 9(6): 978–99. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "23646", "date": "11 Jul 2017", "name": "Juan Manuel Anzola", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this work, Guerrero et al. use mature microRNA in order to detect possible targets of these microRNAs in the human genome, and its population variants, including from Latin American, in order to determine possible associations with cervical cancer.\n\nI found the paper sound and its results, analysis and conclusions within the reach of the methodology, however I find the methods lacking, in particular when it comes to the parameters used in the BLAT search. BLAT uses a default seed of 11 to do nucleotide searches (they call it tileSize). So it would be good if the authors state clearly what were the BLAT parameters used, in particular \"tileSize\" and \"stepSize\". If a 11-word was used for this analysis the authors are running the risk of not being sensistive enough in their searches. High Specificity, Low Sensitivity. It would be interesting to determine how many of the genes reported as being targets for microRNAs are not detected in your search.\n\nmicroRNA have a particular set of rules when it comes to binding to their respective targets, with seeds between 6, 8 or 9 nucleotides. Nothing is stated in the paper to give an idea of how the rules for target detection were used in this paper. See Mullany et al paper.\nIt is assumed throughout the paper that all the hits are true positives. There is no measure as to how good is BLAT to detect true vs false positives.\nThe paper: In your introduction you mention that microRNAs are involved in cancer. The paragraph suggest this is the only role of microRNAs, however they are involved in processes such as development and morphogenesis, so please rephrase this paragraph because cancer is not the only role of microRNAs.\nFigure 7D is better represented as percentage, as in the body of the paper.\nYour phrase: \"Because some chromosomes have a greater number of miRNA binding sites, it provides evidence of a non-random distribution of miRNAs within the chromosomes.\" could be the result of chromosome length. Please provide statistical support for your statement.\nPage 14: Not all Pre microRNAs produce mature ones from both strands, in fact in the great majority of cases is only one strand that produces the mature one.\nThe paper will be ready for indexing once these observations are addressed.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "3925", "date": "18 Sep 2023", "name": "Milena Guerrero", "role": "Author Response", "response": "We resubmitted the second version of paper after addressing the various concerns raised. We would like to thank you for their time and for their constructive comments to help assist us in improving the manuscript. We made the necessary changes in order to address all the specified concerns. The direct responses to the reviewer’s comments are listed below: Reviewer. JMA Q1. I found the paper sound and its results, analysis and conclusions within the reach of the methodology, however I find the methods lacking, in particular when it comes to the parameters used in the BLAT search. BLAT uses a default seed of 11 to do nucleotide searches (they call it tileSize). So it would be good if the authors state clearly what were the BLAT parameters used, in particular \"tileSize\" and \"stepSize\". If a 11-word was used for this analysis the authors are running the risk of not being sensitive enough in their searches. High Specificity, Low Sensitivity. It would be interesting to determine how many of the genes reported as being targets for microRNAs are not detected in your search. R1: Thanks for your valuable suggestion. The methodology is re writer. In fact, BLAT only works with tile size 11. This mean that the average total length of mature miRNAs around 16 to 22, and consequently the seed sequence surely is represented at least 50% in mapping with this number of nucleotides. About suggestion “to determine how many of genes reports as being for miRNAs are not detected in your search”, we did the search, using programming R. And similar results of reported here we obtained. We not include this new focus on this paper, but If is need, we can send one of R mapping obtained for one chromosome. Q2: microRNA have a particular set of rules when it comes to binding to their respective targets, with seeds between 6, 8 or 9 nucleotides. Nothing is stated in the paper to give an idea of how the rules for target detection were used in this paper. See Mullany et al paper. R2: Thanks for the valuable comment. According to Mullany one of the most important “rules” for binding to mRNA and the role for cancer are length of seed sequence of miRNA. This condition is mentioned in second paragraph of introduction. Despite of this, the analysis no mentioned, but the authors analyzed seed sequences of miRNAs in terms of folding miRNAs (loop, folk, stem), length, 5´UTR extreme, the results was not included for this publication, because is part to another analysis. Q3: It is assumed throughout the paper that all the hits are true positives. There is no measure as to how good is BLAT to detect true vs false positives. R3: Thanks for your valuable annotation. Considering this probability, after that, we use R and bioconductor tools in order to be sure about the mapping results, we found a match between BLAT and R mapping. This data are under analysis ongoing. The paper: Q4. In your introduction you mention that microRNAs are involved in cancer. The paragraph suggest this is the only role of microRNAs, however they are involved in processes such as development and morphogenesis, so please rephrase this paragraph because cancer is not the only role of microRNAs. R4. Correct. The text is re write. Q5. Figure 7D is better represented as percentage, as in the body of the paper. R5. Thanks for the valuable suggestion. The figure was modified, highlighting percentages instead numeric values. File of figures. Q6. Your phrase: \"Because some chromosomes have a greater number of miRNA binding sites, it provides evidence of a non-random distribution of miRNAs within the chromosomes.\" could be the result of chromosome length. Please provide statistical support for your statement. R6: Thanks for the valuable suggestion. The authors included the statistical analysis and confirmed the results, for more clarity the paragraph is re-writing as follow: “In order to confirm the distribution of miRNA binding sites, the analysis for each chromosomal following all chromosomes was done. The statistic W Shapiro-Wilk test, show a p-value 0.02; and the mean comparison analysis by ANOVA with a p-value 0.0046 allowed us to confirm the non-random distribution of miRNA binding sites along the genome”. Q7. Page 14: Not all Pre microRNAs produce mature ones from both strands, in fact in the great majority of cases is only one strand that produces the mature one. R7: Thanks for the suggestion. It was adjusted." } ] }, { "id": "24293", "date": "28 Sep 2017", "name": "Subhash Mohan Agarwal", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn the present study the authors have mapped the miRNA involved in cervical cancer on to Latin American genome using in silico predictions. As cervical cancer has the highest mortality rates in low and middle income countries we do need to advance our understanding on mechanism of its progression. It is an interesting study however, there are few shortcomings in the current MS which need to be addressed.\nIt is not clear how human genes near to viral insertion sites have been identified. It was observer that near integration sites mostly only one or two genes are present. The method and parameters used for finding the genes should be detailed so that the results are reproducible. For example have the genes been identified within a particular distance of the insertion sites. Why the authors have mapped the integration sites for 8 types of HPVs collectively and not HPV-16 and 18 alone which are the high risk HPV. Is there any basis for it? The authors have stated that a total of 2028 miRNA binding sites of which 432 were detected in miRBase. In my opinion the analysis should have been restricted to only these sites as they are experimentally identified sites for miRNA binding. As I understand the authors have mapped 42 miRNAs on Latin American genome. It is not clear how 42 miRNAs were selected for this subsequent step.\nMinor comments:\nIn the supplementary data the headings of the tables should be in English. Are there 578 or 568 integration sites. It appears from Dataset 2 that there are 568 integration sites. Sheet named \"VPH integration sites\" Page 4 (last 2 lines) instead of 12 it should be 13. As per the data in figure 3 there are 13 genes in the intermediate category. Methods in Abstract: miRNA sequences associated with CC ……were obtained from miRBase. Shouldn’t it be literature?\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "3924", "date": "18 Sep 2023", "name": "Milena Guerrero", "role": "Author Response", "response": "Dear Reviewer SMA. We resubmitted the second version of the paper after addressing the various concerns raised. We would like to thank for their time and for their constructive comments to help assist us in improving the manuscript. We made the necessary changes in order to address all the specified concerns. The direct responses to the reviewer’s comments are listed below Reviewer. SMA Q8. It is not clear how human genes near to viral insertion sites have been identified. It was observer that near integration sites mostly only one or two genes are present. The method and parameters used for finding the genes should be detailed so that the results are reproducible. For example have the genes been identified within a particular distance of the insertion sites. R8: Thanks for your valuable suggestion. In a previous work (published in Spanish) it was described Guerrero & Guerrero, 2016. We analyze 42 scientific reports including chromosomal bands, HPV genotype, molecular technique for experimental results, and expression profile of miRNA according to lesions in CC. Here, we use the description and annotation of genes described in Data Bases, identified positions, regions, chromosomes, sequences and other characteristics to limit the regions of IS. All was manually curated for each chromosome along genome. Q9: Why the authors have mapped the integration sites for 8 types of HPVs collectively and not HPV-16 and 18 alone which are the high risk HPV. Is there any basis for it? R9: Thanks for your valuable comment. In fact, our reason were the results of two of our previous studies about genotyping of HPV in our region, which showed remarkable frequency, in addition to HPV16 and HPV 18, to other as HPV 45, 31, 33, 58, 67, 68. That was the main reason to include these genotypes in our analysis of mapping. But also by scientific purposes, because, in the literature is highly frequent to find and to study HPV 16 and 18, but not other genotypes. This information is relevant for us, to understand more depth the natural history of HPV and its mechanisms. Sánchez et al., 2013. Published in Spanish Nicola N, 2014. Bachelor thesis published in Spanish Q10: The authors have stated that a total of 2028 miRNA binding sites of which 432 were detected in miRBase. In my opinion the analysis should have been restricted to only these sites as they are experimentally identified sites for miRNA binding. R10: Thanks for your valuable comment. To our knowledge, the 2028 binding sites to miRNA and its key role in cervical cancer were identified by the first time in this study, using BLAT mapping. Is an important finding to highlight, compared to 432 binding sites previously reported, and a valuable contribution of bioinformatic tools for this kind research. Q11: As I understand the authors have mapped 42 miRNAs on Latin American genome. It is not clear how 42 miRNAs were selected for this subsequent step. R11: Thanks for your valuable comment. Briefly the pipeline was: -BLAT mapping of miRNAs on reference genome -Identification of integration sites (IS) of HPV -From IS HPV, looking for positions of genes of cell cycle, near to these IS. -Functional analysis of IS HPV according to annotations described by Uniprot. -With positions of near genes (regulators of cell cycle) and the positions of binding sites of miRNAs, manual mapping for each chromosome was done. -Finally, miRNAs in proximity to cell cycle genes control, were identified. Minor comments: Q12. In the supplementary data the headings of the tables should be in English. R12: Thanks for your detailed comment. It was corrected. Q13. Are there 578 or 568 integration sites. It appears from Dataset 2 that there are 568 integration sites. Sheet named \"VPH integration sites\" R13: Correct. There are 568 HPV integration sites. It was corrected. Q14. Page 4 (last 2 lines) instead of 12 it should be 13. As per the data in figure 3 there are 13 genes in the intermediate category. R14: Correct. Are 13 genes. It is corrected. Q15. Methods in Abstract: miRNA sequences associated with CC ……were obtained from miRBase. Shouldn’t it be literature? R15: Correct. The abstract is re written to include and express data in a clear way." } ] } ]	1	https://f1000research.com/articles/6-946
https://f1000research.com/articles/7-1893/v1	04 Dec 18	{ "type": "Research Article", "title": "Wearable technology and health: A bibliometric analysis using SciMAT", "authors": [ "Marlon Felipe Burbano-Fernandez", "Gustavo Ramirez-Gonzalez" ], "abstract": "Background: To begin research with cross-thematic topics, an analysis is necessary that collects data and historical accounts about these topics. In the field of health and technology, this principle is not unknown. In this case, it is the interest of this investigation to inquire about the topics \"wearable\" and \"health\" in order to gather enough information to establish how it has been developed, what the current topics are and what areas it is planned to study in a near future. The purpose of this study is to analyse the growth of the wearable technology and health subject matter. This will create an account of the most relevant themes over time in this subject and their evolution. This study will have an emphasis on recent years, in order to see which themes may be important for future research and publications. Methods: This paper aims at making a bibliometric exploration using information from the Scopus database using the search chain “weareable AND health” within the years of 2000 to 2017. The bibliometric software SciMAT was used for data visualization and analysis. Results: The results obtained are an analysis of the growth rate that is presented in Scopus publications between 2000 and 2017. In addition, a bibliometric analysis is obtained through the SciMat software, which shows the relevant issues that arise through time and that may be relevant for future research. Conclusions: Publications associated with the search chain “Wearable AND Health” from 2000 to 2017 were generally in constant growth over the years. With the SciMAT software, the main theme that was found was that of monitoring systems. In addition, it shows how new themes that are highly relevant emerge, such as mobile technologies and the study of algorithms, which emerge as an evolution from signal processing.", "keywords": [ "Wearable", "Health", "SciMAT", "Scopus" ], "content": "Introduction\n\nWearable technology, termed wearables, are increasingly becoming an essential part of activity tracking for people. They are part of the Internet of Things (Miorandi et al., 2012) and are characterized by their capability of being worn by the user, who can, at any moment, give or execute commands, even if the user is in movement or performing daily activities (Munoz-Organero & Lotfi, 2016). Vandrico Inc. states that the ideal wearable device must meet the following characteristics: be wearable, i.e. it must be used on the person’s body; controllable, i.e. the device must be controlled by the user actively or passively; offer evolution, i.e. the device must provide knowledge, make learning easier or make the experience more enjoyable; mobile, i.e. the device must give the users the freedom to act naturally without being limited to a fixed area.\n\nAccording to these characteristics, it is logical to think that wearables can be used to take care of health by making it possible for the data to flow between people and health professionals to the point of making monitoring and alerts automatic (Corti, 2016). As an example, diabetes patients have been monitored through wearables as shown in recent (Lee et al., 2016) and former (Shichiri et al., 1986; Meyerhoff et al., 1993) studies. The former studies indicate that the idea of using wearables is not recent. Consequently, it is now time to perform a bibliometric compilation to observe the growth in publications throughout the years on this topic, looking at the geographical location where this topic is mentioned, who are its main authors and the main journals where it has been published, as well as its evolution over time. This will make it possible to have an idea of the topics to be investigated in the near future.\n\nIn this study, a bibliometric analysis was performed from data collected from the Scopus database between the years of 2000 and 2017.\n\n\nMethods\n\nThe data for this bibliometric study was extracted from the Scoups database, by using the search chain “wearable AND health” and downloading the information of the format .RIS. For the download, it must be done in blocks of 2000 documents (corresponding to the maximum selection in Scopus). Only the aforementioned search chain was used as a search criterion and no documents between the years 2000 and 2017 were discarded. The search was made in June 2018.\n\nThe downloaded data were entered into the SciMAT software (SciMAT-v1.1.04), which can be freely downloaded (http://sci2s.ugr.es/scimat/download.html). The data entered can be viewed with the option \"Knowledge base -> Documents -> Document manager\". In this case, in the first instance we find that there is a total of 6001 downloaded data. Using the Document manager view, the fields shown in an Excel spreadsheet were copied and pasted (In this case, Microsoft Office 2016 was used), and with the use of the option \"Conditional Format\" the repeated data were found by title. In these duplicate data by title, the documents that had the oldest year were excluded from the database compiled in SciMAT. From this search, 5696 documents were found after excluding duplicated documents.\n\nThen, with the SciMAT software, \"Group set -> Word -> Find similar word by plurals (automatic)\" was used in order to limit the number of words with the same meaning between plural and singular. Finally, through \"Knowledge base -> Words -> Words manager\" were joined words or acronyms that had the same meaning, for example \"ML\" and \"Machine Learning\". And others that were irrelevant because they are obvious were eliminated, for example, the word \"wearble\" and \"health\" is removed from the database.\n\nThe adapted database can be found in the file attached to this document. To enter, you must have SciMAT installed, and through the \"File -> Open Project\" option, access the \"AnalysisScimat\" file.\n\nThis data was visualized with the software SciMAT, using a filtering process to avoid duplicity in similar documents and words. A growth analysis was done to see the number of articles published per year. Finally, SciMAT’s analysis tool was used to make connections between different periods and different clusters of relevant topics throughout the years.\n\nFrom the documents found from Scopus, descriptions were made based on the data classified by the software SciMAT-v1.1.04 on three fundamental aspects: period of time between the years of 2000 and 2017. The journals where they were published; the authors who published under the topics of “wearable AND health”.\n\nThe SciMAT software was used for the bibliometric analysis in accordance to the indications of Cobo et al. (2012). The process has three modules. ‘Module to manage the knowledge base’. ‘Module to carry out the science mapping analysis’. And ‘Visualization module’. In the ‘Module to manage the knowledge base’ build the know base, importing files .ris, edited entities, quit duplicated documents and was defined periods. In ‘Module to carry out the science mapping analysis’, was selected periods, selected unit of analysis selections, data reduction, kind of network, network reduction, normalization, clustering, document mappers, quality measures and longitudinal analysis. Finally, in Visualization module we did visualization and interpretation. Table 1 shows the selected parameters.\n\n\nResults\n\nTable 2 shows four columns. In the first one is the published documents year. The second corresponds to the number of documents found in that year. The third column corresponds to the percentage of articles corresponding to each year. Finally, in the fourth column, you can find the growth rate of the previous year (the difference between the publications made in a year subtracted from the previous year). It can be observed there are two periods when there is decrease in the production of articles on the subject of wearables and health. In 2001, there were 3 less articles than in 2000; however, that is not relevant considering both years concentrated less than 1% of the total production of documents in Scopus. Between 2008 and 2009 the production declined by 9 documents, but it is necessary to take into account that in both years the production was 1% of the total, with 163 documents in 2009 and 172 in 2008. This implies that, even if the production of articles declined in general, this decline was not relevant. Table 1 also shows that the bibliographic production of studies on wearables and health is increasing.\n\n\nResults\n\nThe graphics generated according to the visualization module are shown in Figure 1–Figure 3. For this study, the evolution map, the strategic map for the periods 2015, 2016 and 2017, and the cluster associated 2017 maps were created (Cobo et al., 2012; Sánchez et al., 2014).\n\n(A) 2015, (B) 2016 and (C) 2017.\n\nFigure 1 shows the evolution map. This map shows that in the period of 2000–2005, only two themes were relevant: ‘Telemedicine’ and ‘Bluetooth’. In the period of 2006–2010, the term ‘Telemedicine’ converges mainly to ‘Monitoring system’ and the theme ‘Bluetooth’ converges to ‘Mobile technology’. In addition, in this period, new themes appear. In the period of 2011, the themes are reduced to ‘Monitoring system’ and ‘Health care’. In the periods of 2012 and 2013, the theme of ‘Monitoring system’ is mainly maintained. In 2014, the theme ‘Monitoring system’ was diversified into ‘Device’, ‘Accelerometer’, ‘Mobile Technology’ and ‘Health care’. In the period of 2015, the themes of ‘Devices’ and ‘Mobile Technology’ are maintained, and, besides, the themes of ‘wearable antennas’, ‘Signal processing’, ‘wearable devices’ and ‘Sensor’ appear. In the period of 2016, the themes ‘Devices’, ‘mobile technology’ are maintained; ‘signal processing’ evolves into the theme of ‘Algorithms’, while ‘Wearable antennas’ evolve into ‘Body sensor network’, and the theme of ‘wearable computers’ appears. Finally, in 2017, the theme of ‘Devices’ evolves into ‘Monitoring System’ and ‘Aged’; the themes of ‘Wearable computer’ and ‘Algorithms’ are maintained.\n\nIt is interesting to observe that, in the evolutionary analysis from 2000 to 2005, the themes were not very diverse. From 2006 to 2010, several themes of interest appear, but in the periods of 2011 to 2014 they again focus on two themes only. In these periods, the predominant theme is ‘Monitoring system’ and it evolves in the periods of 2015 and 2016 into the ‘Devices’ theme, and then finally, in the period of 2017, the topics of ‘Monitoring system’ and ‘Aged’ are resumed. In the period of 2014, the theme of ‘Mobile technology’ is resumed, which is maintained until the period of 2016, and in the period of 2017 the theme ‘Monitoring system’ evolves. In the period of 2015, the theme of ‘Signal processing’ becomes stronger, which evolves in the periods of 2016 and 2017 into the theme ‘Algorithms’.\n\nIn Figure 2, the Strategic maps for the years 2015, 2016 and 2017 are shown, where the themes that emerge and those that are in motor clusters according to their centrality and their density can be seen. In the period of 2015 it is observed that the topics that in the motor cluster region ‘Signal processing’ and ‘Devices’ appear. In the period of 2016, the theme of ‘Devices’ and ‘Review’ is mainly in the motor cluster region. In this period, the theme of ‘Algorithms’ is also at the center. Finally, in the period of 2017, the themes in the motor cluster are ‘Monitoring System’ and ‘Aged’.\n\nFor this case, the two themes that are found in the motor cluster regions for the period of 2017 were taken into account. In this case, it can be observed that the theme of ‘Monitoring system’ is related to themes that present potential for studies, such as ‘Health care’, ‘Mobile technology’ and ‘Wearable devices’, among other themes. It can be observed that ‘Monitoring system’ is the center of the cluster, but that all the other themes are related to each other.\n\nAnother theme that is found in the motor cluster region is the theme of ‘Aged’, which is the center of the graph. For this case, its related topics are focused on pilot studies, of prospective and feasibility. Other themes related to this theme are ‘Devices’ and ‘Wearable electronic devices’. In this case, it can be observed that ‘Aged’ is related to the other themes, but these do not relate much to each other.\n\n\nDiscussion\n\nIn the evolution diagram performed in this study, it can mainly be observed that the theme that transcends over time is that of monitoring systems, This shows that wearable devices have been thought of as health monitoring systems from the beginning of their development. Between 2016 and 2017, it can be seen that the use of devices in the theme of aged also becomes important. In addition, it can be observed that since 2014 mobile technologies have become very relevant and finally, in 2017, they evolve into monitoring systems. In any case, it can be seen that wearable devices in the health care area have mainly been used for monitoring systems.\n\nIn the strategic maps, what can be found in the evolution diagrams is corroborated, where the rising theme for 2017 is ‘Monitoring system’, which evolves from the theme ‘Devices’ in 2015 and 2016. In 2015, it can be observed that a relevant theme was ‘Signal processing’, which evolves into ‘Algorithms’ in 2016, but which in 2017 becomes a basic and transversal theme. This implies that, although there are many studies on these themes, future studies in this area may not necessarily have an impact.\n\nFinally, the two most relevant topics found for the cluster diagrams show that monitoring systems are the most relevant for possible studies, where the sub-themes of health care, mobile technology and wearable devices are the most relevant. To a lesser extent, it is interesting that there also are relations to the themes connected to accelerometers, procedures, signal processing and electrocardiography. On the other hand, not so relevant but equally important, it is possible to see that the theme of ‘Aged’ is important, whereas associated sub-themes are prospective, pilot and feasibility studies for the use of devices and wearables.\n\n\nConclusions\n\nAs a first conclusion, it could be observed that in Scopus, publications associated with the search chain “Wearable AND Health” from 2000 to 2017 were generally in constant growth over the years.\n\nIn the evolutionary map found with the SciMAT software, it was found that in the studied search chain, the main theme that was found was that of monitoring systems. In addition, it shows how new themes that are highly relevant emerge, such as mobile technologies and the study of algorithms, which emerge as an evolution from signal processing.\n\nIn the strategic map for the years of 2015, 2016 and 2017, it is corroborated that the transcendental themes are devices, which evolve into the themes of monitoring system and aged, which may be relevant for future studies.\n\nThe cluster diagram for the period of 2017 in the monitoring system and aged themes is shown. In the theme of monitoring system, the themes of health care, mobile technology and wearable device are shown as good proposals for future studies. Similarly, the themes of accelerometer, procedures, signal processing and electrocardiogram can also be good proposals, but to a lesser extent. It should be noted that, according to this diagram, there are combinations between all these themes. In the theme of ‘Aged’, the studies are more focused on prospective, pilot or feasibility studies for the use of devices and wearables, these being good alternatives for future studies.\n\n\nData availability\n\nF1000Research: Dataset 1. The SciMatAnalisis.zip folder contains the AnalisisScimat file, which must be opened with the SciMat program. This file contains the adjusted data set. The Analysis folder contains the file Analysis4, which must be opened with the SciMat software analysis tool. And the folder Analisis4HTML, contains the results of this analysis deployed in web pages, https://doi.org/10.5256/f1000research.15622.d224368 (Burbano-Fernandez, 2018).\n\nSciMat software is available from: https://sci2s.ugr.es/scimat/download.html, under a GPLv3 license.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nBurbano-Fernandez MF, Ramirez-Gonzalez G: Dataset 1 in: Wearable technology and health: A bibliometric analysis using SciMAT. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15622.d224368\n\nCobo MJ, López-Herrera AG, Herrera-Viedma E, et al.: SciMAT: A new science mapping analysis software tool. Journal of the Association for Information Science and Technology. 2012; 63(8): 1609–1630. Publisher Full Text\n\nCorti S: Impacto de dispositivos “wearables” en el monitoreo de la salud. 2016. Reference Source\n\nLee H, Choi TK, Lee YB, et al.: A graphene-based electrochemical device with thermoresponsive microneedles for diabetes monitoring and therapy. Nat Nanotechnol. 2016; 11(6): 566–572. PubMed Abstract \| Publisher Full Text\n\nMeyerhoff C, Bischof F, Mennel FJ, et al.: On line continuous monitoring of blood lactate in men by a wearable device based upon an enzymatic amperometric lactate sensor. Biosens Bioelectron. 1993; 8(9–10): 409–414. PubMed Abstract \| Publisher Full Text\n\nMiorandi D, Sicari S, De Pellegrini F, et al.: Internet of things: Vision, applications and research challenges. Ad Hoc Networks. 2012; 10(7): 1497–1516. Publisher Full Text\n\nMunoz-Organero M, Lotfi A: Human movement recognition based on the stochastic characterisation of acceleration data. Sensors (Basel). 2016; 16(9): pii: E1464. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSánchez MÁM, Herrera MD, Fernández AIL, et al.: Un análisis bibliométrico de la producción académica española en la categoría de Trabajo Social del \"Journal Citation Report\"/A bibliometric analysis of Spanish production of Social Work category according to the Journal Citation Report. Cuadernos de Trabajo Social. 2014; 27(2): 429. Publisher Full Text\n\nShichiri M, Asakawa N, Yamasaki Y, et al.: Telemetry glucose monitoring device with needle-type glucose sensor: a useful tool for blood glucose monitoring in diabetic individuals. Diabetes Care. 1986; 9(3): 298–301. PubMed Abstract \| Publisher Full Text" }	[ { "id": "41481", "date": "11 Dec 2018", "name": "Peter Kokol", "expertise": [ "Reviewer Expertise Software engineering", "intelligent systems", "bibliometrics", "data mining" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors performed a bibliometric analysis of the employment of wearable technology in health using the SciMat software. They showed that the literature production is increasing in the period 2000 - 2017. They found that “monitoring systems” is the main research theme. And that new themes like mobile technologies and the study of algorithms are emerging. The paper is interesting and is analysing two topics, which is becoming more and more important and popular.\nFirst, they describe the use of a not so well known bibliometric analysis tool, SciMat. In that respect I would advise that the authors clearly state why they decided to use it, instead of more known tools like VOSViewer. I would also propose that in the discussion they state the advantages and disadvantages of SciMat compared to VOSViewer and other bibliometric tools.\nSecond, the authors reveals the dynamics of literature production in quite a novel way, and identify most important topics of research. I would suggest that author add the most recent or cited references for each topic they identified. They should also extend the discussion with the implications for practice and further research from the point of maintaining health and development of IoT devices. They should also identify the research gaps. Some standard descriptive bibliometric analyses like geographical dispersion of research and most prolific journals, could also be helpful.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "41646", "date": "21 Dec 2018", "name": "Manuel Jesus Cobo", "expertise": [ "Reviewer Expertise Bibliometric and science mapping analysis." ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript presents a bibliometric analysis on wearable health, which is based on co-words networks. To do that, authors employed SciMAT.\n\nIn what follows, some comments and suggestion are listed.\nThe English is very poor. The authors should make a great effort to improve it, and describe in a clear manner their finding. I don’t understand why there are duplicate documents in the database. There are more than 300 duplicates items, that seems too much. If the authors download the data correctly from Scopus, it is impossible to get duplicates. On the other hand, the number of documents listed in the document manager is equal to the number of records returned by the search? The methodology seems like a SciMAT tutorial. Authors do not need to describe every single step with the software. There are two section with the same label. In the strategic diagram, there are some themes that don’t make sense. For instance, Algorithm, review, etc. This broad theme should be marked as stop group in SciMAT, then redo the analysis. The organization of the results section is strange. First, the authors should describe the periods (each strategic diagram), and then the evolution map. The references section must be improved. Authors should add citations to the methodology employed, to co-words analysis, bibliographic networks, etc.\n\nIs the work clearly and accurately presented and does it cite the current literature? No\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1893
https://f1000research.com/articles/7-1321/v1	20 Aug 18	{ "type": "Research Article", "title": "The root cause of Duchenne muscular dystrophy is the lack of dystrophin in smooth muscle of blood vessels rather than in skeletal muscle per se", "authors": [ "Nadesan Gajendran" ], "abstract": "Background: The dystrophin protein is part of the dystrophin associated protein complex (DAPC) linking the intracellular actin cytoskeleton to the extracellular matrix. Mutations in the dystrophin gene cause Duchenne and Becker muscular dystrophy (D/BMD). Neuronal nitric oxide synthase associates with dystrophin in the DAPC to generate the vasodilator nitric oxide (NO). Systemic dystrophin deficiency, such as in D/BMD, results in muscle ischemia, injury and fatigue during exercise as dystrophin is lacking, affecting NO production and hence vasodilation. The role of neuregulin 1 (NRG) signaling through the epidermal growth factor family of receptors ERBB2 and ERBB4 in skeletal muscle has been controversial, but it was shown to phosphorylate α-dystrobrevin 1 (α-DB1), a component of the DAPC. The aim of this investigation was to determine whether NRG signaling had a functional role in muscular dystrophy. Methods: Primary myoblasts (muscle cells) were isolated from conditional knock-out mice containing lox P flanked ERBB2 and ERBB4 receptors, immortalized and exposed to CRE recombinase to obtain Erbb2/4 double knock-out (dKO) myoblasts where NRG signaling would be eliminated. Myotubes, the in vitro equivalent of muscle fibers, formed by fusion of the lox P flanked Erbb2/4 myoblasts as well as the Erbb2/4 dKO myoblasts were then used to identify changes in dystrophin expression. Results: Elimination of NRG signaling resulted in the absence of dystrophin demonstrating that it is essential for dystrophin expression. However, unlike the DMD mouse model mdx, with systemic dystrophin deficiency, lack of dystrophin in skeletal muscles of Erbb2/4 dKO mice did not result in muscular dystrophy. In these mice, ERBB2/4, and thus dystrophin, is expressed in the smooth muscle of blood vessels allowing normal blood flow through vasodilation during exercise. Conclusions: Dystrophin deficiency in smooth muscle of blood vessels, rather than in skeletal muscle, is the main cause of disease progression in DMD.", "keywords": [ "Dystrophin", "DAPC", "Duchenne muscular dystrophy", "Smooth muscle", "Blood vessels", "Neuregulin", "ERBB2/4", "HER2/4" ], "content": "Introduction\n\nSignaling from neuregulin 1 (NRG), through its epidermal growth factor (EGF) family of receptors ERBB1-4, has major functions in several organs such as heart, breast, and nervous system including central and peripheral synapses. The role of NRG signaling in skeletal muscle has been controversial. To investigate signaling events in muscle fibers, myotubes formed by fusion of myoblasts, are routinely used as the in vitro equivalent of muscle fibers. We already reported that in myotubes, formed from C2C12 myoblasts, NRG signaling through ERBB2/4 heterodimeric receptors phosphorylated α-dystrobrevin 1 (α-DB1)1, one of the components of the dystrophin associated protein complex (DAPC). DAPC links the intracellular actin cytoskeleton to the extracellular matrix and is thereby thought to provide structural stability during muscular activity. DAPC, apart from containing dystrophin which is a 427 kDa protein, consists of several other proteins such as α- and β-dystrobrevins, dystroglycans, sarcoglycans, sarcospan, syntrophins, and laminins. At the neuromuscular synapse, the DAPC is also formed with utrophin, also a 427 kDa protein, instead of dystrophin. The phosphorylation of α-DB1 through NRG/ERBB signaling stabilized acetylcholine receptors (AChRs) at the neuromuscular synapse2. Duchenne and Becker muscular dystrophy (D/BMD) patients have mutation(s) in the dystrophin gene, resulting in the expression of a truncated dystrophin protein3–6. Taken together, the main body of research on DMD argues for the lack of dystrophin in skeletal muscle as the cause for DMD.\n\nIn mice, apart from muscular dystrophy, absence of dystrophin causes neuromuscular junction (NMJ) fragmentation7 similar to the NMJ fragmentation associated with a loss of NRG/ERBB signaling1. Lack of dystrophin, besides causing muscular dystrophies, results in cardiomyopathy8 and is also responsible for several disease states in the brain6. The importance of NRG signaling for normal cardiac development in mice was firmly established by the fact that ablation of NRG, ERBB4, or ERBB2 resulted in premature death during midgestation9–11. In cardiac muscle NRG/ERBB4 signaling is sufficient for cardiomyocyte proliferation and repair of heart injury12, but knowledge of the detailed signaling mechanisms and the target proteins through which this was achieved are lacking. The aim of this investigation was to identify the function of NRG/ERBB signaling in muscle and, as it phosphorylated α-DB1 in the DAPC complex, determine if it had a functional role in muscular dystrophy by identifying downstream signaling targets.\n\n\nMethods\n\nerbb2/4 dKO and loxP flanked erbb2/4 myoblasts (a kind gift from M. Courtet, and previously described1) were cultured on laminin-coated dishes (Roche) and upon reaching 70–80% confluency, were allowed to form myotubes by changing to differentiation media (2% horse serum, 1% penicillin/streptomycin (Sigma-Aldrich), DMEM (Sigma-Aldrich)).\n\nMyotubes from 10-cm culture dishes were harvested in 600 μl lysis buffer and protein complexes were immunoprecipitated as described previously1,13 with modifications. In brief, myotubes harvested in ice-cold lysis buffer (10 mM Na3PO4, pH 7.8, 150 mM NaCl, 5 mM EDTA, 1 mM EGTA, 1% Triton X-100, protease inhibitor mixture (Roche), and phosphatase inhibitors Pic1 and Pic2 (Sigma-Aldrich)) were homogenized in a Dounce homogenizer and incubated for 3 h at 4°C with protein G-coupled mouse monoclonal syntrophin antibody 1351 (4 μl/80 μl protein G beads, Abcam, catalog number ab11425). Beads were then washed in lysis buffer containing protease inhibitors, but without Triton X-100, resuspended in 3 x SDS loading buffer (150 mM Tris-HCl [pH 6.8], 300 mM dithiothreitol [added just before use], 6% SDS, 0.3% bromophenol blue, and 30% glycerol), and denatured (94°C, 3 min) before loading on an 6–8% gradient/0.8% bis-acrylamide SDS-PAGE gels buffered with Tris-glycine.\n\nGels were transferred onto PVDF membranes (Millipore) and subject to ECL (Thermo Fisher Scientific) development after incubation with primary and secondary antibodies. BSA (3%) was used as a blocking reagent. The following primary antibodies were used: rabbit anti-dystrophin (H300) polyclonal (diluted 1:400, catalog number sc-15376) and mouse anti-utrophin (55) monoclonal (diluted 1:400, catalog number sc-136116) were from Santa Cruz Biotechnology, Inc., mouse monoclonal anti-syntrophin 1351 (4 μl antibody/80 μl protein G beads for lysate from a 10 cm culture dish of myotubes, catalog number ab11425) from abcam and rabbit anti-α-syntrophin 259 (5 μg/ml for Westerns; a kind gift from Stanly C. Froehner and Marvin Adams, University of Washington, Seattle, WA). Goat anti-mouse IgG-HRP (catalog number sc-2005) and goat anti-rabbit IgG-HRP (catalog number sc-2004) secondary antibodies (Santa Cruz Biotechnology, Inc.) were used at a 1:5,000 dilution.\n\nRNA isolation and qPCR were performed as previously described14 and the 2−ΔΔCt method was used to analyze relative changes in gene expression. RNA from myotube cultures was isolated with TRIzol (Invitrogen) according to their protocol. DNase I (Promega) treatment and reverse transcription was performed on 1 µg total RNA with random primers and superscript reverse transcriptase from Invitrogen according to their protocol. cDNA was diluted 1:5 before use in qPCR, which was performed with SyBR Green mix (Applied Biosystems) using the Applied Biosystems StepOne machine with two-step PCR (60°C, 1 min and 95°C 15 s) for 40 cycles using the standard program. The quantitative PCR mix was prepared as follows: 12.5 μl SyBR Green mix, 2.5 μl of a 3 µM solution each of forward and reverse primer, 1 μl of diluted cDNA and made up to 25 µl total volume with sterile water. Each sample for real time PCR was done in triplicate and the mean of the resulting three values were taken. The following primers, designed to recognize exons with at least one intron in between for each primer pair, were used for dystrophin, utrophin, and rL8 amplifications: dystrophin forward, 5′-GATGATGAACATTTGTTAATCCAGC-3′ and reverse, 5′-CATATTCTGCTTGCAGATTCCTG-3′; utrophin forward, 5′-CTAAACTCCTGCGGCAGCAC-3′ and reverse, '-GTGTCAAGTGAGTAGCTCAATGC-3′ and rL8 (normalization gene) forward, 5′-ACTGGACAGTTCGTGTACTG-3′ and reverse, 5′-GCTTCACTCGAGTCTTCTTG-3′.\n\n\nResults\n\nWe reported previously that on western blots following immunoprecipitation, there were two isoforms of α-dystrobrevin1 associated with DAPC, a 75 kDa and 89 kDa protein1. We also demonstrated that ablation of ERBB2/4 receptors resulted in a lack of phosphorylation of the 75 kDa protein1. In addition, the amount of the 75 kDa protein detected on western blots, compared to the 89 kDa protein, was reduced in Erbb2/4 dKO myotubes. However in myotubes with intact ERBB2/4 receptors, such as in C2C12 myotubes, it was the other way around1 i.e. less 89 kDa protein. As absence of dystrophin in the DMD mouse model mdx, also resulted in a reduction in the 75 kDa protein compared to wild-type mice15, it was possible that the reduction in the 75 kDa protein in the Erbb2/4 dKO myotubes was due to a reduction in dystrophin. Furthermore the AChR fragmentation observed in mdx mice7, paralleled those seen in Erbb2/4 dKO mice1 raising the possibility that another reason for the observed destabilization of the AChR cluster in Erbb2/4 dKO myotubes could be due to the reduced levels of dystrophin in these myotubes. These observations taken together suggest that one of the targets of NRG/ERBB signaling is dystrophin. To investigate this, Erbb2/4 dKO myotubes derived from immortalized Erbb2/4 dKO myoblasts1 and myotubes formed from myoblasts, before transfection of myoblasts with Cre recombinase, containing loxP flanked Erbb2/4 genes were used. Both, ERBB2 and ERBB4 receptors were ablated to eliminate NRG signaling1 through these receptors in muscle, because ERBB4 receptors, apart from forming heterodimers, can also form homodimers and ERBB2 can heterodimerize with ERBB316. Furthermore, as cultured myotubes secrete NRG17, the external addition of NRG was not necessary, as demonstrated for phosphorylation of α-dystrobrevin 1 by NRG/ERBB1.\n\nThree independent experiments (Figure 1A) each using myotubes formed from a different myoblast clone containing loxP-flanked Erbb2/4 genes, clearly detected dystrophin (lanes 1–3). However dystrophin was not detected in Erbb2/4 dKO myotubes (lanes 4–6) where ERBB2/4 receptors were ablated after CRE recombination of loxP flanked Erbb2/4 genes. Utrophin on the other hand was detected in myotubes with and without ERBB2/4 receptors (Figure 1B) demonstrating that NRG/ERBB signaling selectively regulates dystrophin expression (Figure 2).\n\n(A and B), Western blots of immunoprecipitated proteins from myotubes with loxP flanked exons of Erbb2 and Erbb4 gene (lanes 1 to 3) and cre mediated knock-out of Erbb2 and Erbb4 genes (lanes 4 to 6) detected with dystrophin (A) and utrophin (B) antibodies. Lanes 1 to 3 and 4 to 6 each represent the same experiment performed independently and loaded on the same gel. The lower panel shows detection of syntrophin that served as a loading control. Immunoglobulin G (IgG) detected is the syntrophin antibody used for the immunoprecipitation. As the western blot in (A) was stripped of antibodies and used in (B), the loading control in (B) applies to both A and B. (C and D) qPCR data of dystrophin and utrophin levels in C2C12 myotubes, relative to Erbb2/4 dKO (Erbb2/4-/-) myotubes. Expression levels were normalized to ribosomal protein L8 (rL8) expression. This experiment was performed at least twice with similar results. This figure was previously published in a patent (Patent Link: WO 2017/036852 A1), but the copyright is the author’s own.\n\nUnder normal circumstances, NRG signaling through ERBB2/4 (HER2/4) receptors stimulates dystrophin expression, allowing the formation of a normal dystrophin-associated protein complex (DAPC). If either the dystrophin gene contains mutations or NRG signaling is blocked, then in the absence of functional dystrophin, a normal DAPC is not formed, resulting in various disease states such as dilated cardiomyopathy, Duchenne/Becker muscular dystrophy (D/BMD), and neuromuscular synapse instability.\n\nThe qPCR estimation of dystrophin and utrophin levels (Figure 1C, D) confirmed that dystrophin expression is absent in Erbb2/4 dKO (shown in Figure 1 as erbb2/4-/-) myotubes, as there was no detectable mRNA and signals in qPCR were only observed above threshold at about 34 cycles which is essentially detection of non-specific amplification or background signal, whereas detection of ribosomal protein L8, used to normalize expression of dystrophin and utrophin, was above threshold at about 22 cycles in Erbb2/4 dKO and C2C12 samples. Detection of dystrophin mRNA in C2C12 cells by qPCR confirmed that the primers used to amplify dystrophin functioned. It is only possible to conclude that dystrophin is present in C2C12 and the level cannot be estimated since the level of dystrophin in C2C12 was relative to that in Erbb2/4 dKO myotubes, for which essentially background non-specific values were obtained in qPCR due to the absence of dystrophin mRNA. Utrophin detection (Figure 1D), using the same RNA/cDNA preparation used for dystrophin detection, confirmed that the cDNA preparation from Erbb2/4 dKO myotubes used for dystrophin detection was intact. Utrophin expression was reduced to only less than half the amount (Figure 1D) in Erbb2/4 dKO myotubes compared to C2C12 myotubes which may be due to the lack of NRG signaling since NRG stimulates utrophin expression to some extent18. Myotubes formed from C2C12 myoblasts were used as a control for qPCR instead of myotubes containing loxP flanked Erbb2/4 genes (used for the western blots in Figure 1A, B) as the loxP-flanked Erbb4 gene19 is a hypomorph due to the insertion of the neo selection cassette. Hence C2C12 myotubes were used instead of Erbb2/4 dKO myotubes to exclude the possibility that levels of dystrophin mRNA may have been affected (NRG signals through ERBB2/4 to stimulate dystrophin expression and reduced Erbb4 expression may have affected this). This is not a problem for dystrophin protein detection (not estimation) in immunoprecipitated samples from myotubes containing loxP flanked Erbb2/4 genes. qPCR on Erbb2/4 dKO confirmed the absence of dystrophin expression (Figure 1C) as observed on the western blot (Figure 1A). Hence NRG/ERBB signaling is necessary for dystrophin expression.\n\n\nDiscussion\n\nEven though dystrophin is lacking in skeletal muscles of Erbb2/4 dKO mice, they do not show dystrophic symptoms20. The promoter used for CRE expression in generating Erbb2/4 dKO mice, the human skeletal actin promoter (HSA), is expressed in the striated muscles, skeletal and heart muscle15,21,22. Therefore ERBB2/4 and dystrophin levels in smooth muscle of blood vessels would not be affected, as CRE is not expressed in smooth muscle of Erbb2/4 dKO mice, allowing the formation of a normal functional DAPC. Hence smooth muscle of blood vessels in these mice allows for increased blood flow to skeletal and cardiac muscle during exercise. This is because neuronal nitric oxide synthase (nNOS) associating with dystrophin and generating nitric oxide (NO) that signals to soluble guanylate cyclase, generating cyclic guanosine 3’,5’-monophosphate (cGMP) in smooth muscle of blood vessels, causes vasodilation enabling exercise-induced increase of blood flow and thereby prevents muscle ischemia23. Thus the absence of obvious dystrophic symptoms in Erbb2/4 dKO skeletal muscle where there is a lack of dystrophin strongly suggests that the main cause of muscular dystrophy is not the lack of dystrophin in skeletal muscle per se but systemic lack of functional dystrophin, especially in smooth muscle of blood vessels, resulting in impaired sympatholysis and muscle ischemia during exercise23. This hypothesis is consistent with published data using phosphodiesterase type 5 (PDE5) inhibitors, which interfere with breakdown of NO by PDE5 and thereby prolong the half-life of cGMP, the target of NO24. This was demonstrated in mdx mice where PDE5 inhibition alleviates the dystrophic phenotype23 and also in DMD patients where PDE5 inhibition with either tadalafil or sildenafil treatment in Duchenne muscular dystrophy boys restored normal blood vessel function and blood flow during exercise24. Thus the muscle has an extensive vasculature to provide more oxygenated blood through vasodilation the absence of which would result in muscle ischemia, injury and fatigue during exercise.\n\nWe previously reported that the blockade of NRG signaling through ERBB2/4 receptors prevented phosphorylation of α-dystrobrevin1 and hence affected NMJ stability1. However ablation of ERBB2/4 receptors, and thus elimination of NRG signaling through them, results in a lack of dystrophin expression (Figure 1). Thus NRG/ERBB signaling maintains NMJ stability through at least two pathways, one where it phosphorylates α-dystrobrevin 11,2 and the other where it stimulates dystrophin expression and thereby allowing the formation of a DAPC that stabilizes acetylcholine receptors (Figure 2).\n\nNRG/ErbB signaling also induces cardiomyocyte proliferation and repairs heart injury12 and is essential for normal cardiac development9–11, whereas dystrophin deficiency does not impair cardiac development but does result in dilated cardiomyopathy (DCM)8. Since ERBB4 can heterodimerize with ERBB2 and a function blocking ERBB2 antibody treatment results in DCM in mice21 and DCM in cancer patients with a function blocking HER2 antibody treatment25, this is consistent with NRG/ERBB signaling stimulating dystrophin expression, since DMD patients and mdx mice, both lacking functional dystrophin, develop DCM (Figure 2). Hence the data presented here provides a mechanism for the reported beneficial effects of ERBB4 in repairing heart injury whereby NRG/ERBB2/4 signaling stimulates dystrophin expression. Thus NRG/ERBB carries out different functions in cardiac development and maintenance through different signaling targets, in the latter case through regulating dystrophin expression.\n\nIncreasing NRG signaling through ERBB2/4, especially in the smooth muscle of blood vessels, could be a way to increase truncated dystrophin expression in D/BMD patients. This would ameliorate dystrophic symptoms in those patients where the mutation in dystrophin does not affect association of nNOS26 and thereby enabling normal blood flow during exercise. Furthermore, as dystrophin is also present in all regions of the brain, being most abundant in the cerebellum27, and since NRG/ERBB signaling regulates dystrophin expression, levels of dystrophin could be the underlying cause of some of the disease states such as schizophrenia, associated with NRG/ERBB function in the brain.\n\n\nConclusions\n\nNRG signaling through ERBB2/4 receptors is necessary for stimulation of dystrophin expression. However, when ERBB2/4 receptors are lacking in skeletal muscle but expressed in smooth muscle, mice do not exhibit dystrophic symptoms demonstrating that lack of dystrophin expression in smooth muscle is the root cause of the onset of D/BMD.\n\n\nData availability\n\nDataset 1. Uncropped western blot images and raw Ct values from qPCR. DOI: https://doi/org/10.5256/f1000research.15889.d21462828.", "appendix": "Competing interests\n\n\n\nThe author declares that the figure and corresponding data reported in this paper was used as part of the information for filing a patent application on ERBB4 receptor modulators based on the discovery that NRG signaling through ERBB2/4, via the cleaved ERBB4 intracellular domain, stimulated dystrophin expression.\n\n\nGrant information\n\nThis work was supported by funds from the University of Basel, Basel, Switzerland.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nSchmidt N, Akaaboune M, Gajendran N, et al.: Neuregulin/ErbB regulate neuromuscular junction development by phosphorylation of α-dystrobrevin. J Cell Biol. 2011; 195(7): 1171–1184. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGrady RM, Akaaboune M, Cohen AL, et al.: Tyrosine-phosphorylated and nonphosphorylated isoforms of alpha-dystrobrevin: roles in skeletal muscle and its neuromuscular and myotendinous junctions. J Cell Biol. 2003; 160(5): 741–752. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nConte GGL: Scrofola del sistema muscolare. Annali Clinici dell’ Ospedale degl' Incurabili di Napoli. 1836; 2: 66–79.\n\nMeryon E: On Granular and Fatty Degeneration of the Voluntary Muscles. Med Chir Trans. 1852; 35: 73–84.1. PubMed Abstract \| Free Full Text\n\nDuchenne G: Recherches sur la paralyse musculaire pseudohypertrophique, ou paralysie myosclerosique. Arch Gen Med. 1868; 11: 2–25,179–209,305–21,421–43,552–88.\n\nBecker PE, Kiener F: [A new x-chromosomal muscular dystrophy]. Arch Psychiatr Nervenkr Z Gesamte Neurol Psychiatr. 1955; 193(4): 427–448. PubMed Abstract\n\nKong J, Anderson JE: Dystrophin is required for organizing large acetylcholine receptor aggregates. Brain Res. 1999; 839(2): 298–304. PubMed Abstract \| Publisher Full Text\n\nCox GF, Kunkel LM: Dystrophies and heart disease. Curr Opin Cardiol. 1997; 12(3): 329–343. PubMed Abstract\n\nMeyer D, Birchmeier C: Multiple essential functions of neuregulin in development. Nature. 1995; 378(6555): 386–390. PubMed Abstract \| Publisher Full Text\n\nGassmann M, Casagranda F, Orioli D, et al.: Aberrant neural and cardiac development in mice lacking the ErbB4 neuregulin receptor. Nature. 1995; 378(6555): 390–394. PubMed Abstract \| Publisher Full Text\n\nLee KF, Simon H, Chen H, et al.: Requirement for neuregulin receptor erbB2 in neural and cardiac development. Nature. 1995; 378(6555): 394–398. PubMed Abstract \| Publisher Full Text\n\nBersell K, Arab S, Haring B, et al.: Neuregulin1/ErbB4 signaling induces cardiomyocyte proliferation and repair of heart injury. Cell. 2009; 138(2): 257–270. PubMed Abstract \| Publisher Full Text\n\nKramarcy NR, Vidal A, Froehner SC, et al.: Association of utrophin and multiple dystrophin short forms with the mammalian M(r) 58,000 dystrophin-associated protein (syntrophin). J Biol Chem. 1994; 269(4): 2870–2876. PubMed Abstract\n\nGajendran N, Kapfhammer JP, Lain E, et al.: Neuregulin signaling is dispensable for NMDA- and GABAA-receptor expression in the cerebellum in vivo. J Neurosci. 2009; 29(8): 2404–2413. PubMed Abstract \| Publisher Full Text\n\nBlake DJ: Dystrobrevin dynamics in muscle-cell signalling: a possible target for therapeutic intervention in Duchenne muscular dystrophy? Neuromuscul Disord. 2002; 12 Suppl 1: S110–7. PubMed Abstract \| Publisher Full Text\n\nUrsini-Siegel J, Schade B, Cardiff RD, et al.: Insights from transgenic mouse models of ERBB2-induced breast cancer. Nat Rev Cancer. 2007; 7(5): 389–397. PubMed Abstract \| Publisher Full Text\n\nMeier T, Masciulli F, Moore C, et al.: Agrin can mediate acetylcholine receptor gene expression in muscle by aggregation of muscle-derived neuregulins. J Cell Biol. 1998; 141(3): 715–726. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKhurana TS, Rosmarin AG, Shang J, et al.: Activation of utrophin promoter by heregulin via the ets-related transcription factor complex GA-binding protein alpha/beta. Mol Biol Cell. 1999; 10(6): 2075–2086. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLong W, Wagner KU, Lloyd KC, et al.: Impaired differentiation and lactational failure of Erbb4-deficient mammary glands identify ERBB4 as an obligate mediator of STAT5. Development. 2003; 130(21): 5257–5268. PubMed Abstract \| Publisher Full Text\n\nEscher P, Lacazette E, Courtet M, et al.: Synapses form in skeletal muscles lacking neuregulin receptors. Science. 2005; 308(5730): 1920–1923. PubMed Abstract \| Publisher Full Text\n\nCrone SA, Zhao YY, Fan L, et al.: ErbB2 is essential in the prevention of dilated cardiomyopathy. Nat Med. 2002; 8(5): 459–465. PubMed Abstract \| Publisher Full Text\n\nMcCarthy JJ, Srikuea R, Kirby TJ, et al.: Inducible Cre transgenic mouse strain for skeletal muscle-specific gene targeting. Skelet Muscle. 2012; 2(1): 8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKobayashi YM, Rader EP, Crawford RW, et al.: Sarcolemma-localized nNOS is required to maintain activity after mild exercise. Nature. 2008; 456(7221): 511–515. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNelson MD, Rader F, Tang X, et al.: PDE5 inhibition alleviates functional muscle ischemia in boys with Duchenne muscular dystrophy. Neurology. 2014; 82(23): 2085–2091. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSparano JA: Cardiac toxicity of trastuzumab (Herceptin): implications for the design of adjuvant trials. Semin Oncol. 2001; 28(1 Suppl 3): 20–27. PubMed Abstract \| Publisher Full Text\n\nLai Y, Thomas GD, Yue Y, et al.: Dystrophins carrying spectrin-like repeats 16 and 17 anchor nNOS to the sarcolemma and enhance exercise performance in a mouse model of muscular dystrophy. J Clin Invest. 2009; 119(3): 624–635. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLidov HG, Byers TJ, Kunkel LM: The distribution of dystrophin in the murine central nervous system: an immunocytochemical study. Neuroscience. 1993; 54(1): 167–187. PubMed Abstract \| Publisher Full Text\n\nGajendran N: Dataset 1 in: The root cause of Duchenne muscular dystrophy is the lack of dystrophin in smooth muscle of blood vessels rather than in skeletal muscle per se. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15889.d214628" }	[ { "id": "38333", "date": "20 Sep 2018", "name": "Kay E. Davies", "expertise": [ "Reviewer Expertise muscular dystrophy" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper concludes that the root cause of DMD is the lack of dystrophin in smooth muscle blood vessels. The only result to support this is their use of ERBB2/ERBB4 knock out myotubes which are negative for dystrophin. In the abstract under results they imply that there are data from the double mutant mice but results are not presented. There is just one analysis in mytotubes showing dystrophin absence. In vivo, this signalling could well be compensated for and dystrophin may be expressed. In the original article on these mice (ref 20; Esther et al1), this signalling is compensated for by agrin in vivo at the NMJ. No comments are made about dystrophin being absent in the muscle of these animals nor their muscle phenotype .\nThus this report does not contain sufficient evidence for the conclusion which is based on one blot from a cell line.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [ { "c_id": "4002", "date": "25 Sep 2018", "name": "Nadesan Gajendran", "role": "Author Response", "response": "I thank the reviewer for taking the time to review the article and for the comments. I realise that the role of neuregulin in muscle, especially in transcription of synapse specific genes and NMJ formation has been controversial. The reviewer suggests that in vivo NRG/ErbB signalling could be compensated by another signalling pathway (when ERBB2/4 receptors are ablated) to stimulate dystrophin expression1 since, the reviewer comments, agrin compensates for this (NRG signalling through ERBB2/4) at the NMJ2. However, in the article that is referred to by the reviewer and that describes the ErbB2/4 dKO mice2, the aim was to determine if neuregulin (NRG) signaling through ERBB2/4 receptors in muscle was required for synapse specific transcription and NMJ formation. That Agrin plays an important role in acetylcholine receptor clustering and NMJ formation was demonstrated by the fact that mice carrying a mutation in the agrin gene lacked NMJs3. Thus the original article on the ErbB2/4 dKO mice concluded that NRG was not required for synapse-specific gene transcription at the NMJ and that development and maintenance of neuromuscular synapses were only marginally affected. Hence, the article does not conclude that NRG signalling is compensated for by Agrin, but rather that NRG is not required for synapse specific gene expression and NMJ formation. In addition, unlike NRG/ERBB signalling stimulating dystrophin expression in muscle1, transcription of synapse specific genes and the formation of the NMJ involves reciprocal interactions between muscle fibers and motor neurons. We subsequently demonstrated that neuromuscular synapses being marginally affected in the ErbB2/4 dKO mice was due to a lack of phosphorylation of alpha-dystrobrevin 1 (α-DB1)4. To demonstrate phosphorylation of α-DB1 by signalling through ERBB2/4 receptors, I used ErbB2/4 dKO myotubes (as I did in this article) as well as myotubes formed from C2C12 and primary myotubes formed from myoblasts from wild type (wt) mice (Schmidt etal. Fig. 7 a & b)4. We observed phosphorylation of α-DB1 in samples prepared from myotubes formed from freshly purified primary myoblasts from ErbB 2/4 dKO mice. This phosphorylation of α-DB1 could be blocked by ERBB inhibitors (Schmidt et al. Fig. 7a)4 demonstrating that: (i) signalling through ERBB2/4 receptors was responsible for phosphorylation of α-DB1, (ii) the observed phosphorylation of α-DB1 in myotubes formed from ErbB2/4 dKO primary myoblasts was likely caused by the presence of contaminating cells in the myoblast preparation. Contaminating cells in the myoblast preparation is obviously lost after extended culture periods as confirmed by the lack of phosphorylation of α-DB1 in ErbB2/4 dKO myotubes in the absence of any ERBB inhibitor (Schmidt et al. Fig. 7c)4 For the same reason I used the ErbB2/4 dKO myotubes instead of muscle tissue, to demonstrate a lack of dystrophin expression1. Once we discovered that signalling through ERBB2/4 receptors phosphorylated α-DB1 (Schmidt et al. Fig. 7 a-c)4 we realised that this explained the fragmentation of acetylcholine receptor (AChR) clusters at the neuromuscular synapse in ErbB2/4 dKO mice. This was subsequently confirmed by experiments in cultured myotubes (wild type and α-DB1 KO) as well as with and without α-DB1 expression constructs (Schmidt et al. Fig. 6, results were not presented in order of discovery)4. We could not show absence of α-DB1 phosphorylation in vivo in the muscles of ErbB2/4 dKO mice for the same reasons mentioned above. Due to the presence of tissues other than skeletal muscle in muscle lysates and cells other than myoblasts in primary myoblast purifications from ErbB2/4 dKO mice, it was not possible to resolve either the lack of phosphorylation of α-DB1 or the lack of dystrophin in the absence of ERBB2/4 signaling in the dKOs. In particular satellite cells present in muscle lysates (and in primary myoblast purifications) were shown to express a high level of dystrophin5 and will still express dystrophin in ErbB2/4 dKO mice since the HSA promoter driving CRE expression is not active6,and hence ERBB2/4 receptors would not be ablated in these cells. The juxtaposition of dystrophin expression in satellite cells and in myofibers makes it very difficult to distinguish dystrophin expression in satellite cells from that in muscle fibers5. It may well be possible that researchers who may have looked for the dystrophin expression in muscle specific ErbB2 and/or ErbB4 KO may have run into this problem and not observed a lack of dystrophin in muscle or muscle sections. Satellite cells, however, appear to have a limited potential to regenerate skeletal muscle6 and hence would not fully explain the lack of dystrophic symptoms in ErbB2/4 dKO mice. Dystrophin and utrophin were detected in loxP-flanked ErbB2/4 myotubes (Gajendran et al., Fig 1.)1 formed from myoblasts, even after the extended culture period to generate them, and it was only after transient CRE transfection and ablation of ErbB2/4 receptors (confirmation of recombination checked by PCR) dystrophin expression was not observed but utrophin was, in different clones. Understandably, it was not possible to either demonstrate a lack of phosphorylation of α-DB1 or the lack of dystrophin expression when using lysates of total muscle tissues due to the presence of other tissues and cells, including the smooth muscle of blood vessels and satellite cells where ErbB2/4 receptors were not ablated (the HSA promoter driving CRE expression is not active there). In addition to dystrophin detection, any other component of the DAPC present in vivo in the muscle that we wanted to detect in myotubes such as utrophin, a-dystrobrevin, and a-syntrophin were also detected on western blots following immunoprecipitation of the dystrophin associated protein complex (DAPC) . If the absence of NRG/ErbB2/4 signaling, and as a consequence absence of dystrophin in skeletal muscle, could be compensated in vivo by another mechanism, even possibly through another ligand other than NRG1, then this compensation mechanism would be independent of ERBB2/4 receptors as they are ablated in the skeletal muscles in these mice. From experiments using α-DB1 KO mytobes in combination with α-DB1 expression constructs, it also became clear that signaling from α-DB1 downregulates dystrophin expression but not utrophin expression (results available at WO 2017/036852 A1 under documents “09.03.2017 Initial Publication with ISR (A1 10/2017)”)7. Thus opposing signals from α-DB1 (downregulation) and NRG/ERBB2/4 (upregulation) regulates dystrophin expression. This regulation appears to be mediated by the cleaved intracellular domain of ERBB4 (4ICD) since increased dystrophin levels in dystrobrevin KO myotubes are accompanied by a strong upregulation of ERBB4 expression most of which is cleaved to generate 4ICD (see figures at WO 2017/036852 A1)7. Together with a wealth of published information, the data presented in this article, in my opinion, is essential and sufficient to conclude that expression of dystrophin is stimulated by NRG/ERBB signalling. It follows that in ErbB2/4 dKO mice that did not show dystrophic symptoms (see below), the absence of dystrophin in skeletal muscle but not in smooth muscle of blood vessels, clearly argues that the root cause of DMD is due to the lack of dystrophin in the smooth muscle of blood vessels. Naturally it is possible that dystrophin expression in the endothelial cells, between the smooth muscle and the lumen of blood vessels, in ErbB2/4 dKO mice also plays a role in vasodilation during exercise, assuming that the HSA promoter used to ablate ERBB2/4 receptors is not active in these cells. Published data on DMD, together with the data presented in this article, would suggest that the lack of dystrophin in skeletal muscle exacerbates the onset of dystrophy caused by a lack of increased blood flow during exercise. This observation is also supported by a very early observation in mdx muscles that only some fibers, and not all, in skeletal muscle initially show dystrophic symptoms8 and also by the fact that expressing the full length dystrophin in smooth muscles in mdx mice, even though it was expressed at significantly lower levels than the level of dystrophin found wild type mice, improved aberrant vasoregulation9. The fact that skeletal muscle in those mice still showed some atrophy could be due to insufficient dystrophin expression in the smooth muscle causing dystrophic symptoms as a consequence of reduced blood flow during exercise and exacerbated due to the lack of dystrophin in skeletal muscle (Reviewed in Ennen JP et al.)10. The exacerbation of dystrophic symptoms in skeletal muscle, apart from structural defects, may also be as a result of disruption of signalling events from the DAPC due to the absence of dystrophin and this is further supported by the observation that α-DB KO results in increased dystrophin and ERBB4 expression in myotubes7. As far as could be determined from the literature, there are no compensatory signals in vivo that can ameliorate the effects caused by deficient ERBB2 or ERBB4 signalling such as defective muscle spindle formation11, dilated cardiomyopathy (DCM) following Herceptin (antibody against HER2/ErbB2) treatment12, DCM in conditional ErbB213,14 or ErbB415 KO mice, AChR cluster disintegration (also observed in mdx mice) in muscle specific ErbB2/4 dKO mice4, and the embryonic lethality observed in ErbB2 or ErbB4 KO mice16,17. NRG and its ERBB receptors appear to exert their function in a spatio-temporal fashion in development and maintenance (in adults). The reviewer is correct that no comments were made about dystrophin in the ErbB2/4 dKO mice2. However in ErbB2/4 dKO mice nuclei in muscle fibers were examined extensively (see figures in supplemental data of that publication)2 and centralized nuclei, a hallmark of DMD18,19 were not observed. It was also reported in the supporting online text (supplemental data)2 that atrophic fibers could not be observed. Hence, based on these observations, there was no indication to look for dystrophin expression. It was also mentioned in the article that sustained muscle strength was not affected and that the ErbB2/4 dKO mice did not have a myasthenic condition. However these mice showed a hind limb extension reflex which was not reported in the article. I trust that I have addressed the reviewers concern sufficiently, and if the reviewer would be willing to review a revised version of this article, I would welcome the opportunity to submit one that would include some of the explanations/clarifications presented here. Bibliography 1. Gajendran, N. The root cause of Duchenne muscular dystrophy is the lack of dystrophin in smooth muscle of blood vessels rather than in skeletal muscle per se [version 1; referees: 1 not approved]. F1000Res. 7, 1321 (2018). 2. Escher, P. et al. Synapses form in skeletal muscles lacking neuregulin receptors. Science 308, 1920–1923 (2005). 3. Gautam, M. et al. Defective neuromuscular synaptogenesis in agrin-deficient mutant mice. Cell 85, 525–535 (1996). 4. Schmidt, N. et al. Neuregulin/ErbB regulate neuromuscular junction development by phosphorylation of α-dystrobrevin. J. Cell Biol. 195, 1171–1184 (2011). 5. Dumont, N. A. et al. Dystrophin expression in muscle stem cells regulates their polarity and asymmetric division. Nat. Med. 21, 1455–1463 (2015). 6. Nicole, S. et al. Intact satellite cells lead to remarkable protection against Smn gene defect in differentiated skeletal muscle. J. Cell Biol. 161, 571–582 (2003). 7. Gajendran, N. ERBB4 RECEPTOR MODULATORS FOR USE IN THE TREATMENT OF DYSTROPHIN-ASSOCIATED DISEASES. (2017). WO 2017/036852 A1 8. Engel, W. K. Muscle biopsies in neuromuscular diseases. Pediatr. Clin. North Am. 14, 963–995 (1967). 9. Ito, K. et al. Smooth muscle-specific dystrophin expression improves aberrant vasoregulation in mdx mice. Hum. Mol. Genet. 15, 2266–2275 (2006). 10. Ennen, J. P., Verma, M. & Asakura, A. Vascular-targeted therapies for Duchenne muscular dystrophy. Skelet Muscle 3, 9 (2013). 11. Leu, M. et al. Erbb2 regulates neuromuscular synapse formation and is essential for muscle spindle development. Development 130, 2291–2301 (2003). 12. Sparano, J. A. Cardiac toxicity of trastuzumab (Herceptin): implications for the design of adjuvant trials. Semin. Oncol. 28, 20–27 (2001). 13. Ozcelik, C. et al. Conditional mutation of the ErbB2 (HER2) receptor in cardiomyocytes leads to dilated cardiomyopathy. Proc. Natl. Acad. Sci. USA 99, 8880–8885 (2002). 14. Crone, S. A. et al. ErbB2 is essential in the prevention of dilated cardiomyopathy. Nat. Med. 8, 459–465 (2002). 15. García-Rivello, H. et al. Dilated cardiomyopathy in Erb-b4-deficient ventricular muscle. Am. J. Physiol. Heart Circ. Physiol. 289, H1153–60 (2005). 16. Lee, K. F. et al. Requirement for neuregulin receptor erbB2 in neural and cardiac development. Nature 378, 394–398 (1995). 17. Gassmann, M. et al. Aberrant neural and cardiac development in mice lacking the ErbB4 neuregulin receptor. Nature 378, 390–394 (1995). 18. Wali, V. B. et al. Overexpression of ERBB4 JM-a CYT-1 and CYT-2 isoforms in transgenic mice reveals isoform-specific roles in mammary gland development and carcinogenesis. Breast Cancer Res. 16, 501 (2014). 19. Wang, B., Li, J. & Xiao, X. Adeno-associated virus vector carrying human minidystrophin genes effectively ameliorates muscular dystrophy in mdx mouse model. Proc. Natl. Acad. Sci. USA 97, 13714–13719 (2000)." }, { "c_id": "4015", "date": "27 Sep 2018", "name": "Dame Kay Davies", "role": "Reviewer Response F1000Research Advisory Board Member", "response": "The author has responded with some reasonable arguments but the issue with the paper is lack of evidence. No new evidence is provided. In vivo rescue of the dystrophic phenotype in dystrophic deficient mice and dogs by delivery to skeletal muscle is a compelling argument for the role of dystrophin in skeletal muscle. If there is a role in smooth muscle, then this needs to be shown by in vivo experiments not a cell line." }, { "c_id": "4028", "date": "02 Oct 2018", "name": "Nadesan Gajendran", "role": "Author Response", "response": "I thank the reviewer for responding to my comments. Although the reviewer does not question the discovery described in the manuscript, that NRG signalling is required for dystrophin expression as demonstrated in myotube cultures, concerns remain regarding in vivo evidence to confirm that the root cause of muscular dystrophy is due to the lack of dystrophin in smooth muscle of blood vessels rather than in skeletal muscle. The reviewer’s concern is understandable given the vast number of publications that focused on the lack of dystrophin in skeletal muscle as the cause of muscular dystrophy. These reports are indeed in contrast to our observation that ErbB2/4 dKO mice don’t show dystrophic symptoms1. In these mice, NRG signalling, and as a consequence also dystrophin expression (based on our in vitro results/ new findings), is abolished in skeletal muscles but not in vascular smooth muscle (VSM). Ideally, to address the reviewers concern, a conditional transgenic mouse where either ERBB2 (or ERBB4) is knocked out in both skeletal and VSM should be examined to have a situation where dystrophin would be lacking in both. On the other hand it would also be good to demonstrate that dystrophin absence in skeletal muscle does not cause dystrophic symptoms, confirming our observations on the ErbB2/4 dKO mouse. I address below the reviewers comments. In vivo evidence for a role for dystrophin in smooth muscle: Mice where ErbB2 is conditionally knocked out in skeletal muscle, HSA-CRE/ErbB2, following CRE expression driven by the HSA promoter2 show a similar phenotype as ErbB2/4 dKO1 mice mentioned in this article (in both mouse strains the same HSA promoter is used to express CRE). As the HSA promoter is not active in smooth muscle, ERBB2 receptors would not be ablated and dystrophin will still be expressed in VSM, in contrast to skeletal muscles in these mice. The loxP flanked ErbB2 mouse is the same mouse that was used in breeding with an ERBB4 conditional KO mouse and an HSA-CRE transgenic to finally generate the ErbB2/4 dKO mouse1. The abnormal muscle spindle formation reported for the HSA-CRE/ErbB2 mouse was not looked at in the ErbB2/4 dKO mouse probably because it had already been described in the HSA-CRE/ErbB2 KO mouse. In addition the study describing the ErbB2/4 dKO mouse was focused on the role of NRG/ERBB signaling in synapse specific transcription and NMJ formation. Interestingly an ErbB2 conditional knock-out mouse with CRE expression driven by the muscle creatine kinase (MCK) promoter, MCK-CRE/ErbB2 was made that was different from the one mentioned above as it showed impaired muscle regeneration and a requirement of ERBB2 for survival of muscle spindles and myoblasts3. As VSM express both the brain and muscle isoforms of creatine kinase4, the MCK promoter would be active in VSM and hence ERBB2 would be ablated resulting in a loss of dystrophin expression in VSM in these mice, explaining the different histopathology characters between the two ErbB2 KO mice that differ in the promoter driving CRE expression. The muscle phenotype described for the MCK-CRE/ErbB2 KO mouse was not observed in the HSA-CRE/ERBB2 KO mouse where the HSA promoter is not active in VSM, and hence ErbB2 would not be ablated and therefore dystrophin would be expressed. MCK-CRE/ERBB2 KO mouse primary myoblasts lacking ErbB2, were reported to undergo extensive apoptosis when differentiating into myofibers3, an abnormality that I did not observe that when using ErbB2/4 dKO myotubes5. In another study, siRNA mediated silencing of dystrophin expression in the muscles of adult mice was meticulously analysed6. The authors concluded that in spite of the clear absence of dystrophin in the skeletal muscle, they did not observe any of the histopathology characters observed in mdx mice6. This paper describes silencing of dystrophin expression in skeletal muscles of adult mice, which results in a delay before the existing dystrophin is depleted, and the authors suggest that the dystrophic pathology observed in dystrophin deficiency may be developmentally regulated. However in the ERBB2/4 dKO mice, dystrophin would be ablated early during development in skeletal muscles since the HSA promoter driving CRE expression is active on E9.5, when dystrophin would also start to be expressed7, which would argue against the dystrophic pathology being developmentally regulated. In vivo rescue of the dystrophic phenotype in mice and dogs may be due to delivery of the therapeutic (construct, anti-sense oligoribonucleotides, gene editing components) into skeletal and vascular smooth muscle There have been numerous attempts to express dystrophin in skeletal muscle that ameliorated the dystrophic pathology to some extent, but none have led to applicable therapy. As the focus of most therapies are to restore skeletal, and sometimes also cardiac muscle it was not of concern whether VSM would or would not be targeted during therapy. Gene therapy using Rous sarcoma virus (RSV) or cytomegalovirus (CMV) promoters8 would also result in the expression of dystrophin, or smaller versions of it, in surrounding VSM after intra-muscular (IM) injection or particle bombardment9. In a Phase I study of gene therapy for Duchenne/Becker muscular dystrophy (D/BDM) the CMV promoter was also used10 and hence also here dystrophin would be restored in VSM. Systemic delivery of antisense oligoribonucleotides11, or restoration of dystrophin expression by (intravenous) stem cell transplantation12, would also result in dystrophin restoration in VSM. Although in vivo rescue of the dystrophic phenotype in dystrophin deficient mice and dogs focused on delivery of the construct to skeletal muscle, it would also target VSM and hence it is neither possible to exclude the role of VSM nor to conclude that delivery to skeletal muscle is sufficient, to rescue the dystrophic phenotype. Furthermore, simply expressing dystrophin in skeletal muscle fibers lacking dystrophin would restore the dystrophin associated protein complex (DAPC) without saying much about whether muscle function is restored. If indeed muscle function is restored, then it should be looked at whether dystrophin may also have been expressed in VSM through leaking of the therapeutic into those tissues, sometimes through the circulation especially when using viral based therapies as described below. Gene editing of mutated dystrophin using adeno-associated viruses (AAV) to deliver CRISPR components is a promising therapeutic strategy and has recently been demonstrated to function in a canine model of Duchenne muscular dystrophy13. But it still needs to be determined if the gene editing will be sustainable over long term. In this study an AAV serotype 9 that shows tropism for heart and skeletal muscle was used to deliver the CRISPR components. To drive the expression of Cas9 (one of the gene editing components), a muscle specific creatine kinase regulatory cassette was used that should also result in gene editing of dystrophin in VSM especially when the viruses are injected intravenously for systemic delivery. In a second experiment in the same study, 6 weeks following delivery through direct injection into muscles, dystrophin expression and assembly of the DAPC in dystrophic muscles was observed. Dystrophin expression in contralateral muscles that were not injected was also observed and it was attributed to leakage of AAV9 into circulation. In contrast to the siRNA silencing of dystrophin mentioned earlier, which did not affect muscle function, a very high amount of virus carrying the gene editing components was used in the case of CRISPR mediated gene editing in the canine, which will increase the likelihood for some of it leaking into the circulation and transducing other tissues, including VSM. Given that therapy for Duchenne patients is largely focused on skeletal muscle, this article would suggest that development of therapeutics should (also) focus on VSM (and possibly also vascular endothelial cells) for any long term therapeutic benefit for patients with Duchenne and Becker muscular dystrophy. In addition the demonstration that NRG/ERBB signalling stimulates dystrophin expression would help to improve treatment regimens with Herceptin/Trastuzumab to ameliorate cardiomyopathy. Finally, this article suggests that investigation of changes in functional dystrophin levels as a possible cause of disease states associated with NRG/ERBB signalling is warranted. I trust I have addressed the concerns the reviewer raised. Bibliography 1. Escher, P. et al. Synapses form in skeletal muscles lacking neuregulin receptors. Science 308, 1920–1923 (2005). 2. Leu, M. et al. Erbb2 regulates neuromuscular synapse formation and is essential for muscle spindle development. Development 130, 2291–2301 (2003). 3. Andrechek, E. R. et al. ErbB2 is required for muscle spindle and myoblast cell survival. Mol. Cell. Biol. 22, 4714–4722 (2002). 4. Clark, J. F. The creatine kinase system in smooth muscle. Mol. Cell. Biochem. 133-134, 221–232 (1994). 5. Soriano-Arroquia, A., Clegg, P. D., Molloy, A. P. & Goljanek-Whysall, K. Preparation and Culture of Myogenic Precursor Cells/Primary Myoblasts from Skeletal Muscle of Adult and Aged Humans. J. Vis. Exp. (2017). doi:10.3791/55047 6. Ghahramani Seno, M. M. et al. RNAi-mediated knockdown of dystrophin expression in adult mice does not lead to overt muscular dystrophy pathology. Hum. Mol. Genet. 17, 2622–2632 (2008). 7. Houzelstein, D., Lyons, G. E., Chamberlain, J. & Buckingham, M. E. Localization of dystrophin gene transcripts during mouse embryogenesis. J. Cell Biol. 119, 811–821 (1992). 8. Acsadi, G. et al. Human dystrophin expression in mdx mice after intramuscular injection of DNA constructs. Nature 352, 815–818 (1991). 9. Cheng, L., Ziegelhoffer, P. R. & Yang, N. S. In vivo promoter activity and transgene expression in mammalian somatic tissues evaluated by using particle bombardment. Proc. Natl. Acad. Sci. USA 90, 4455–4459 (1993). 10. Romero, N. B. et al. Phase I study of dystrophin plasmid-based gene therapy in Duchenne/Becker muscular dystrophy. Hum. Gene Ther. 15, 1065–1076 (2004). 11. Lu, Q. L. et al. Systemic delivery of antisense oligoribonucleotide restores dystrophin expression in body-wide skeletal muscles. Proc. Natl. Acad. Sci. USA 102, 198–203 (2005). 12. Gussoni, E. et al. Dystrophin expression in the mdx mouse restored by stem cell transplantation. Nature 401, 390–394 (1999). 13. Amoasii, L. et al. Gene editing restores dystrophin expression in a canine model of Duchenne muscular dystrophy. Science (2018). doi:10.1126/science.aau1549" } ] }, { "id": "39666", "date": "29 Oct 2018", "name": "Alberto Malerba", "expertise": [ "Reviewer Expertise Muscular dystrophy", "gene therapy" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis manuscript suggests the possibility that the lack of dystrophin expression in cell types different than myofibers is the cause of Duchenne muscular dystrophy (DMD). In particular smooth muscle cells are proposed to be the cells responsible for the disease.\n\nThe idea that DMD is not necessarily due to a lack of dystrophin in skeletal muscle fibres is interesting but of course it challenges a plethora of previous studies and needs to be demonstrated properly. The main issue with this study is that the claims are not supported by results. At the end of the Introduction it is stated that “The aim of this investigation was to identify the function of NRG/ERBB signalling in muscle and, as it phosphorylated á-DB1 in the DAPC complex, determine if it had a functional role in muscular dystrophy by identifying downstream signaling targets…”. Clearly the manuscript fails to show this as only a very short dataset is provided.\n\nTo support the claim of the title the authors should provide evidences that …“Even though dystrophin is lacking in skeletal muscles of Erbb2/4 dKO mice, they do not show dystrophic symptoms”…This crucial point is mentioned in Discussion by citing the ref paper #20 which anyway does not include any information about a lack of dystrophin in muscles.\n\nThe suggestion that the lack of dystrophin in smooth muscle cells (or blood vessel cells) is causing the disease is speculative and it is not supported by the data provided. A solid set of data demonstrating that the dystrophin expressed in smooth muscle cells (or other blood vessel cells) only (ie not in myofibres) is sufficient to rescue the pathology should be provided. Clearly a proper set of experiments in vivo is needed to support the authors’ claim.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [ { "c_id": "4100", "date": "01 Nov 2018", "name": "Nadesan Gajendran", "role": "Author Response", "response": "αI thank the reviewer for reading the article and for giving comments/suggestions. The reviewer refers to previous studies that are challenged by this article. This comment was also made by reviewer 1 to which I have responded in detail below referring to published studies. Briefly, although several of the previous studies to restore dystrophin expression targeted skeletal muscle, dystrophin or mini dystrophin would also be expressed in vascular smooth muscle (VSM) due to the delivery route, promoter used for expression of dystrophin or gene editing components, or the vehicle used for delivery and hence it is neither possible to conclude that restoring dystrophin expression in skeletal muscle is sufficient to rescue the dystrophic pathology nor to exclude the role of VSM in the onset of dystrophic symptoms. Furthermore, studies that reported restoring dystrophin expression in skeletal muscle will result in the dystrophin associated protein complex (DAPC) to be restored and enable signalling from this complex to be re-established, but this does not mean that on the long term the aberrant vasoregulation due to a lack of dystrophin in VSM will not cause again the onset of dystrophy in skeletal muscle. As the reviewer is aware, the current article concludes that the root cause of DMD is the lack of dystrophin in VSM but it does not claim that restoring dystrophin expression in VSM will restore the dystrophic pathology in skeletal muscle to the normal state although it may ameliorate it. If lack of dystrophin in VSM is responsible for the onset of DMD as the current article concludes, then from what is known so far, the lack of dystrophin in skeletal muscle may exacerbate the dystrophic pathology caused by aberrant vasoregulation, possibly due to structural defects as well as the interruption of signalling events from the dystrophin associated protein complex (DAPC). This hypothesis is supported by a published study where they demonstrated that satellite cells from mdx mice retained their regenerative capacity and that the host environment was critical for satellite cell function1. Hence if the host environment is in a dystrophic state (onset of which may have been due to aberrant vasoregulation), such as in mdx skeletal muscle, impaired muscle regeneration is observed. Thus the current article suggests that treatment strategies should (also) target VSM for any long term benefit for DMD patients. The reviewer comments that the manuscript (article) fails to show if NRG/ERBB signaling had a functional role in muscular dystrophy by identifying downstream signalling targets as written at the end of the introduction. Clearly the results in Fig. 1 shows that dystrophin is a signalling target of NRG/ERBB and the fact that, based on HSA promoter activity, ERBB2/4 dKO mice do not show dystrophic symptoms (explained below) is due to ERBB2/4 and hence also dystrophin still being expressed in VSM (and possibly endothelial cells). Thus NRG/ErbB signalling is essential for dystrophin expression, and the systemic lack of dystrophin causes muscular dystrophy. The conditional HSA-CRE/ErbB2 KO mouse2 (is the same ErbB2 lox P flanked mouse used in breeding to generate the HSA-CRE/ErbB2/4 dKO mouse) shows similar pathology to the HSA-CRE/ErbB2/4 dKO mouse. However when dystrophin is lacking in skeletal muscle and VSM as in the conditional MCK-CRE/ErbB2 KO mouse3 where CRE expression is driven by the muscle creatine kinase promoter, it shows impaired muscle regeneration similar to the mdx mouse that lacks dystrophin and hence would be expected to show a more severe pathology compared to the MCK-CRE/ErbB2 KO mouse. The reviewer commented that only a very short dataset is provided. Studies supporting the conclusion of the current article that the lack of dystrophin in VSM is the root cause of DMD are: 1) Ablation of ErbB2/4 receptors results in a lack of dystrophin in myotubes; 2) Skeletal muscle specific HSA-CRE/ErbB2/4 dKO mice that would lack dystrophin in skeletal muscle but express dystrophin in VSM do not show a dystrophic pathology4; 3) HSA-CRE/ErbB2 KO2 mice used in breeding to generate HSA-CRE/ErbB2/4 dKO show a similar pathology to the latter; 4) MCK-CRE/ErbB2 mice3 that would lack dystrophin in skeletal muscle and VSM (as MCK promoter is active in both), show impaired muscle regeneration; 5) Lack of dystrophin in skeletal muscle fibers through RNAi knock down did not show an overt dystrophic pathology5; 6) expression of dystrophin in VSM, even at significantly lower levels compared to wild type controls, improved aberrant vasoregulation in mdx mice6; 7) Taken together, the current article and several published studies involving rescue of dystrophic muscle would suggest, that aberrant vasoregulation is the cause of the onset of DMD (current article) and that lack of dystrophin in skeletal muscle exacerbates the dystrophic pathology. Other data are available in the publication of the patent filing7 (patent link: WO 2017/036852 A1, link also in legend to Fig. 2) reporting that signalling (stimulation/upregulation) from NRG/ERBB is most likely mediated by the intracellular domain of ERBB4 (4ICD) whereas signalling from α-dystrobrevin 1 (α-DB1) downregulates dystrophin expression since genetic deletion of dystrobrevin results in increased dystrophin and ERBB4 expression with most of the ERBB4 being cleaved to generate 4ICD. Data in the published patent filing also show that restoring α-DB1 in DB KO myotubes downregulates dystrophin expression to levels found in myotubes formed from C2C12 cells. These data, although clearly demonstrate signalling events occur from the DAPC, are not necessary to support the conclusion in the current article and hence were not included. The reviewer comments that the ref provided (ref. number 20) in the article does not include any information about a lack of dystrophin in muscles. As I wrote in my comment to the report of reviewer 1, the supplemental data for that paper includes information that reports the absence of centralized nuclei and lack of atrophic fibers thus there was no indication to look for dystrophin expression. The paper4 also reports that sustained muscle strength was not affected nor was a myasthenic condition observed. These mice did show a hind limb extension reflex which was not reported in the paper. As I also wrote to in my comment to the report of reviewer 1, it was not possible to resolve dystrophin expression (nor phosphorylation of α-DB18) in vivo or in muscle lysates of ErbB2/4 dKO mice for the reasons explained in my comment to the report of reviewer 1 below. The reviewer also comments that “…data demonstrating that the dystrophin expressed in smooth muscle cells (or other blood vessel cells) only (i.e. not in myofibers) is sufficient to rescue the pathology should be provided. This experiment has been published whereby dystrophin was expressed in VSM that improved aberrant vasoregulation in mdx mice even though the amount of dystrophin was significantly lower6 than that in wild type control mice. Although the current article concludes that lack of dystrophin in VSM is the root cause (onset) of DMD, this does not necessarily mean that expressing dystrophin in VSM when the skeletal muscle is already is in a dystrophic state (due to systemic dystrophin deficiency) would rescue the dystrophic phenotype. Thus experiments designed to express full length dystrophin in VSM only at embryonic day 9.5 (when dystrophin is normally expressed) is not feasible given the size of the dystrophin cDNA and complicated by the fact that expression levels, even of a mini dystrophin, may not be sufficient. Hence an alternative experiment to obtain a situation in vivo where dystrophin is expressed in VSM as well as in other cells and tissues but not in skeletal muscle is to prevent NRG signalling by ErbB2/4 dKO and hence dystrophin expression in skeletal muscle i.e. the HSA-CRE/ErbB2/4 dKO mouse mentioned in the current study. The loss of dystrophin when ErbB2/4 receptors are ablated by CRE expression driven by the HSA promoter in skeletal muscle does not result in dystrophic symptoms (when dystrophin is expressed in VSM and other cells and tissues) is consistent with published observations whereby RNAi knock-down of dystrophin in skeletal muscles did not show an overt dystrophic pathology5. I would like to invite the reviewer to read my comments below to reviewer 1 as several of the concerns raised were similar. As I will submit a revised version that would address the concerns raised by reviewer 1 (report and comment) and reviewer 2, if the reviewer has any additional concerns that is communicated as a comment to this response, I would try and address those as well. Bibliography 1. Boldrin, L., Zammit, P. S. & Morgan, J. E. Satellite cells from dystrophic muscle retain regenerative capacity. Stem Cell Res. 14, 20–29 (2015). 2. Leu, M. et al. Erbb2 regulates neuromuscular synapse formation and is essential for muscle spindle development. Development 130, 2291–2301 (2003). 3. Andrechek, E. R. et al. ErbB2 is required for muscle spindle and myoblast cell survival. Mol. Cell. Biol. 22, 4714–4722 (2002). 4. Escher, P. et al. Synapses form in skeletal muscles lacking neuregulin receptors. Science 308, 1920–1923 (2005). 5. Ghahramani Seno, M. M. et al. RNAi-mediated knockdown of dystrophin expression in adult mice does not lead to overt muscular dystrophy pathology. Hum. Mol. Genet. 17, 2622–2632 (2008). 6. Ito, K. et al. Smooth muscle-specific dystrophin expression improves aberrant vasoregulation in mdx mice. Hum. Mol. Genet. 15, 2266–2275 (2006). 7. Gajendran, N. ERBB4 RECEPTOR MODULATORS FOR USE IN THE TREATMENT OF DYSTROPHIN-ASSOCIATED DISEASES. (2017). WO 2017/036852 A1 8. Schmidt, N. et al. Neuregulin/ErbB regulate neuromuscular junction development by phosphorylation of α-dystrobrevin. J. Cell Biol. 195, 1171–1184 (2011)." }, { "c_id": "4162", "date": "05 Nov 2018", "name": "Alberto Malerba", "role": "Reviewer Response", "response": "I thank the author for replying to my concerns. As I mentioned when revising the manuscript, I do consider interesting the concept that non-skeletal muscle cells have a role in DMD. This is plausible and, as the author states, there are several published studies showing that cells other than myofibres have a role in DMD but the general consensus is that the main issue is a lack of dystrophin in myofibres and that the lack of dystrophin in other cell types exacerbate the disease (which is pretty much the opposite of the author’s claim). For example the paper Ito et al 2006 concludes that “these data suggest that dystrophin in VSMCs may play an important role in the local autocrine regulation of α-adrenergic constriction, and that the loss of this regulatory mechanism may exacerbate muscle fiber necrosis”.However, I am not criticizing the claim of the current manuscript per se, my only concern is that this claim is not supported by experimental evidences (as also observed by the other reviewer). The author mentioned several publications supporting the conclusions of this manuscript but, clearly, the manuscript must sustain the claim with solid original data. Showing that in vitro “Ablation of ErbB2/4 receptors results in a lack of dystrophin in myotubes” is simply not enough to persuade the readers that the lack of dystrophin in smooth muscle cells is “The root cause of Duchenne muscular dystrophy”. The manuscript should stand on its own and observations from other published studies supporting the conclusion are valuable but should not replace original data.I would be pleased to see a revised version that addresses these concerns, but I advise the author to include on it solid in vivo data to support its claims." } ] } ]	1	https://f1000research.com/articles/7-1321
https://f1000research.com/articles/7-1133/v1	25 Jul 18	{ "type": "Method Article", "title": "Using zebrafish to study the function of nephronophthisis and related ciliopathy genes", "authors": [ "Elisa Molinari", "Simon A. Ramsbottom", "Veronica Sammut", "Frances E. P. Hughes", "John A. Sayer", "Elisa Molinari", "Simon A. Ramsbottom", "Veronica Sammut", "Frances E. P. Hughes" ], "abstract": "Zebrafish are a valuable vertebrate model in which to study development and characterize genes involved in cystic kidney disease. Zebrafish embryos and larvae are transparent, allowing non-invasive imaging during their rapid development, which takes place over the first 72 hours post fertilisation. Gene-specific knockdown of nephronophthisis-associated genes leads to ciliary phenotypes which can be assessed in various developmental structures. Here we describe in detail the methods used for imaging cilia within Kupffer’s vesicle to assess nephronophthisis and related ciliopathy phenotypes.", "keywords": [ "Kupffer’s vesicle", "acetylated alpha-tubulin", "primary cilia", "somite" ], "content": "Introduction\n\nThe zebrafish, Danio rerio, is a powerful model in which to study inherited diseases1,2. A total of 70% of human genes have a zebrafish orthologue, which can be exploited in this model to determine expression and function3. Zebrafish can be used in a huge variety of manipulations, including forward genetic screens, targeted gene editing and pharmacological screening of new compounds. Here we describe techniques to visualize cilia within the developing node called Kupffer’s vesicle (KV)4 using fluorescence microscopy. These techniques can be successfully utilized to study nephronophthisis and related ciliopathy (NPHP-RC) genes5–7.\n\nNephronophthisis (NPHP) is an autosomal recessive inherited cystic kidney disease and a common cause of childhood and adolescent end-stage renal disease8. Genetically the disease is heterogeneous, with more than 20 genes involved in pathogenesis9,10. Notably, extra-renal manifestations are seen in 15% of patients, including disorders of the brain, retina, heart, liver and skeletal system. Remarkably, all genes known to cause NPHP and its associated syndromes express their protein products in the primary cilium, basal body or centrosome, leading to the umbrella term NPHP-RC for this group of conditions11. The disease pathogenesis of NPHP-RC is intimately related to primary ciliary structure and function (in terms of both sensing and signalling). Studying these diseases and how they affect the cilia (especially during development) has led to important mechanistic insights12.\n\nKV in the zebrafish is a ciliated organ that functions in an analogous manner to the murine embryonic node13,14 and is important for the establishment of left-right (LR) asymmetry. The KV forms at around 10–12 hours post-fertilization (hpf) from a cluster of dorsal forerunner cells that have migrated to the tail bud. KV cells are ciliated and cilia mediated flow establishes LR body patterning and organ asymmetry. Typically there are around 50–60 cilia within the KV, of length 3–5 µm4,15. KV cilia are a mixture of motile and immotile. Changes in cilia number, motility and length can have a significant impact on the generation of flow and hence laterality (KV function)16–18. Despite being a well-defined structure, locating and imaging KV is challenging and we hope to demystify this in the detailed methodological review presented here.\n\nWe aim to provide a precise methodological approach for the identification and staining of KV in developing zebrafish embryos. We hope to provide details of how to exploit this ciliated structure as a readout and disease model for inherited ciliopathy syndromes.\n\n\nMethods\n\n1. Phosphate buffered saline (PBS; 10x): 25.6 g Na2HPO4·7H2O, 80 g NaCl, 2 g KCl, 2 g KH2PO4. Make up to 1 litre with H2O and autoclave at 121°C.\n\n2. 4% Paraformaldehyde (PFA) (Sigma-Aldrich ; catalogue number 16005) in PBS (note 1).\n\n3. Methanol (VWR; catalogue number 20846.326).\n\n4. Acetone (Sigma-Aldrich; catalogue number 24201).\n\n5. Blocking solution: 5% Bovine serum albumin (BSA) (Sigma-Aldrich; catalogue number A7906) in PBS containing 0.01% Tween-20 (Sigma-Aldrich; catalogue number P9416) and 0.1% DMSO (Sigma-Aldrich ; catalogue number D8779).\n\n6. Glass vials and lids (Qmx laboratories; catalogue number V0040, V0309).\n\n7. Microscope slides (VWR; catalogue number 631-1550).\n\n8. PVC insulation tape (Onecall; catalogue number CBBR7213) (note 7).\n\n9. Primary antibodies for ciliary and KV staining e.g. anti-acetylated α-tubulin (Sigma-Aldrich; catalogue number T6793); anti-aPKC (1:500, SCBT) Santa Cruz Biotechnology; sc-216).\n\n10. Secondary antibodies e.g. Alexa Fluor® 488 Donkey Anti-Mouse IgG Antibody (Thermo Fisher Scientific; catalogue number A-21202); Alexa Fluor® 568 Donkey Anti-Rabbit IgG Antibody (Thermo Fisher Scientific; catalogue number A10042).\n\n11. Mounting medium (Vectashield H-1200, Vector Laboratories).\n\nImaging KV allows for easy analysis of cilia structure and function. It can be initially seen from the 5 somite stage (ss), but is most easily visualized between the 8 and 12 ss as it increases in size over this developmental window (Figure 2 and note 2).\n\nTo assess the age of the fish embryos, roll the embryo on to its side and count along the number of fully-formed somites, starting from the head to the tail (Figure 1). At the 10 ss the tail will begin to bud off from the yolk and this can be used as a secondary measure (Figure 2).\n\n1. Fix 8-12 ss Zebrafish embryos in 4% paraformaldehyde in phosphate buffered saline (4% PFA in PBS) overnight at 4°C. Do not remove the chorions before fixation (note 3).\n\n2. Dehydrate the embryos in Methanol in a graded fashion from 25% to 100% methanol (in distilled water) in steps of 25%, changing the wash every 10 minutes.\n\n3. Store the embryos in 100% Methanol at -20°C until required.\n\nA schematic (A) and light-microscopy image (B) showing a lateral view of a 10 somite stage zebrafish embryo. The most anterior somite (red arrow) is slightly shorter and broader than more posterior somites. The generation of somites is tightly linked to the overall development of the embryo and therefore allows for accurate aging. e, eye; kv, Kupffer’s vesicle; ov, otic vesicle; s, somites; tb, tail bud.\n\nLateral (top panels) and ventral (bottom panels) views of zebrafish from the 8 somite stage (ss) to the 12 ss. Over this developmental window, Kupffer’s vesicle (white arrowhead) expands and then shrinks, reaching its greatest size at the 10 ss. At 28.5°C, zebrafish will form approximately two somites per hour, so the above panels represent a 4-hour period.\n\nThroughout this procedure you should use glass vials and pipettes (note 4).\n\n1. Rehydrate embryos slowly through a graded series of Methanol from 100% to 0% (in distilled water), changing the wash every 5 minutes. Keep embryos on ice throughout (note 5).\n\n2. Wash twice more with distilled water to remove all traces of methanol\n\n3. Incubate in chilled acetone at -20°C for 5 minutes, to permeabilise the tissue\n\n4. Replace acetone with distilled water\n\n5. Wash twice more with distilled water to remove all traces of acetone\n\n6. Replace water with PBS (to reduce osmotic stress in de-chorionated tissues)\n\n7. Manually de-chorionate the embryos using watchmaker’s forceps or Green 21 gauge needles\n\n8. Transfer embryos into 2ml glass vials\n\n9. Incubate embryos in blocking solution for 2 hours\n\n10. Incubate embryos in primary antibody in blocking solution overnight at 4°C (note 6)\n\n11. Wash embryos 3 times in PBS for 15 minutes or longer per wash\n\n12. Incubate embryos in secondary antibody in blocking solution overnight at 4°C\n\n13. Wash embryos 3 times in PBS for 15 minutes or longer per wash\n\nAt the end of the staining procedure, embryos can be stored for up to two weeks at 4°C in PBS, or long-term at 4°C in 2% PFA in PBS. Staining of embryos does not appear to be significantly diminished by longer-term storage.\n\nOnce zebrafish embryos have been stained, the easiest way to visualize KV is by mounting the tail only in relief slides. Many tails can be imaged on a single slide, making slide preparation, imaging and storage easier.\n\nPrepare slides\n\n1. Take a glass slide and lay one layer of PVC tape on to it (note 7).\n\n2. Cut a rectangular hole in the middle of the tape and remove this portion to create a well.\n\nPrepare and mount embryos\n\n1. Remove the tail tip of the embryos using forceps (note 8 and Figure 3)\n\n2. Place up to 10 tail tips into the prepared well\n\n3. Orient the tail tips so that they face upwards\n\n4. Carefully remove the surrounding PBS (note 9)\n\n5. Fill the well with mounting medium (note 10)\n\n6. Place a cover slip over the well and seal the edges with nail varnish\n\nSchematic showing a 10 somite stage zebrafish embryo (A). Forceps can be used as depicted to separate the tail tip from the rest of the embryo. Excess yolk is then removed, and the tail tip mounted flat on to a microscope slide. (B) Bright field image overlaid with fluorescence image of flat-mounted tail tip. The white box depicts the location of Kupffer’s vesicle. Fluorescence from acetylated-α tubulin (red) can be seen within this region. (C) Confocal image of Kupffer’s vesicle, visualised using acetylated-α tubulin (red), aPKC (green) and DAPI (blue).\n\nImaging can be undertaken with any fluorescent microscope that is able to remove out-of-focus light. While the KV can be seen with a traditional global excitation and capture system, there is too much noise resulting from the surrounding tissue. A confocal microscope offers the best image quality, but optical sectioning using structured illumination also gives good results.\n\nLight microscopy images were acquired using a Leica MZ16F stereo microscope and Leica Application Suite V4.2. The presence of fluorescently labelled acetylated-α tubulin was confirmed in flat-mounted embryos using a Leica MZ16F stereo microscope with a Leica external light source for fluorescence excitation EL6000 with a red fluorescence protein filter.\n\nFluorescent microscopy images were captured using an Axio Imager Z1 fluorescent microscope (Zeiss), using a Colibri light source and filter sets for DAPI (blue), Alexa Fluor488 (green) and Alexa Fluor 594 (Red) with a 20X lens.\n\nZebrafish were maintained in a Home Office aquarium facility at a temperature of 28°C. All zebrafish procedures were performed under Home Office UK license regulations.\n\n1. Preparation of paraformaldehyde\n\nParaformaldehyde is dangerous in its powdered form and should be handled with care using a fume hood and mask. To successfully get PFA into solution it is necessary to warm it to 60°C and add sodium hydroxide to raise the pH. At the correct temperature and pH (pH 7), the powder will go into solution rapidly. After cooling check the pH is correct. You can freeze aliquots (e.g. 10–20 ml) of PFA for later use. Once thawed, keep at 4°C for up to three days.\n\n2. Fixing zebrafish at the 10 ss\n\nAt 28.5°C, embryos reach the 10 ss approximately 14 hpf, making collection difficult (as this is typically very late at night, assuming fish lay eggs in the morning in a standard aquarium with unmodified night/day cycles). To facilitate collection of the 10 ss embryos, it is possible to slow the development of the embryos down by placing them at a slightly lower temperature. At around 24°C the embryos will not reach the 10 ss until approximately 21 hpf (i.e. the following morning).\n\n3. Embryo fixation\n\nIt is usual to remove the chorion before fixation of fish embryos. At early stages, however, the embryos are very fragile and also the yolk tends to stick to almost anything when rehydrated. Thus it is best to leave the chorion on when undertaking the rehydration procedure, prior to immunostaining.\n\n4. Immunostaining in glass\n\nEmbryos should be transferred using glass pipettes; they will stick to plastic and disintegrate easily. Immunostaining should be carried out in glass vials only. Using 2-ml glass vials allows for staining in a small volume of liquid (<200 µl) which helps to reduce reagent costs and reduces the storage space required.\n\n5. Embryo rehydration\n\nEmbryos are very fragile at this stage and should be handled with care. When rehydrating, only remove 50% of the solution they are in and replace with the next one (e.g. remove half of the 100% methanol, and replace with half of 75% methanol to give 87.5%, then remove most of this and place into 75% methanol such that the steps are of roughly 12.5%). Mixing methanol with water leads to an exothermic reaction so embryos should be kept on ice throughout to maintain their low temperature.\n\n6. Primary antibodies\n\nAny antibody which specifically detects ciliary proteins can be used for analysis of KV, provided that it cross-reacts with zebrafish antigen. Staining of the intra-flagellar transport (IFT) machinery, the axoneme, or the ciliary membrane are all possible with this technique and allow for analysis of both structural and functional defects within zebrafish cilia. For simple detection of cilia structure the best antibody is anti-acetylated-α tubulin (Sigma-Aldrich; catalogue number T6793) at 1:500 for the ciliary axoneme. To visualise KV epithelium, counterstain with anti-aPKC at 1:500.\n\n7. Generation of imaging chamber\n\nWhile it is possible to purchase slides with wells of varying sizes, most commercially available slides have a circular bottom making it difficult to get a uniform distance from the cover slip to the sample. As a result only one or two samples can be placed in each. To generate a uniformly flat imaging surface, PVC tape can be used to convert any flat slide to a relief slide in which you can create an imaging chamber to fit your own size requirements. For KV visualization, one layer of tape provides ample depth to prevent the sample being crushed, while keeping the sample close to the cover slip. Detailed information and a visual demonstration of imaging chamber generation has been reported previously19.\n\n8. Embryo dissection for KV visualization\n\nTo get a good image of KV it is best to remove the tail tip from the rest of the embryo and lay it flat. Using two pairs of forceps, pinch two-thirds of the way along the back of the embryo, to separate the posterior part of the embryo from the anterior, and between the head and tail to detach the tail from the yolk sac. Remove as much yolk from the remaining tissue as possible.\n\n9. Removal of PBS from the relief slide\n\nDue to the small size of the dissected zebrafish tail tip, it can be difficult to remove the PBS from the imaging chamber without disturbing the samples. To facilitate this process, use a razor blade to cut off a medium-sized (20–200 µl) plastic micropipette tip at 45°. The cut tip can be placed flat on to the microscope slide allowing for aspiration of PBS without disturbing the samples.\n\n10. Mounting medium\n\nMounting medium with an anti-fading agent is essential for the successful imaging of immunofluorescence-stained KV. Some mounting media is available in a hard-setting format e.g. Vectashield H-1500, but it is preferable to use a non-setting medium such as Vectashield H-1200.\n\n\nDiscussion\n\nThe use of zebrafish for modelling genetic diseases has proven to be very significant. Zebrafish are advantageous given the presence of transparent developmental stages. They are straightforward to maintain and genetic manipulations can be performed with relative ease. Here we have focused on accessing and imaging the zebrafish KV, a transient developmental structure with a lumen and ciliated epithelial cells lining it. The KV functions, via its cilia, serves as a ‘left-right’ organiser in order to determine left-right asymmetry within the zebrafish embryo13,14. Within the KV, a single layer of epithelial cells surround a fluid-filled lumen into which they extend a motile cilium, generating asymmetric fluid flows that allow left-right patterning signals to be transmitted. Thus the KV is a simple yet accessible structure that is ideal for investigating cilia structure and function, and their role in inherited human ciliopathy syndromes. Nearly all new ciliopathy syndromes described in man use zebrafish, and their KV in particular, to gain insights into disease pathogenesis. Genetic knockdowns can cause defects in KV that may affect the size and structure of the organ as well as the cilia, and RNA rescue can determine the specificity of this affect. The study of KV therefore provides an accessible tool for the modelling of ciliopathies and interventions, such as genetic therapies and pharmacological agents, which may be used to rescue underlying cilia defects.\n\n\nData availability\n\nDataset 1. Raw, uncropped images from which Figures 1-3 were generated. Leftmost image, Figure 1; middle six images, Figure 2; rightmost two images, Figure 3. DOI: 10.5256/f1000research.15511.d21072720.", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by Northern Counties Kidney Research Fund, Kidney Research UK (PDF_003_20151124) and the Medical Research Council (MR/M012212/1).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nDrummond IA: Kidney development and disease in the zebrafish. J Am Soc Nephrol. 2005; 16(2): 299–304. PubMed Abstract \| Publisher Full Text\n\nSong Z, Zhang X, Jia S, et al.: Zebrafish as a Model for Human Ciliopathies. J Genet Genomics. 2016; 43(3): 107–20. PubMed Abstract \| Publisher Full Text\n\nHowe K, Clark MD, Torroja CF, et al.: The zebrafish reference genome sequence and its relationship to the human genome. Nature. 2013; 496(7446): 498–503. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAmack JD: Salient features of the ciliated organ of asymmetry. Bioarchitecture. 2014; 4(1): 6–15. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSimms RJ, Hynes AM, Eley L, et al.: Modelling a ciliopathy: Ahi1 knockdown in model systems reveals an essential role in brain, retinal, and renal development. Cell Mol Life Sci. 2012; 69(6): 993–1009. PubMed Abstract \| Publisher Full Text\n\nChaki M, Airik R, Ghosh AK, et al.: Exome capture reveals ZNF423 and CEP164 mutations, linking renal ciliopathies to DNA damage response signaling. Cell. 2012; 150(3): 533–48. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSlaats GG, Saldivar JC, Bacal J, et al.: DNA replication stress underlies renal phenotypes in CEP290-associated Joubert syndrome. J Clin Invest. 2015; 125(9): 3657–66. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSimms RJ, Eley L, Sayer JA: Nephronophthisis. Eur J Hum Genet. 2009; 17(4): 406–16. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWolf MT, Hildebrandt F: Nephronophthisis. Pediatr Nephrol. 2011; 26(2): 181–94. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSrivastava S, Molinari E, Raman S, et al.: Many Genes-One Disease? Genetics of Nephronophthisis (NPHP) and NPHP-Associated Disorders. Front Pediatr. 2018; 5: 287. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHildebrandt F, Otto E: Cilia and centrosomes: a unifying pathogenic concept for cystic kidney disease? Nat Rev Genet. 2005; 6(12): 928–40. PubMed Abstract \| Publisher Full Text\n\nSlaats GG, Giles RH: Are renal ciliopathies (replication) stressed out? Trends Cell Biol. 2015; 25(6): 317–9. PubMed Abstract \| Publisher Full Text\n\nEssner JJ, Vogan KJ, Wagner MK, et al.: Conserved function for embryonic nodal cilia. Nature. 2002; 418(6893): 37–8. PubMed Abstract \| Publisher Full Text\n\nKramer-Zucker AG, Olale F, Haycraft CJ, et al.: Cilia-driven fluid flow in the zebrafish pronephros, brain and Kupffer's vesicle is required for normal organogenesis. Development. 2005; 132(8): 1907–21. PubMed Abstract \| Publisher Full Text\n\nWang G, Yost HJ, Amack JD: Analysis of gene function and visualization of cilia-generated fluid flow in Kupffer's vesicle. J Vis Exp. 2013; (73): e50038. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMontenegro-Johnson TD, Baker DI, Smith DJ, et al.: Three-dimensional flow in Kupffer's Vesicle. J Math Biol. 2016; 73(3): 705–25. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSmith DJ, Montenegro-Johnson TD, Lopes SS: Organized chaos in Kupffer's vesicle: how a heterogeneous structure achieves consistent left-right patterning. Bioarchitecture. 2014; 4(3): 119–25. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLin CY, Tsai MY, Liu YH, et al.: Klf8 regulates left-right asymmetric patterning through modulation of Kupffer's vesicle morphogenesis and spaw expression. J Biomed Sci. 2017; 24(1): 45. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFellgett SW, Ramsbottom SA, Maguire RJ, et al.: Using Confocal Analysis of Xenopus laevis to Investigate Modulators of Wnt and Shh Morphogen Gradients. J Vis Exp. 2015; (106): e53162. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMolinari E, Ramsbottom S, Veronica Sammut V, et al.: Dataset 1 in: Using zebrafish to study the function of nephronophthisis and related ciliopathy genes. F1000Research. 2018. Data Source" }	[ { "id": "36493", "date": "13 Aug 2018", "name": "Weibin Zhou", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nMolinari et al. presented a detailed description of the method they have developed to assess the function of ciliopathy-related genes using the zebrafish KV as a model system. In particular, they developed this method to image fluorescently labeled KV cilia such that (I assume) the morphology of those cilia can be analyzed. Overall the description is well written but a little more detail is needed to allow the readers to reproduce and use this method in their own studies. My major questions are as follows:\n\nThe KV cilia are either motile and immotile. I think the method in this manuscript is primarily addressing the morphology of cilia, but the authors did not present any method of analyzing the ciliary morphology after the illustrated imaging process. It would be critical for them to present their quantitative analysis of ciliary morphology based on the fluorescent images, to ensure the completeness of this manuscript. It is not mentioned how many embryos were placed into each glass vial for immunostaining and the volume of staining solution for each sample. This is also pertinent to the quantitative analyses, as I wonder about the typical sample size needed for the quantitative analyses. I think the intro and the discussion needs to be edited by a native English speaker to ensure there is no grammatical error.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Partly\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Partly", "responses": [] }, { "id": "37381", "date": "30 Aug 2018", "name": "Glenn P. Lobo", "expertise": [ "Reviewer Expertise Ciliogenesis", "vitamin A receptors", "zebrafish", "mouse models", "microscopy", "retinal degeneration", "photoreceptors", "kidney" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nMolinari and colleagues present a detailed methodology to functionally access cilia structure in KV vessels of zebrafish.\nI am especially impressed with the details provided, methodology setup and the clarity of the images in this manuscript.\nSuggestions:\nIt would be beneficial to readers if authors provide an image of the \"relief slides/ imaging chamber\" used for mounting.\n\nWhat type of microscope was used? Upright or inverted? An image of the actual zebrafish in the imaging chamber/ slides on the mounting stage of the microscope would also provide additional information about the set-up. What type of objective was used? Air or oil objective? Additional information on microscopy could be provided.\n\nA more in-depth detail on optical sectioning using structured illumination (I presume using an Zeiss Apotome 2) would be beneficial.\n\nWhich program was used to calculate cilia length? Fiji or Image J? Please provide some additional details.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Partly\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes", "responses": [] }, { "id": "37382", "date": "05 Sep 2018", "name": "Roslyn Simms", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nMolinari et al. present a well written and clear description of the methodology required to image the transient developmental zebrafish organelle, Kupffer's Vesicle (KV). Overall, they clearly show KV development using light microscopy and a confocal image of cilia in KV stained using acetylated-alpha tubulin.\n\nTheir method will definitely facilitate research to study the structure of KV and its cilia. They provide useful tips on how to overcome practical challenges. However it would be informative to include information on how to measure cilia length (for example software used) and how to calculate KV volume.\nSimilarly, additional methodological information will have to be added relating to studying the \"function of KV\", perhaps using live confocal imaging to study cilia motility, or this comment amended.\nMinor points, which would help the reader:\nI suggest delete \"easy\" in the Methods section on \"Fixing for KV imaging\" as this relatively contradicts the challenge of imaging KV which you describe in the introduction. Consider including a magnified image of the white arrowed area of the lateral views in Figure 2. Add information on how you identified the location of KV from the bright field image in Figure 3B. In the notes section, add the number of KV you advise mounting on each slide for imaging. Finally in the discussion, add some references to support your statements:\"Nearly all new cilioopathy syndromes,,,\" and \"genetic therapies and pharmacological agents\" (the time window to do this will be relatively short).\n\nOverall an informative methodological article for the zebrafish research community.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Partly", "responses": [] } ]	1	https://f1000research.com/articles/7-1133
https://f1000research.com/articles/7-1888/v1	03 Dec 18	{ "type": "Research Article", "title": "INCISION e-learning program as a useful teaching tool to enhance surgeons’ knowledge and skills: An Indonesian multi-center cross-sectional pilot study", "authors": [ "Achmad Kemal Harzif", "M Nurhadi Rahman", "Pungky Mulawardhana", "Nadia Shafira", "Tricia Dewi Anggraeni", "Kanadi Sumapraja", "Dwiana Ocviyanti", "M Nurhadi Rahman", "Pungky Mulawardhana", "Nadia Shafira", "Tricia Dewi Anggraeni", "Kanadi Sumapraja", "Dwiana Ocviyanti" ], "abstract": "Background: Media aids are one of the most important components in the teaching and learning process. This pilot study program was conducted in order to assess the effectiveness of the INCISION e-learning program as teaching media in the surgical teaching and learning process, and its ability to improve surgical skills and knowledge achievement. Methods: One intervention group and one control group were involved in this study. The intervention group used the hysterectomy INCISION e-learning module, while the control group used conventional teaching approaches. The study was conducted with 14 resident surgeons in three universities in Indonesia: Universitas Indonesia, Universitas Airlangga, and Universitas Gajah Mada. The testing components used were a pre-test, post-test questionnaire (a modified Ritzman questionnaire) and direct observation of procedural skills in the operating room (OR). Data were analyzed descriptively using Mann-Whitney and Wilcoxon tests. Results: Using a Mann-Whitney test, we found the differences between the average scores of the intervention group and the control group to be statistically significant (p=0.046). A Wilcoxon test also revealed significant differences (p=0.028). The modified Ritzman questionnaire also revealed that the residents in the intervention group felt more confident in their surgical knowledge (82%), and made more efficient use of their time in the OR (81%). Conclusions: These findings reveal a significant improvement in knowledge and skill achievement in residents that underwent training via the INCISION e-learning module, compared to residents taught via conventional teaching strategies.", "keywords": [ "INCISION", "e-learning", "surgeon", "skills", "knowledge" ], "content": "Introduction\n\nThe use of multimedia as a learning tool is one of the best educational techniques as it is able to engage more than one sense simultaneously, generally the senses of sight and hearing. Multimedia programs provide a variety of different stimuli, including elements of text, speech, sound and music, graphics, animations and still pictures1–3. In undergraduate and graduate medical training, the type of teaching media used depends on the institution as well as the individual teacher and the subject matter being taught4,5.\n\nTraditionally, medical teachers explain theories and demonstrate procedures, followed by practice by the trainees (“see one, do one”)3,6. There are four teaching approaches to surgical education including: standardized/simulated patients; procedure courses, videos, textbooks; web-based training; cadavers and live animals7. Nowadays, technological development has influenced the ways in which learning and information presentation takes place, with a variety of technological tools now supplementing and partly replacing paper books2,3,6,8–11. The INCISION e-learning module is a new learning and teaching approach, comprised of an online learning platform designed to transfer procedure-specific knowledge to surgeons, gynaecologists and residents. INCISION also provides some information on pre- and postoperative care; however the primary focus is on procedure and the relevant surgical anatomy.\n\nIn this pilot study, we sought to evaluate the effectiveness of the INCISION approach on the transfer of relevant surgical knowledge, as well as to assess the strengths and weaknesses of e-learning via INCISION from the point of view of the surgery resident.\n\n\nMethodology\n\nThe INCISION pilot study program involved 14 surgeon residents specializing in obstetrics and gynecology (OBGYN). Inclusion criteria included residents in 3–6th semester who had never been trained for hysterectomies, while exclusion criteria included participants who were not willing to participate in this study or had incomplete filling of the questionnaire.\n\nWe divided residents into an intervention group and a control group by random number generation, consisting of 7 surgeon residents in each group. Recruitment of the participants was done by asking residents in 3–6th semester in person on July until August 2015 during the break after class at three universities: Universitas Indonesia, Universitas Airlangga, and Universitas Gajah Mada.\n\nThe intervention group used the hysterectomy INCISION e-learning module, while the control group used conventional teaching approaches. The distribution of participants is shown in Table 1.\n\nThis study was approved by the Ethics Committee of Faculty of Medicine, Univesitas Indonesia, on July 6th 2015 (reference number: 564/UN2.F1/ETIK/2015). Permission was obtained to perform this study in all three sites. Written consent for participation was obtained from all study participants.\n\nThis study was conducted between July and August 2015, in the medical facilities and hospitals of: Universitas Indonesia, Universitas Airlangga and Universitas Gajah Mada.\n\nFour evaluation methods were used. First, we included a pre-test with the purpose of evaluating the residents’ knowledge of procedures prior to training (Supplementary File 1). Second, a post-test was conducted after having received the course material (Supplementary File 1). Third, a Ritzman questionnaire was administered to evaluate the residents’ perception of the INCISION approach on the training material, and how it impacted their knowledge (Supplementary File 2). The Ritzman questionnaire was not taken by the control group, only the intervention group. Every question is rated on a score of 1 to 7, which when converted into a percent is able to be grouped as >80% = good, 50-79% = average, and <50% = bad.\n\nFinally, residents performed the procedure under the supervision of a qualified trainer, which was not assessed.\n\nIn-depth interviews were also conducted with the intervention group for assessment value of INCISION e-learning. Residents were interviewed by the research team in the Obstetrics and Gynecology Department's Meeting Room. The research team wrote notes during the interview only. The residents were asked about added value of using INCISION e-learning; weakness, difficulties or obstacles encountered during the study with INCISION e-learning; the difference between INCISION e-learning and conventinal teaching; and whether the residents will recomend INCISION e-learning for their friends.\n\nResidents participating in the study were given a pre-test during the first week of the study period. This test served as an entry exam to gauge their incoming level of expertise with the particular procedure. Starting in the second week, the intervention group of residents commenced learning via the class with INCISION e-learning 2D module, while the control group of residents were taught via presentation methods.\n\nOn the third week, trainers and residents from the intervention group took the class with INCISION e-learning 3D module and discussed it together, while the control group watched the trainer and followed the operation of a patient. Finally, all residents conducted the post-test. The residents took the exam at least once during the week after the study period.\n\nSupplementary File 3 contains information about the conventional course.\n\nData were analyzed descriptively using SPSS 21 version statistical software, using Mann-Whitney and Wilcoxon tests for statistical analysis. We used a non-parametric test due to the small sample size. The Mann-Whitney test was used to examine the differences in knowledge based on the pre-test examination. The Wilcoxon signed-rank test was also used to evaluate the difference between courses, using data from a paired-sample design. We used the Wilcoxon test to determine whether there was a difference between the intervention group and the control group at the end of the pilot program, based on the administered post-test. A p value <0.05 was considered statistically significant.\n\n\nResults\n\nAll subjects completed the study; no participants declined to be part of the study.\n\nFigure 1a shows the score distribution of the pre- and post-tests in the control group. Five residents decreased their score and two showed an increase. However, in the intervention group, the majority of residents increased their score, as shown in Figure 1b.\n\nDistribution of pre- and post-test scores from (a) the control group and (b) the intervention group.\n\nWe evaluated differences between the two groups’ average pre-test scores in order to determine whether there were any differences in knowledge and skill prior to intervention. Using a Mann-Whitney test, we demonstrated that the average value of the pre-test in the control group was not significantly different from that of the intervention group (p=0.561;).\n\nFurther evaluation of the average post-test scores between the control and intervention groups was important to determine whether there were differences in knowledge and skill following intervention. Using a Mann-Whitney test, we found that the average score of the intervention group (67), was significantly greater than that of the control group (53). This difference was statistically significant (p=0.046).\n\nWe then compared the pre- and post-test results within each group. The average pre-test score in the control group was 52.71, while the average post-test score in the control group was 52.71. There was no difference between the pre- and post-test results in the control group. The average pre-test score in the intervention group was 50.14, while the average post-test score was 66.93. A Wilcoxon test revealed that this difference was significant (p=0.028).\n\nResidents in the intervention group were satisfied with the training outcome, with an average Ritzman questionnaire score of 81% (Table 2). The highest score, 86%, was given for the statement that residents believe the content of INCISION is useful for their job. On average, residents felt that they appreciated the course (82%), that the learning atmosphere was encouraging (78%), that the learning was fun (73%), and that they obtained beneficial knowledge from the course (80%). The residents also felt that the INCISION e-learning was comprehensible (82%). The content and the language (foreign words and technical terms) was also found to be easy to comprehend (86 and 83%, respectively). They felt that they kept-up thematically with the course (82%), and that the time spent was sufficient for the theme covered (80%).\n\nAdditionally, residents in the intervention group felt that INCISION e-learning provides an adequate gain in knowledge (77%). They had the impression that their knowledge had expanded on a long-term basis (82%), that they would be able to remember the new themes well (78%), and that they think they will be able to sufficiently report what they had learned some time after the course (61%). They also expressed that they will apply what they learned to their day-to-day work (76%), and that they would recommend the INCISION approach to their collegues (80%). Overall, the residents rated the INCISION media as good (84%). They preferred 3D film (90%) over 2D film (76%), for ease of understanding the content. With respect to content presentation, they also found 3D film (94%) to be more suitable than 2D film (80%). They felt that the online academy both aided understanding (88%), and was an acceptable medium for presenting the content (80%).\n\nThe residents also felt more confident in their surgical knowledge (82%), and made more efficient use of their time in the OR (81%) after following the INCISION approach. They thought that the understanding of the procedure was aided by the step-by-step approach (88%), and feel patient safety would be increased due to the INCISION approach (82%). Regarding product feedback, the residents thought that more images should be provided (80%), as well as more (78%) and longer (84%) video segments.\n\nIn the in depth-interview, most residents said that INCISION e-learning via its 3D videos was able to improve knowledge and skills. They said that the advantage of using INCISION e-learning was that it can be accesed anywhere and at anytime, but there were limitations to access the program, for example, when they be located in remote areas that don’t have internet access the program couldn’t be used. Finally, all residents in the intervention group said that they will recommend INCISION e-learning due to their positive learning experience to their colleagues.\n\n\nDiscussion\n\nThe results of this study correspond to our expected outcome regarding the INCISION e-learning paradigm. The initial knowledge and skills of residents in both the control and intervention groups were similar, whereas following INCISION e-learning training of residents in the intervention group, the knowledge outcome was significantly increased. We can therefore conclude that INCISION e-learning is able to increase residents’ knowledge compared to conventional learning.\n\nThe limitation of this study design was the small number of participants; however, this meant that we were able to complete this report with in-depth interviews of the residents in the intervention group, as shown in Table 3. They discussed the value of INCISION e-learning, including weaknesses, difficulties, and obstacles encountered during the study, as well as the difference between learning with or without the INCISION module. They also discussed whether they would recommend the program to others.\n\nMost residents reported that INCISION e-learning was able to improve knowledge and skills via its use of 3D videos. Another reported advantage was that it can be accessed anywhere and at anytime. Reported weaknesses of the program was limitations to access should they be located in remote areas. Residents also reported that they will recommend INCISION e-learning to their colleagues due to their positive learning experience.\n\nThese results are consistent with previous studies that report the efficacy, and the satisfaction amongst users, of multimedia as learning tool for medical purposes, in particular for surgical learning8,9,11.\n\n\nConclusion\n\nThese findings reveal that there were significant differences in knowledge and skill achievement between students who underwent training via the INCISION e-learning module and students who were trained via conventional teaching strategies. In addition, a questionnaire revealed that resident surgeons in the intervention group appreciated the use of INCISON e-learning.\n\n\nData availability\n\nF1000Research: Dataset 1. Answers for Ritzman questionnaire about the perception of INCISION e-learning in the intervention group., https://doi.org/10.5256/f1000research.15799.d22725512\n\nF1000Research: Dataset 2. Raw data for pre and post test scores in both control and intervention groups., https://doi.org/10.5256/f1000research.15799.d22725613", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nThe authors would like to thank all participants and supervisors that contributed to the work in this study.\n\n\nSupplementary material\n\nSupplementary File 1: Pre- and post-test used.\n\nClick here to access the data\n\nSupplementary File 2: Ritzman questionnaire about the perception of INCISION e-learning.\n\nClick here to access the data\n\nSupplementary File 3: Conventional Hysterectomy Course Schedule.\n\nClick here to access the data\n\n\nReferences\n\nAloraini S: The impact of using multimedia on students’ academic achievement in the college of education at king saud university. Journal of King Saud University-Languages and Translation. 2012; 24(2): 75–82. Publisher Full Text\n\nIssa N, Mayer RE, Schuller M, et al.: Teaching for understanding in medical classrooms using multimedia design principles. Med Educ. 2013; 47(4): 388–396. PubMed Abstract \| Publisher Full Text\n\nShariff U, Seretis C, Lee D, et al.: The role of multimedia in surgical skills training and assessment. Surgeon. 2016; 14(3): 150–163. PubMed Abstract \| Publisher Full Text\n\nMasic I: E-learning as new method of medical education. Acta Inform Med. 2008; 16(2): 102–17. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIssa N, Mayer RE, Schuller M, et al.: Research on the applications of multimedia auxiliary teaching system with the computer to the education of clinical medicine. Medical Education. 2015; 47(4).\n\nShariff U, Kullar N, Haray PN, et al.: Multimedia educational tools for cognitive surgical skill acquisition in open and laparoscopic colorectal surgery: a randomized controlled trial. Colorectal Dis. 2015; 17(5): 441–450. PubMed Abstract \| Publisher Full Text\n\nParmar DJ, Shah C, Parmar RD: Effectiveness of four step approach for procedures (skill) teaching: Rct. The Southeast Asian Journal of Case Report and Review. 2013; 2(5): 281–288. Reference Source\n\nChoules AP: The use of elearning in medical education: a review of the current situation. Postgrad Med J. 2007; 83(978): 212–216. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFunke K, Bonrath E, Mardin WA, et al.: Blended learning in surgery using the Inmedea Simulator. Langenbecks Arch Surg. 2013; 398(2): 335–340. PubMed Abstract \| Publisher Full Text\n\nFingeret AL, Martinez RH, Hsieh C, et al.: Watch what happens: using a web-based multimedia platform to enhance intraoperative learning and development of clinical reasoning. Am J Surg. 2016; 211(2): 384–389. PubMed Abstract \| Publisher Full Text\n\nNorman E, Furnes B: The relationship between metacognitive experiences and learning: Is there a difference between digital and non-digital study media? Comput Hum Behav. 2016; 54: 301–309. Publisher Full Text\n\nHarzif AK, Rahman MN, Mulawardhana P, et al.: Dataset 1 in: INCISION e-learning program as a useful teaching tool to enhance surgeons’ knowledge and skills: An Indonesian multi-center cross-sectional pilot study. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15799.d227255\n\nHarzif AK, Rahman MN, Mulawardhana P, et al.: Dataset 2 in: INCISION e-learning program as a useful teaching tool to enhance surgeons’ knowledge and skills: An Indonesian multi-center cross-sectional pilot study. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15799.d227256" }	[ { "id": "41413", "date": "21 Dec 2018", "name": "Arunita Jagzape", "expertise": [ "Reviewer Expertise Medical Education Technology" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe efforts are laudable and adequate weightage can be given since it is a multi-center study but I would like to mention that in the title it is mentioned \"to enhance knowledge and skills\", out of which only knowledge is tested. It is mentioned only in the Abstract regarding Direct observation of Procedural Skills, but it was nowhere mentioned in the main text, neither was it evaluated which it should have been, as it could test the skill component as well as performance. The module could also be explained in detail. In the Methodology, study design and sampling technique is not mentioned. Kindly explain the part in the Methodology - Procedure; that the residents took the exam at least once during the week after the study period. In the Discussion, the present study is not discussed in light of other studies. In the conclusions, it is mentioned that there were significant differences in knowledge and skill. If skills were nor assessed, this statement can't be made.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/7-1888
https://f1000research.com/articles/7-1464/v1	14 Sep 18	{ "type": "Research Article", "title": "Establishing an international laboratory network for neglected tropical diseases: Understanding existing capacity in five WHO regions", "authors": [ "Laura Dean", "Janet Njelesani", "Charles Mulamba", "Russell Dacombe", "Pamela S. Mbabazi", "Imelda Bates", "Janet Njelesani", "Charles Mulamba", "Russell Dacombe", "Pamela S. Mbabazi", "Imelda Bates" ], "abstract": "Background. Limited laboratory capacity is a significant bottleneck in meeting global targets for the control and elimination of neglected tropical diseases (NTD). Laboratories are essential for providing clinical data and monitoring data about the status and changes in NTD prevalence, and for detecting early drug resistance. Currently NTD laboratory networks are informal and specialist laboratory expertise is not well publicised, making it difficult to share global expertise and provide training, supervision, and quality assurance for NTD diagnosis and research. This study aimed to identify laboratories within five World Health Organisation regions (South-East Asia, Eastern Mediterranean, Americas, Western Pacific and Europe) that provide NTD services and could be regarded as national or regional reference laboratories, and to conduct a survey to document their networks and capacity to support NTD programmes. Methods. Potential NTD reference laboratories were identified through systematic searches, snowball sampling and key informants. Results. Thirty-two laboratories responded to the survey. The laboratories covered 25 different NTDs and their main regional and national roles were to provide technical support and training, research, test validation and standard setting. Two thirds of the laboratories were based in academic institutions and almost half had less than 11 staff. Although greater than 90 per cent of the laboratories had adequate technical skills to function as an NTD reference laboratory, almost all laboratories lacked systems for external verification that their results met international standards. Conclusions. This study highlights that although many laboratories believed they could act as a reference laboratory, only a few had all the characteristics required to fulfil this role as they fell short in the standard and quality assurance of laboratory processes. Networks of high quality laboratories are essential for the control and elimination of disease and this study presents a critical first step in the development of such networks for NTDs.", "keywords": [ "Neglected Tropical Diseases", "Capacity Building", "Laboratory Networks", "Quality Assurance", "Americas", "Eastern Mediterranean", "Europe", "South-East Asia", "Western Pacific" ], "content": "Introduction\n\nLaboratories are recognised as one of the weakest elements of health systems due to chronic under-investment. Lack of investment results in poor infrastructure, inadequate numbers and skills of technical staff, insufficient and uncoordinated technical assistance, and lack of diagnostic tools appropriate for low-resource settings1,2. Yet laboratory services are integral to interventions for the surveillance, control and elimination of neglected tropical diseases (NTDs). Laboratories provide clinical and monitoring data about disease prevalence and trends, and are essential for flagging up early signs of drug resistance1,3–6. Appropriate management of clinical cases of NTDs depends on laboratories providing accurate diagnoses for identifying cases5,6. Preventive chemotherapy interventions through mass drug administration (MDA) rely on laboratory data to make decisions regarding intervention effectiveness and for reliably documenting progress towards zero transmission5.\n\nAccelerated scale-up of existing interventions is critical to reach the 2020 NTD Roadmap targets on the control and elimination of NTDs, however, lacking laboratory capacity is a critical bottleneck preventing the international NTD community from meeting targets. There needs to be enhanced laboratory ability in areas with significant NTD prevalence to provide technical and scientific support for the diagnosis, surveillance, monitoring and evaluation of national NTD programmes7. The World Health Organisation’s (WHO) Strategic and Technical Advisory Group (STAG) for Neglected Tropical Diseases have therefore prioritised strengthening the capacity of NTD laboratories and establishing a formal NTD laboratory network which can provide a quality assurance and referral function8.\n\nGlobally, few laboratories specialise in NTDs. Laboratory support for NTD programmes is generally provided by parasitology laboratories within national health care systems or research institutions9. Most of these laboratories focus on malaria, and to a lesser extent on soil transmitted helminths (STH), with very little laboratory expertise in other NTDs10. No central register of specialist NTD laboratories exists. NTD laboratory expertise is fragmented and un-coordinated, with no formal referral system or network to provide high level support from internationally-accredited reference laboratories for quality assurance of NTD testing. Consequentially, much laboratory data available on NTDs, that has been used to make important strategic decisions about programme implementation and transmission rates, may have been generated by laboratories working in isolation that are not enrolled in any external quality assurance scheme. The global laboratory infrastructure for NTD control programmes lags behind many other global health programmes, such as those for tuberculosis, malaria, poliomyelitis, measles and hepatitis, which successfully established a globally connected network of laboratories and systems for externally validating disease-specific laboratory data as recommended by the World Health Assembly11,12.\n\nTo identify and harness existing capacity and to improve efficiency, laboratories that support national NTD programmes need to be mapped and organised into a functional international network. At the top tier, there should be internationally accredited and interlinked national reference laboratories. Each of which should head a pyramidal referral structure comprising laboratories at, for example, provincial level who support more peripheral district and primary care sites involved in front-line diagnosis and surveillance. The role of the NTD reference laboratories is to maintain their own accreditation and service quality, and to facilitate provision of quality services by lower level laboratories through, for example, offering training on good laboratory practice and quality management systems, external quality assessment and referral testing, and monitoring performance standards through the organisation of regular proficiency testing1.\n\nInformation about the location and expertise of laboratories with specialist NTD expertise across WHO regions is scarce and difficult to access. It is not generally known whether these laboratories meet international accreditation standards or have the capacity, expertise, and networks that would enable them to operate as national or regional reference centres. Creating a database of laboratories that includes a description of what support they can provide for NTD programmes is an essential first step in the process of establishing an international and regional NTD laboratory network.\n\nThis study aimed to identify laboratories that provide NTD services and could be regarded as national or regional reference laboratories within WHO regions, and to document their capacity to support NTD programmes. It covered five of WHO’s six regions since the WHO Africa region office conducted its own complementary study and the results could not be collated due to differing study methods. Our study mapped the geographical distribution and networks of these laboratories, and collated information about the skills and services they provided to support NTD programmes. Scoping the current situation provides a platform on which to design strategies to build an international network of accredited NTD reference laboratories. Such a network is essential to overcome the laboratory bottleneck which is a key barrier in accelerating intervention scale-up to meet 2020 NTD Roadmap goals7.\n\n\nMethods\n\nThere is no existing global register of specialist NTD laboratories. This scoping study developed an unbiased and comprehensive way of identifying potential NTD reference laboratories in the five WHO regions – Americas, Eastern Mediterranean, Europe, South-East Asia and Western Pacific. As there was no pre-existing definition of an ‘NTD reference laboratory’, we extracted information from published literature13–17 about the laboratory characteristics needed to fulfil diagnostic, research, supervision, training, quality, and networking requirements of a national or regional reference laboratory for NTDs, and verified them with NTD control programme specialists. These characteristics were:\n\n• able to conduct verifiable quality diagnosis in one or more NTDs\n\n• able to support research into NTDs prevalent in their region\n\n• able to train and mentor staff in national or tertiary level laboratories within the region\n\n• actively networked with other national and international NTD laboratories and research institutions\n\n• evidence of accreditation to international standards (e.g., ISO 15189, Good Clinical Laboratory Practice (GCLP))\n\nDetailed information from the literature about each of these characteristics was used to design an electronic survey administered through Bristol On-line Survey (now Online surveys) to laboratories in the five WHO regions with potential to be national or regional NTD reference laboratories. Topics covered were: location and geographical coverage, NTD tests available, accreditation status, staffing, ability to provide training and technical support, and any capacity gaps the laboratory perceived they had in relation to NTDs.\n\nTo identify as many potential regional reference laboratories to include in the survey, and to avoid bias, two wide-ranging search strategies were used. Firstly, key informants were identified from international NTD programmes and research institutions and through WHO regional offices. These included WHO officers in each of the five regions, representatives of multi-lateral agencies supporting laboratory networks and centres, and NTD funders and researchers. Snowballing was used to identify further key informants.\n\nEach key informant was asked to identify which laboratories they were aware of that could be considered an NTD reference laboratory based on our pre-defined list of characteristics. Laboratories did not have to focus exclusively on NTDs, since NTDs may be part of a larger portfolio of work but needed to have a reputation as a referral laboratory (or laboratory unit) for NTDs. 25 key informants provided contact details for 69 laboratories that they considered may be perceived as an NTD reference laboratory.\n\nSecondly, an internet search for potential NTD reference laboratories was conducted. Countries affected by NTDs in each of the five WHO regions were identified from information on NTD strategies and/or activities in documents on WHO regional websites18–23, from country-specific information in the WHO NTD roadmap, and from individuals in the WHO Global Working Group on Capacity Strengthening for national NTD programmes. Overall 60 countries were identified as being affected by and prioritising NTDs in the five WHO regions: The Americas 17 countries, South-East Asia 11 countries, Europe 8 countries, Eastern Mediterranean 14 countries, and Western Pacific 10 countries. Potential NTD reference laboratories were identified by searching websites of national NTD programmes in NTD-affected countries, and the websites of the WHO regional offices, and by following additional links and references provided on these websites. The internet search strategy identified 98 laboratories that may potentially be NTD reference laboratories.\n\nFor each identified laboratory, contact details of laboratory heads were obtained from the websites or through key informants. Overall the combined searches yielded 167 contacts in potential reference laboratories. Each contact person was provided with information about the purpose and content of the survey by e-mail and asked to complete the survey. In order to increase response rates, the Modified Dillman approach24 was used which involved fortnightly reminders about the survey for a period of five weeks between October 2013-January 2014 until the survey closed. The survey was also offered in Spanish as appropriate. Following closure of the surveys, 35 telephone calls were made to collect information from non-respondents. These focussed particularly on the European and Americas regions due to low response rates and generated one additional completed survey.\n\nData from the survey was entered into an excel spreadsheet and anonymised. Data was analysed to provide quantitative, descriptive information, and content analysis was conducted to identify NTD laboratories that met or were close to meeting the characteristics of a reference laboratory.\n\n\nResults\n\nNineteen percent (n=32) of the 167 of the laboratory heads contacted responded to the survey. The majority of respondents were from the Eastern Mediterranean region (34%, n=11) (Table 1). No responses were received from the European region. The majority of laboratories (53%, n=17) provided a national level service and 25% (n=8) operated at the international level, predominantly within their own WHO region. The main regional role of surveyed laboratories was the provision of technical support (22%, n=7) and training (22%, n=7) to other laboratories though some were also involved in research, test validation and standard setting (Table 1).\n\nThe 32 laboratories covered 25 different NTDs across five WHO regions with 23 laboratories (72%) covering two or more NTDs. Most laboratories within the Eastern Mediterranean region (73%, n=8) specialised in leishmaniasis. Within the Western Pacific region, laboratories tended to focus on STH (50%, n=4) and schistosomiasis (50%, n=4). Laboratories in South-East Asia focused on STH (60%, n=6) and lymphatic filariasis (60%, n=6). Chagas disease, taeniasis, cysticercosis, echinococcosis, onchocerciasis and dengue were the only NTDs covered in the Americas region.\n\nLaboratories tended to be small with almost half (47 %, n=15) employing 1–10 staff and three-quarters (76%) employing 30 or less. The South-East Asia region had the highest number of staff per laboratory. The Eastern Mediterranean region had the least staff and commonly lacked quality officers, management and administrative staff. Ninety one percent (n=29) of laboratories indicated that their staff had the necessary technical skills to function as an NTD reference laboratory and only one laboratory strongly believed it lacked this technical capacity. Two thirds (69%, n=22) of the laboratories were based in academic institutions and felt they were strong in supporting NTD research. The majority of laboratories (78 %, n=25) identified at least one gap in their capacity, most commonly external quality assurance, which was reported as lacking by 9 (28%) laboratories. The type of capacity gaps was similar across all five regions.\n\nThere was variation in adherence to laboratory quality standards. Only four (13%) stated adherence to international standards such as Good Laboratory Practice, ISO 15189 and ISO 9000. 47 percent (15) of laboratories adhered to national quality standards and 40% (13) did not adhere to any. Although 47% (19) of laboratories reported a quality officer, only five (16%) participated in an external quality assurance (EQA) programme for NTD tests. Four of these were in the Eastern Mediterranean region. All five stated that less than 5% of their results within the last 3 years had been unsatisfactory. In the Western Pacific, South-East Asia and Americas regions, at least 90% (12) of laboratories did not participate in an external quality assurance scheme. Fourteen (44%) laboratories had regular interactions with international NTD networks including Drugs for Neglected Diseases Initiative, Tropical Disease Research and regional NTD elimination programmes. International conferences, regional meetings, and NTD workshops, were the predominant modes of networking reported. Despite the majority (91%) of laboratories believing they have the capacity to carry out the role of a reference laboratory, only 14% (n=3) met all the pre-determined characteristics.\n\n\nConclusions\n\nThis study marks an important step in a process towards creating the need and awareness for an international network of NTD expert laboratories. We used a systematic and wide-ranging approach to identify 32 laboratories distributed across four of five WHO regions which have potential to be regional or national reference laboratories for NTDs. Between them, these laboratories reported that they have the technical skills to provide expertise in 25 different NTDs with each laboratory focussing on NTDs that are prevalent in their region. These laboratories could form the top tier of an interlinked network of laboratories capable of providing quality information for NTD programmes and able to act as NTD research and training centres. Half the laboratories operated at national level and a quarter at regional level and this two-level geographical focus forms a sound basis for creating national and international NTD laboratory networks.\n\nAlthough over 90% of the laboratories surveyed in this study believed they could act as a reference laboratory, only 3 had all the characteristics required to fulfil this role. Almost all laboratories fell short in the area of standards and quality assurance of laboratory processes. A consistent finding was that 87% of laboratories did not adhere to international quality standards and 40% did not adhere to national quality standards. Further evidence of the paucity of quality systems is that only five of the 32 laboratories participated in an EQA programme for NTD tests; four of these were in the Eastern Mediterranean region. This means there is no independent verification that laboratory results meet international standards. Such verification is imperative in order to generate reliable results and lack of verification undermines confidence in data concerning NTD prevalence, trends and reduced drug efficacy25. Unless NTD data originates from quality assured, accredited laboratories, reports about progress towards global NTD targets will lack credibility.\n\nIt is not clear why laboratories felt that they were able to act as reference laboratories even though many did not adhere to national or international quality standards and most were not enrolled in external quality schemes. This finding suggests that the importance of being able to demonstrate that test results are reliable may be under-recognised even among laboratory professionals and that this validation is not demanded by NTD programme managers and other decision makers.\n\nOverall, we contacted 167 laboratories and received information from 32, a third of which were located within the Eastern Mediterranean region; none were in the European region. We used a broad search strategy, so a high response rate was not anticipated since it was likely that many laboratories contacted were not involved in reference-level NTD work. However, it is possible that our search missed some relevant laboratories or that we did not identify some laboratories because contact details were incorrect, or language barriers prevented some managers from responding.\n\nThe majority of laboratories covered at least two different NTDs. This diversity raises an important question about whether each laboratory should focus on one NTD or several. This will in part be dictated by existing expertise and by the burden of different NTDs in the vicinity. The most efficient use of resources may be to centralise expertise for several NTDs within one laboratory. This would facilitate throughput of large numbers of samples and centralise the expensive, state-of-the-art diagnostic tools needed to provide high levels of diagnostic specificity and sensitivity26–28. Amalgamation of laboratory services for several NTDs is complex and would need to be carefully managed to maintain rigorous systems and quality standards across a large range of services and to get buy-in from national programme managers and other laboratories25.\n\nMost of the laboratories specialising in NTDs are based in research institutions. This is an important factor to consider when planning an international NTD laboratory network since the primary goal of these laboratories is to generate research. This goal may not always be aligned with the priorities of national NTD programmes to provide routine service delivery and training. Research laboratories are characterised by short-term projects with high staff turnover and are strongly influenced by the topical interests of donors. The difference in priorities faced by research laboratories and NTD programmes means that potential tensions need to be anticipated and managed if a research laboratory is the primary provider of national or regional NTD laboratory expertise.\n\nAlthough only three of the identified laboratories had all the characteristics of a reference laboratory, within each of the four WHO regions surveyed there are at least two laboratories that could be strengthened to reach international accreditation status and fulfil a role as a regional reference centre for NTDs. Responses from these laboratories indicated they have some experience in implementing international laboratory standards, in using advanced diagnostic tools and in providing technical support and training to other laboratories. At least some of the surveyed laboratories have the potential to be in the top global tier as regional reference laboratories once they are able to adhere to international quality standards. If they are integrated into a formal laboratory network, they will be able to, for example, share equipment, develop standardised indicators and regular monitoring protocols for laboratory performance and staff competence29, and provide training, supervision and mentoring for lower tier laboratories.\n\nThis study provides preliminary information about the location and expertise of high-level laboratories specialising in NTDs. The data we collected was self-reported by laboratory personnel, so an important next step will be for additional information from European and African NTD laboratories to be incorporated, and for selected laboratories to undergo a more in-depth and independent assessment of their capacity. Criteria then need to be agreed and used to strategically select a small number of laboratories in each region which will be supported technically and financially to achieve international accreditation and formal recognition as regional NTD reference laboratories. These laboratories can then form foci around which to construct a global network of NTD laboratories with the longer term aim of encompassing national and sub-national laboratories within the network. A wide-ranging stakeholder consultation process, likely conducted under the stewardship of the NTD STAG which reports directly to the Director General of WHO, will be needed to define the criteria for selecting laboratories and to define the goal and operation of the NTD laboratory network30. Only when this laboratory network is operational will it be possible to have effective and rapid regional and global referral and quality assurance systems and to have confidence in the NTD test results that are essential to support NTD programme operations and research needs 12,27.\n\n\nData availability\n\nDataset 1: De-identified survey data 10.5256/f1000research.16196.d21750131\n\n\nConsent\n\nPotential participants were informed about the purpose and content of the survey and could choose whether or not to complete the survey.", "appendix": "Grant information\n\nThe study was commissioned by the World Health Organization WHO Global Working Group on Capacity Strengthening for National Neglected Tropical Diseases (NTD) Programmes [APW200811893] which reports to the NTD Strategic Technical Advisory Group.\n\n\nSupplementary material\n\nSupplementary File 1: World Health Organization (WHO) regions survey. Includes questions and possible responses\n\nClick here to access the data.\n\n\nReferences\n\nNkengasong JN, Nsubuga P, Nwanyanwu O, et al.: Laboratory systems and services are critical in global health: time to end the neglect? Am J Clin Pathol. 2010; 134(3): 368–373. PubMed Abstract \| Publisher Full Text\n\nPEPFAR Conceptual Framework: Integrated Laboratory Strengthening for High Burden Diseases (Tuberculosis, HIV, and Neglected Tropical Diseases). Geneva: Undated. (accessed 2016 Jan 20). Reference Source\n\nRegional Strategic Plan for Neglected Tropical Diseases in the African Region 2014-2020. Brazzaville: World Health Organisation. 2013; (accessed 2016 Jan 20). Reference Source\n\nSchistosomiasis Progress Report 2001–2011 and Strategic Plan 2012–2020. Geneva: World Health Organisation. 2013; (accessed 2016 Jan 20). Reference Source\n\nAssuring Safety of Preventive Chemotherapy Interventions for the Control of Neglected Tropical Diseases. Geneva: World Health Organisation. 2011; (accessed 2016 Jan 20). Reference Source\n\nUtzinger J, Becker SL, Knopp S, et al.: Neglected tropical diseases: diagnosis, clinical management, treatment and control. Swiss Med Wkly. 2012; 142: w13727. PubMed Abstract \| Publisher Full Text\n\nAccelerating work to overcome the Global Impact of Neglected Tropical Diseases: A road map for implementation. Executive summary. Geneva: World Health Organisation. 2012; (accessed 2016 Jan 20). Reference Source\n\nSalle D: Report of the Sixth meeting of the WHO Strategic and Technical Advisory Group for Neglected Tropical Disease. Geneva 29–30 April. 2013. Geneva: World Health Organisation; 2013; (accessed 2016 Jan 20). Reference Source\n\nRichardson H, Fleming C, Palmer J, et al.: An assessment of the utilization of diagnostic parasitology laboratory services in Ontario. Can J Infect Dis. 1996; 7(4): 237–42. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChinery WA: Diagnostic laboratory parasitology--a stepping stone to medical research in tropical Africa. J Hyg Epiddemiol Microbiol Immunol. 1992; 36(4): 356–61. PubMed Abstract\n\nProtocol for Assessing National Surveillance and Response Capacities for the International Health Regulations (2005). Geneva: World Health Organisation. 2010; (accessed 2016 Jan 20). Reference Source\n\nRoush S, Baldy L: Laboratory support for the surveillance of vaccine-preventable diseases. In: Roush S, Baldy L, ed. VPD Surveillance Manual. Atlanta: Centers for Disease Control and Prevention. 2008; (accessed 2016 Jan 20). Reference Source\n\nDatema TA, Oskam L, Engelberts MF, et al.: Global laboratory initiative tool for a stepwise process towards tuberculosis laboratory accreditation. Int J Tuberc Lung Dis. 2012; 16(5): 704–705. PubMed Abstract \| Publisher Full Text\n\nMartin R, Barnhart S: Global laboratory systems development: needs and approaches. Infect Dis Clin North Am. 2011; 25(3): 677––91, x. PubMed Abstract \| Publisher Full Text\n\nYao K, McKinney B, Murphy A, et al.: Improving quality management systems of laboratories in developing countries: an innovative training approach to accelerate laboratory accreditation. Am J Clin Pathol. 2010; 134(3): 401–409. PubMed Abstract \| Publisher Full Text\n\nMaruta T, Motebang D, Mathabo L, et al.: Impact of mentorship on WHO-AFRO Strengthening Laboratory Quality Improvement Process Towards Accreditation (SLIPTA). Afr J Lab Med. 2012; 1(1): 6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLaboratory Quality Management System Training Toolkit. Geneva: World Health Organisation. 2016; (accessed 2016 Jan 20). Reference Source\n\nAccelerating work to overcome the global impact of neglected tropical diseases. A roadmap for implementation. Geneva: World Health Organisation. 2011; (accessed 2016 Jan 20). Reference Source\n\nRegional office for the Eastern Mediterranean: Neglected Tropical Diseases. Country Activities. Eastern Mediterranean: World Health Organisation. 2015; (accessed 2016 Jan 20). Reference Source\n\nAddressing diseases of poverty: An initiative to reduce the unacceptable burden of neglected tropical diseases in the Asia Pacific region. Western Pacific Region: World Health Organisation and Asian Development Bank. 2014; (accessed 2016 Jan 20). Reference Source\n\nRegional Strategic Plan for Integrated Neglected Tropical Disease Control in South-East Asia Region 2012-2016. India: World Health Organisation SEARO. 2012; (accessed 2016 Jan 20). Reference Source\n\nRegional Action Plan for Neglected Tropical Diseases in the Western Pacific Region (2012-2016). Philippines: World Health Organisation WPRO; (accessed 2016 Jan 20). 2013. Reference Source\n\nSalle B: Report of the WHO Strategic and Technical Advisory Group for Neglected Tropical Diseases. WHO Headquarters 8-9 April 2014. Geneva: World Health Organisation. 2014; (accessed 2016 Jan 20). Reference Source\n\nDillman DA: Mail and internet surveys: The Tailored Design Method. New York: Wiley, 2000. Reference Source\n\nHotez P, Raff S, Fenwick A, et al.: Recent progress in integrated neglected tropical disease control. Trends Parasitol. 2007; 23(11): 511–514. PubMed Abstract \| Publisher Full Text\n\nChappuis F, Stivanello E, Adams K, et al.: Card agglutination test for trypanosomiasis (CATT) end-dilution titer and cerebrospinal fluid cell count as predictors of human African Trypanosomiasis (Trypanosoma brucei gambiense) among serologically suspected individuals in southern Sudan. Am J Trop Med Hyg. 2004; 71(3): 313–7. PubMed Abstract\n\nKeiser J, Duthaler U, Utzinger J: Update on the diagnosis and treatment of food-borne trematode infections. Curr Opin Infect Dis. 2010; 23(5): 513–520. PubMed Abstract \| Publisher Full Text\n\nvan Lieshout L, Polderman AM, Deelder AM: Immunodiagnosis of schistosomiasis by determination of the circulating antigens CAA and CCA, in particular in individuals with recent or light infections. Acta Trop. 2000; 77(1): 69–80. PubMed Abstract \| Publisher Full Text\n\nPeter TF, Shimada Y, Freeman RR, et al.: The need for standardization in laboratory networks. Am J Clin Pathol. 2009; 131(6): 867–74. PubMed Abstract \| Publisher Full Text\n\nKirk CJ, Shult PA: Developing laboratory networks: a practical guide and application. Public Health Rep. 2010; 125 Suppl 2: 102–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDean L, Njelesani J, Mulamba C, et al.: Dataset 1 in: Establishing an international laboratory network for neglected tropical diseases: Understanding existing capacity in five WHO regions. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16196.d217501" }	[ { "id": "38347", "date": "10 Oct 2018", "name": "Charles S. Mgone", "expertise": [ "Reviewer Expertise Diseases of poverty", "capacity building" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript highlights the lack of international, regional and national capacity for reference laboratories and networks for neglected tropical disease diagnosis, research and training. This includes lack of hierarchical referral system that allows escalation of problem solving based on the nature of their complexity and offers supervision and quality assurance along the chain. The paper highlights these problems very well. The data collection was based on laboratory personnel self-reporting after the laboratories were identified by informants and web searching; both inherent with bias. The identification methods used are likely to select well known or well advertised centres while self-reporting is likely to be associated with higher claims of excellence that is real. The lack of information from African and European regions is very glaring and needs to be rectified. Informed willingness of these centres to work as regional reference laboratories needs to be assed, especially after the centres having been made aware of the terms of reference.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4066", "date": "17 Oct 2018", "name": "Susie Crossman", "role": "Reader Comment", "response": "We are grateful to Prof Mgone for his thoughtful review of our paper. We agree that the way that we selected the laboratories and the self-reporting of laboratory capacity could have biased the results towards more well-known laboratories and over-estimation of capacity. We have therefore expanded the limitations section of the paper to include a statement to this effect. We agree that there is a glaring lack of information from the African and European regions and in the discussion section we had already emphasized the need to rectify this." } ] } ]	1	https://f1000research.com/articles/7-1464
https://f1000research.com/articles/7-1879/v1	02 Dec 18	{ "type": "Research Article", "title": "Strong association between higher-risk sex and HIV prevalence at the regional level: an ecological study of 27 sub-Saharan African countries", "authors": [ "Chris R. Kenyon", "Jozefien Buyze", "Ilan S. Schwartz", "Jozefien Buyze", "Ilan S. Schwartz" ], "abstract": "Background: It is unclear why HIV prevalence varies by nearly two orders of magnitude between regions within countries in sub-Saharan Africa. In this ecological study, we assess if HIV prevalence by region is associated with any of four markers of higher risk sexual behavior: lifetime number of partners, multiple partners in past year, higher risk sex (defined as sex with non-cohabiting, non-marital partners) and age at debut. Methods: We performed Pearson’s correlation between the 4 behavioral risk factors and HIV prevalence by region in 47 nationally representative surveys from 27 sub-Saharan African countries, separately by gender. In addition, principal components analysis was used to reduce the eight risk factors (four for each gender) to two principal components (PCs). Mixed effects linear regression was used to assess the relationship between the resulting two PCs and HIV prevalence after controlling for the prevalence of male circumcision. Results: HIV prevalence varied by a median 3.7 fold (IQR 2.9-7.9) between regions within countries. HIV prevalence was strongly associated with higher risk sex and, to a lesser extent, the other risk factors evaluated. Both PCs were strongly associated with HIV prevalence when assessed via linear regression. Conclusions: Differences in sexual behavior may underpin the large differences in HIV-prevalence between subpopulation within sub-Saharan African countries.", "keywords": [ "HIV prevalence", "ecological", "sexual network", "high risk sex", "sub-Saharan Africa" ], "content": "Abbreviations\n\nDHS - Demographic Health Survey\n\nPC - Principal components\n\nPCA - Principal components analysis\n\nSTI - Sexually transmitted infection\n\n\nBackground\n\nWhereas individual-level parameters may influence which individuals in a given population acquire infection, it is population-level parameters that affect the prevalence of infection.\n\nAral et al. (Aral et al., 2007)\n\nAral and others have argued persuasively that population-level parameters such as the structure of sexual networks determine the prevalence of sexually transmitted infections (STIs) (Aral et al., 2007; Morris et al., 2010; Morris et al., 2009). If this is true, then it follows that ecological studies are necessary to assess which markers of network structure are associated with higher HIV prevalence (Kenyon & Colebunders, 2012; Morris et al., 2010).\n\nWhilst a few ecological studies have found a difference in sexual behavior could explain differences in HIV prevalence (De Walque, 2006; Kenyon et al., 2013; Kenyon & Colebunders, 2012; Morris et al., 2010; Morris et al., 2009), many have not (Auvert et al., 2001b; Cleland & Ferry, 1995; Drain et al., 2004; Sawers & Stillwaggon, 2010; Wellings et al., 2006), and the methodologies have been questioned. For instance, most involved cross-country comparisons; these are suboptimal because national populations may be constituted by several relatively separate sexual networks (Kenyon & Colebunders, 2013; Laumann & Youm, 1999). It would therefore be more appropriate to conduct these ecological assessments at the level of coherent sexual networks rather than national populations (Kenyon et al., 2015b). If there is both a high degree of sexual partner homophily by ethnic group, and large differences in HIV prevalence between ethnic groups within countries, then it would be appropriate to use ecological analyses to assess if differences in sexual network structure could explain the variations in HIV prevalence (Kenyon et al., 2013). This type of study from the USA, UK, Kenya, Uganda and South Africa has revealed positive ecological associations between various markers of sexual behavior (number of partners in past year and lifetime, partner concurrency and age of sexual debut), by ethnic group and HIV prevalence (Kenyon et al., 2014a; Kenyon et al., 2013; Kenyon et al., 2014b; Kenyon et al., 2014c; Kenyon et al., 2016b; Morris et al., 2009).\n\nResults from our recent ecological analysis of factors associated with HIV prevalence in Ethiopia based on data from Demographic Health Surveys (DHS) implied that it may be more appropriate to evaluate differences by region than by ethnic group (Kenyon et al., 2015b). For instance, 80 different ethnic groups have been grouped into 9 ethnically-based regions in Ethiopia’s DHS surveys (Central Statistical Agency [Ethiopia] & ICF International, 2012). Surveys from other sub-Saharan African countries also combine similar ethnic groups into regions and, as in Ethiopia, are designed to provide representative samples from each of these regions. We established that there was a high degree of sexual partner homophily by ethnic group region in Ethiopia (Kenyon et al., 2015b). There was also a high (seven-fold) difference in HIV prevalence between regions and this correlated closely at an ecological level with a number of behavioral variables such as lifetime number of partners, reporting sex with a non-marital, non-cohabiting partner, and age at first sex (Kenyon et al., 2015b). In this paper, we extend these earlier analyses to assess systematically if differences in HIV prevalence by region within countries in sub-Saharan Africa are associated with markers of sexual risk behavior.\n\n\nMethods\n\nNationally representative HIV-serolinked DHS’s are available for 29 sub-Saharan African countries from MeasureDHS (http://www.measuredhs.com). We limited our analyses to the 47 surveys from 27 countries where the HIV prevalence varied between regions by at least two-fold; we dropped 3/3 surveys from Lesotho, 1/1 from Swaziland and 2/3 from Zimbabwe (2005 and 2010). The selected variables were downloaded in regional format for all HIV-serolinked surveys from included countries via the STATcompiler function of MeasureDHS.\n\nBehavioral risk factors:\n\nHigher risk sex (0/47): Percentage of women/men who have had sex with a non-marital, non-cohabiting partner in the last 12 months of all women/men reporting sexual activity in the last 12 months\n\nMultiple partners in past-year (0/47): Percentage of women/men aged 15–49 years who have had sexual intercourse with more than one partner in the last 12 months.\n\nLifetime partners (11/47): Mean number of lifetime sexual partners among women/men who ever had sexual intercourse.\n\nWomen/mens’ debut (5/47 for women and 30/47 for men): Median age at first sexual intercourse in years among women/men aged 20–49 years.\n\nControl variables:\n\nMen circumcised (14/47): Percentage of men who report being circumcised.\n\nOutcome variable:\n\nHIV prevalence (0/47): Percentage of 15–49 year old men and women who tested positive for HIV.\n\nThese variables were selected for analysis based on associations with HIV from previous studies and coverage in a sufficient number of surveys (Kenyon et al., 2014a; Kenyon et al., 2013; Kenyon et al., 2015b; Mishra et al., 2009; Morris et al., 2009).\n\nOur analysis used Pearson’s correlation to assess the association between lifetime partners, multiple partners in the past year, high-risk sex, sexual debut and HIV prevalence (15–49 years, men and women combined). The analyses were conducted by region within each survey and done separately for women and men. A Pearson’s r ≥ 0.3 or ≤ -0.3 was considered positive or negative, respectively. For each risk factor, we also compared the fold-difference in the prevalence of the risk factor between the regions with the highest and lowest HIV prevalence in each survey.\n\nMultivariable analysis. Pearson’s correlation was used to assess the associations between the four behavioral risk factors separately between women and men. Principal components analysis (PCA) was used to reduce seven of the eight risk factors to two principal components. Sexual debut of men was not included in this process as data was missing for 30/47 surveys.\n\nMixed effects linear regression with a random intercept for individual surveys was then used to assess the relationship between HIV prevalence and these two principal components, controlling for prevalence of circumcision. The following sensitivity analyses were conducted. 1. Analyses stratified by gender. 2. Using principal components calculated excluding the lifetime partners of women and men variables (as these were missing in 11 surveys).\n\nThe study was conceived as being exploratory rather than hypothesis testing. Because of this and the fact that the sample sizes in each survey were relatively small, we did not use Bonferroni corrections (Armstrong, 2014; Perneger, 1998; Rothman, 1990). A P-value below 0.05 was regarded as statistically significant. The analyses were conducted using STATA 13.0 (College Station, TX). Each DHS received ethical committee clearance for data analyses such as the one performed here. Consequently, no specific ethics committee approval was necessary for this study.\n\n\nResults\n\nA total of 47 HIV-serolinked surveys from 27 countries were included. This sample included countries with a wide range of national HIV prevalences in 15- 49 year olds, ranging from 0.4% (95% confidence interval [CI] 0.2-0.5%) in Niger 2012 to 13.8 (95% CI, 12.9–14.8%) in Zimbabwe 2015. The HIV prevalence differed a median 3.8-fold (interquartile range [IQR] 2.9-8.5) between regions within surveys (Table 1).\n\naHIV ratio: HIV prevalence in highest HIV prevalence region/ HIV prevalence in lowest HIV prevalence region in each survey.\n\nP<0.05 represented by bold font.\n\nThe prevalence of higher-risk sex between regions within surveys varied widely in women (median 4.8-fold difference [IQR 2.3-8.1]) and men (median 2.5-fold difference [IQR 1.8-5.5]).\n\nWomen: There was a positive association between higher risk sex and HIV in 39/47 surveys, of which 26 were statistically significant. There were no surveys where the association was negative (Table 1; Figure 1). Within surveys, the region with the highest HIV prevalence had a median 2.5 (IQR 1.4 to 6.5) times increased prevalence of higher-risk sex for women than the lowest HIV prevalence population.\n\nMen: The association was positive in 38/47 surveys, 22 of which were significant, and negative in none of the surveys. In each survey, the region with the highest HIV prevalence had a median 1.9 (IQR 1.2 to 2.9) times higher prevalence of higher risk sex for men than the lowest HIV prevalence population.\n\nThe reported number of lifetime partners between regions in each survey varied by a median 1.8-fold (IQR 1.5-2.4) for women and 2.5-fold (2.0-3.3) for men.\n\nWomen: There was a positive association between number of lifetime partners and HIV prevalence in 18/36 surveys, of which 9 were statistically significant (Table 1, Figure 2). Of the three surveys for which the association was negative, this was statistically significant for one (Namibia, 2013). Within surveys, the region with the highest HIV prevalence had a median 1.3 (IQR 1.0 to 1.5) times higher prevalence of lifetime partners for women than the lowest HIV prevalence population.\n\nMen: The association was positive in 22/36 surveys (significant in 13 surveys). In the three surveys where the association was negative, this was only significant in the case of Namibia 2013. Within each survey, the region with the highest HIV prevalence had a median 1.5 (IQR 1.1 to 2.1) times higher prevalence of lifetime partners for men than the lowest HIV prevalence population.\n\nThe percent with multiple partners in the past year varied between regions by a median 6.5-fold (IQR 3.5-13.7) for women and 2.8-fold (IQR 2.0-4.8) for men.\n\nWomen: There was evidence of a positive association in 30/47 surveys but this association was only significant in five surveys (Table 1, Figure 3). The association was negative in three surveys, one of which was significant (Namibia, 2013). In each survey, the region with the highest HIV prevalence had a median 2.1 (IQR 1.4 to 4.1) times higher prevalence of multiple partners for women than the lowest HIV prevalence population.\n\nMen: The association was positive in 21/47 surveys, of which four were statistically significant and negative in 10 surveys, of which one was significant – Chad 2014. Within surveys, the region with the highest HIV prevalence had a median 1.3 (IQR 1.0 to 2.1) times higher prevalence of multiple partner for men than the lowest HIV prevalence population.\n\nWomen: There was evidence of a positive association between women’s age at sexual debut and HIV prevalence in 22/42 surveys, but this association was only significant in three surveys (Chad, 2014; Guinea, 2012; Niger, 2012; Table 1, Figure 4). The association was negative in six surveys, of which this was significant in two (Kenya, 2008; Senegal, 2010).\n\nMen: The association was positive in 1/17 surveys, which was statistically significant (Gabon, 2012) and negative in three surveys, none of which was significant.\n\nWithin both women and men, the three behavioral risk factors (lifetime partners, past year partners and high-risk sex) were positively associated with one another (Table 2). Within men (but not women) these three risk factors were negatively associated with age at first sex.\n\nPrincipal components: Two principal components were able to explain 80% of the variation in the seven variables. The component loadings of the principal components are depicted in Figure 5. Principal component one (PC1) represented a summation of the variables high-risk sex (women and men), lifetime partners (women and men) and multiple partners (women and men), minus the variable debut age women (Figure 5).\n\nIn the unadjusted model, PC1 and PC2 were associated with HIV prevalence (coeff. 0.51, 95% CI 0.31-0.71, and coeff. 0.72, 95% CI 0.38-1.05, both P<0.005, respectively; Table 3). Adjusting for the prevalence of circumcision made little difference to the strength of the association (Table 3).\n\nP<0.05, *P<0.005. PC, principal component.\n\nSensitivity analyses repeating the analyses stratified by gender, using the principal components that were calculated excluding the lifetime partners (women and men) variables in the analyses made little difference to the results (Table 3).\n\n\nDiscussion\n\nWe found wide variations in the prevalence of HIV and the four behavioral risk factors evaluated. All the risk factors, excluding women’s/men’s debut, were positively associated with one another and with HIV prevalence. Combining the four risk factors via principal components analysis generated two components which were strongly associated with HIV prevalence.\n\nThere are a number of interpretations of and limitations to our results. Demographic and Health Surveys are not designed to accurately assess sensitive information such as sexual behavior (Beauclair et al., 2013; Morris et al., 2014). As such, they have been shown to underestimate the prevalence of behaviors that may be particularly sensitive to a respondent bias, such as number of partners (Glick & Sahn, 2007; Kenyon et al., 2013; Morris et al., 2014). There is, however, no evidence that we could find that this bias would operate differentially by ethnic or regional group (Johnson et al., 2009; Kenyon et al., 2014a). This bias should thus result in underestimated the prevalence of reported risk behavior, but this effect should not differ by region. We did not control for a broad range of variables that may confound our results. Age of respondents, which is known to affect HIV prevalence and sexual behaviors such as lifetime partner number, was not controlled for. However, the DHS sampling strategy typically produces populations that do not differ by age between regions (Central Statistical Agency [Ethiopia] & ICF International, 2012; De Walque, 2006; Kenyon et al., 2013; Kenyon et al., 2015b). We also decided not to control for condom usage or upstream determinants such as socioeconomic status. In keeping with a number of other African studies (Kenyon et al., 2014a; Kenyon et al., 2013; Kenyon et al., 2015b), we found that condom usage tended to be higher in areas more affected by HIV (data not shown). As a result, we did not control for condom usage.\n\nThe relationship between socioeconomic status and HIV is complex, with most evidence pointing to a positive association between wealth and HIV in sub-Saharan Africa (Hajizadeh et al., 2014; Mishra et al., 2009; Susser, 1994). We found the same to be true at regional level (data not shown). Because our conceptual framework (Figure 6) conceived the upstream determinants as operating via sexual behavior, we considered it would be inappropriate to control for these in our model. The surveys were done at different stages in the HIV epidemics of the various countries. A wide range of studies have found that populations in sub-Saharan Africa have responded to the HIV epidemic by reducing a range of risk behaviors including casual partners and partner numbers (Glick & Sahn, 2007; Hajizadeh et al., 2014; Halperin et al., 2011; Kirby, 2008). By preferentially affecting persons with higher numbers of partners, AIDS mortality may also reduce the average number of partners reported in high prevalence settings (Chesson et al., 2003; Kenyon et al., 2016a). Not controlling for age of the epidemic, behavior change and AIDS mortalities, effect on partner number should, however, serve to dilute any relationship between sexual risk behaviors and HIV prevalence. We did not assess heterogeneity in HIV prevalence and high-risk behaviors within regions or the extent to which sexual networks were coterminous with regions within countries. Once again unconsidered heterogeneity would be expected to reduce the strength of the association between HIV and risk-behavior.\n\n(A) Men in Gambela (HIV prevalence 7%) report a higher number of lifetime sex partners (both median and percent with more than 6 partners – vertical red line) than those in the Somali region (HIV prevalence 0.9%) (Reproduced from [14] with permission). (B) This in conjunction with other determinants of enhanced network connectivity will result in a larger reachable path for HIV in Gambela than Somali. The reachable path is represented by the edges between nodes in the sexual networks. (C) The prevalence of other risk factors such as condom usage and circumcision influence the probability of HIV transmission per sexual contact (probability of transmission is depicted as proportional to edge width between nodes). (D) The composite of A, B and C determine the sexual network transmission index, which combines the reachable path determined by A and B with the probability of transmission (C) and results in a higher HIV prevalence (red nodes) in Gambela than Somali (E). Within both Gambela and Somali there is a stepwise increase in HIV prevalence with increasing number of partners, but given the higher sexual network transmission index in Gambela, the probability of HIV is higher for someone with a low number of partners in Gambela than someone with a high number of partners in Somali. (F) represents the HIV prevalence by number of lifetime sex partners for men 15–49 years from all DHS countries with available data. Graphic reproduced from [20] with permission. Somali most closely approximates India and Gambela, Tanzania.\n\nBecause HIV prevalence is a product of a number of risk factors operating over decades, one would not expect to consistently find a linear relationship between HIV prevalence and any particular risk factor by region. For this reason, in additional to the Pearson’s correlations, we compared the prevalence of each of the risk factors between the lowest and highest HIV prevalence region in each survey. These provided results commensurate with those provided by linear regression. If non-participation in the surveys or the HIV-testing component were not random this could result in non-response biases. However, response rates in most of the surveys were high. This study involves the ecological analysis of cross sectional data. As such we can only describe associations found and not attribute causation.\n\nConvincing evidence of a strong link between rate of partner change and risk of STIs including HIV has been provided by empirical individual level studies, modelling studies, and the theoretical importance of rate of partner change in the formula for the basic reproductive number (Auvert et al., 2001a; Chen et al., 2007; Mishra et al., 2009; Xu et al., 2006). Women reporting sex with a non-marital, non-cohabiting person have also been shown to be more likely to be HIV positive in all of 18 DHS’s where this was assessed (Mishra et al., 2009). The same was true for men in only 1/18 surveys (Vietnam) (Mishra et al., 2009). These findings at an individual level reduce the chance that the association we found between these same factors and HIV at an ecological level is due to an ecological inference fallacy. Furthermore, studies from nationally representative samples in several sub-Saharan African countries have found that region and/or ethnic group remain strongly associated with HIV after controlling for a range of standard individual level risk factors (adjusted odds ratios of up to 14.3 (95% CI 6.1- 33.4)(Barankanira et al., 2016; Fraser-Hurt et al., 2011; Johnson & Way, 2006; Johnson & Budlender, 2002; Mermin et al., 2008; Oluoch et al., 2011). This suggests that there is a risk factor at the level of region and/or ethnic group that was not adequately controlled for in these analyses. This risk factor could be a wide range of factors, such as HSV-2 prevalence (Weiss et al., 2001), composition of the vaginal microbiome (Buve et al., 2014), host genetic factors (Ramsuran et al., 2011) or differential network connectivity (Halperin et al., 2011; Kenyon et al., 2013; Kirby, 2008). Our ecological study is unable to assess which of these factors is responsible for the variations in HIV-prevalence by region. However, our results are compatible with the network-connectivity theory (Kenyon et al., 2017). In brief, this theory posits that populations vary in how connected their sexual networks are (determined by factors such as rate of partner change, concurrency). This, combined with the prevalence of other risk factors that affect the probability of transmission per contact (such as condom use, circumcision and STI prevalence), determines the prevalence of HIV (Kenyon et al., 2017) (see Figure 6 for a more detailed explanation). Why was higher risk sex the behavioral risk factor most strongly associated with HIV-prevalence? Having sex with a non-marital, non-cohabiting person may be an independent risk factor but it may also be associated with another risk behavior that was either unmeasured or inaccurately measured and this other risk behavior is a driver of HIV transmission. Given that the DHS methodology is poor at ascertaining socially sensitive information such as respondent concurrency, questions regarding less sensitive information (such as if a partner was non-marital, non-cohabiting) may be more accurately ascertained.\n\n\nConclusion\n\nThe available evidence, including the results from this study, suggests that a variety of combinations of behavioural and other risk factors result in high HIV prevalence (Buve, 2006; Kenyon et al., 2014a; Kenyon et al., 2013; Kenyon et al., 2015b; Laumann & Youm, 1999). A striking finding of this study was the strong positive associations between lifetime partners, multiple partners and higher risk sex and the negative association with debut in men. These associations suggest that these risk factors may be underpinned by a common factor. Evidence of varying strengths has been advanced for a wide range of upstream determinants including demographic, socioeconomic and norm-related factors (Abramsky et al., 2014; Barnighausen et al., 2007; Bowleg et al., 2011; Carter et al., 2007; Chen et al., 2007; Gorbach et al., 2002; Kenyon et al., 2014a; Kenyon et al., 2015a; Leclerc-Madlala, 2009; Mulawa et al., 2016; Richardson et al., 2014; Yamanis et al., 2016). More research is required to better delineate the relationship between these upstream factors, sexual-behaviors and the resulting sexual-networks and HIV prevalence. Future behavioral and HIV surveys would be strengthened by using audio-computer-assisted self-interview technology to collect sexual behavioral information (Le & Vu, 2012). In countries with large variations in HIV–prevalence thought could be given to using the sexual behaviour of the lower HIV-prevalence communities as positive examples for what could be achieved with behavior change in high HIV-prevalence communities (The Positive Deviance Approach in Female Genital Mutilation Eradication, 1999).\n\n\nData availability\n\nThe datasets analyzed during the current study are available in the MEASURE DHS repository, (http://www.measuredhs.com). Access to the dataset requires registration, and is granted to those that wish to use the data for legitimate research purposes. A guide for how to apply for dataset access is available at: https://dhsprogram.com/data/Access-Instructions.cfm. The exact datasets analyzed are detailed in Table 1.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in funding this study.\n\n\nAcknowledgements\n\nWe would like to thank MEASURE DHS for access to the DHS data of the 47 surveys included.\n\n\nReferences\n\nAbramsky T, Devries K, Kiss L, et al.: Findings from the SASA! Study: a cluster randomized controlled trial to assess the impact of a community mobilization intervention to prevent violence against women and reduce HIV risk in Kampala, Uganda. BMC Med. 2014; 12: 122. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAral SO, Lipshutz J, Blanchard J: Drivers of STD/HIV epidemiology and the timing and targets of STD/HIV prevention. Sex Transm Infect. 2007; 83 Suppl: 1 i1–4. PubMed Abstract \| Publisher Full Text\n\nArmstrong RA: When to use the Bonferroni correction. Ophthalmic Physiol Opt. 2014; 34(5): 502–508. PubMed Abstract \| Publisher Full Text\n\nAuvert B, Ballard R, Campbell C, et al.: HIV infection among youth in a South African mining town is associated with herpes simplex virus-2 seropositivity and sexual behaviour. AIDS. 2001a; 15(7): 885–898. PubMed Abstract \| Publisher Full Text\n\nAuvert B, Buvé A, Ferry B, et al.: Ecological and individual level analysis of risk factors for HIV infection in four urban populations in sub-Saharan Africa with different levels of HIV infection. AIDS. 2001b; 15 Suppl 4: S15–30. PubMed Abstract \| Publisher Full Text\n\nBarankanira E, Molinari N, Niyongabo T, et al.: Spatial analysis of HIV infection and associated individual characteristics in Burundi: indications for effective prevention. BMC Public Health. 2016; 16: 118. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBarnighausen T, Hosegood V, Timaeus IM, et al.: The socioeconomic determinants of HIV incidence: evidence from a longitudinal, population-based study in rural South Africa. AIDS. 2007; 21 Suppl 7: S29–38. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBeauclair R, Meng F, Deprez N, et al.: Evaluating audio computer assisted self-interviews in urban South African communities: evidence for good suitability and reduced social desirability bias of a cross-sectional survey on sexual behaviour. BMC Med Res Methodol. 2013; 13(1): 11. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBowleg L, Teti M, Massie JS, et al.: ‘What does it take to be a man? What is a real man?’: ideologies of masculinity and HIV sexual risk among Black heterosexual men. Cult Health Sex. 2011; 13(5): 545–559. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBuve A: The HIV epidemics in sub-Saharan Africa: why so severe? Why so heterogenous? An epidemiological perspective. The HIV/AIDS Epidemic in Sub-Saharan Africa in a Historical Perspective. 2006; 41–55.\n\nBuve A, Jespers V, Crucitti T, et al.: The vaginal microbiota and susceptibility to HIV. AIDS. 2014; 28(16): 2333–2344. PubMed Abstract \| Publisher Full Text\n\nCarter MW, Kraft JM, Koppenhaver T, et al.: \"A bull cannot be contained in a single kraal\": concurrent sexual partnerships in Botswana. AIDS Behav. 2007; 11(6): 822–830. PubMed Abstract \| Publisher Full Text\n\nCentral Statistical Agency [Ethiopia], ICF International: Ethiopia Demographic and Health Survey 2011. Retrieved from Addis Ababa, Ethiopia and Calverton, Maryland, USA. 2012. Reference Source\n\nChen L, Jha P, Stirling B, et al.: Sexual risk factors for HIV infection in early and advanced HIV epidemics in sub-Saharan Africa: systematic overview of 68 epidemiological studies. PLoS One. 2007; 2(10): e1001. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChesson HW, Dee TS, Aral SO: AIDS mortality may have contributed to the decline in syphilis rates in the United States in the 1990s. Sex Transm Dis. 2003; 30(5): 419–424. PubMed Abstract \| Publisher Full Text\n\nCleland JG, Ferry B: Sexual behaviour and AIDS in the developing world. Taylor & Francis Group. 1995. Reference Source\n\nDe Walque D: Who Gets AIDS and How?: The Determinants of HIV Infection and Sexual Behaviors in Burkina Faso, Cameroon, Ghana, Kenya, and Tanzania. World Bank Publications 2006; 3844. Publisher Full Text\n\nDrain PK, Smith JS, Hughes JP, et al.: Correlates of national HIV seroprevalence: an ecologic analysis of 122 developing countries. J Acquir Immune Defic Syndr. 2004; 35(4): 407–420. PubMed Abstract\n\nFraser-Hurt N, Zuma K, Njuho P, et al.: The HIV epidemic in South Africa: What do we know and how has it changed? Johannesburg: SANAC. 2011. Reference Source\n\nGlick P, Sahn DE: Are Africans practicing safer sex? Evidence from demographic and health surveys for eight countries. Evidence from Demographic and Health Surveys for Eight Countries (April 2007). Cornell Food and Nutrition Policy Program Working Paper (193). 2007. Reference Source\n\nGorbach PM, Stoner BP, Aral SO, et al.: \"It takes a village\": understanding concurrent sexual partnerships in Seattle, Washington. Sex Transm Dis. 2002; 29(8): 453–462. PubMed Abstract \| Publisher Full Text\n\nHajizadeh M, Sia D, Heymann SJ, et al.: Socioeconomic inequalities in HIV/AIDS prevalence in sub-Saharan African countries: evidence from the Demographic Health Surveys. Int J Equity Health. 2014; 13(1): 18. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHalperin DT, Mugurungi O, Hallett TB, et al.: A surprising prevention success: why did the HIV epidemic decline in Zimbabwe? PLoS Med. 2011; 8(2): e1000414. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJohnson K, Way A: Risk factors for HIV infection in a national adult population: evidence from the 2003 Kenya Demographic and Health Survey. J Acquir Immune Defic Syndr. 2006; 42(5): 627–636. PubMed Abstract \| Publisher Full Text\n\nJohnson L, Budlender D: HIV risk factors: A review of the demographic, socio-economic, biomedical and behavioural determinants of HIV prevalence in South Africa. Centre for Actuarial Research Monograph No. 8, University of Cape Town. 2002. Reference Source\n\nJohnson LF, Dorrington RE, Bradshaw D, et al.: Sexual behaviour patterns in South Africa and their association with the spread of HIV: insights from a mathematical model. Demogr Res. 2009; 21(11): 289–340. Publisher Full Text\n\nKenyon C, Buyze J, Colebunders R: HIV prevalence by race co-varies closely with concurrency and number of sex partners in South Africa. PLoS One. 2013; 8(5): e64080. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKenyon C, Colebunders R: Strong association between point-concurrency and national peak HIV prevalence. Int J Infect Dis. 2012; 16(11): e826–827. PubMed Abstract \| Publisher Full Text\n\nKenyon C, Colebunders R: Birds of a feather: homophily and sexual network structure in sub-Saharan Africa. Int J STD AIDS. 2013; 24(3): 211–215. PubMed Abstract \| Publisher Full Text\n\nKenyon C, Menten J, Vu L, et al.: Male circumcision and sexual risk behaviors may contribute to considerable ethnic disparities in HIV prevalence in Kenya: an ecological analysis. PLoS One. 2014a; 9(8): e106230. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKenyon C, Osbak K, Buyze J, et al.: Variations of Sexual Scripts Relating to Concurrency by Race, Class, and Gender in South Africa. J Sex Res. 2015a; 52(8): 878–86. PubMed Abstract \| Publisher Full Text\n\nKenyon CR, Buyze J, Klebanoff M, et al.: Incident Trichomonas vaginalis Is Associated With Partnership Concurrency: A Longitudinal Cohort Study. Sex Transm Dis. 2017; 44(11): 695–699. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKenyon CR, Osbak K, Buyze J: The Prevalence of HIV by Ethnic Group Is Correlated with HSV-2 and Syphilis Prevalence in Kenya, South Africa, the United Kingdom, and the United States. Interdiscip Perspect Infect Dis. 2014b; 2014: 284317. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKenyon CR, Osbak K, Buyze J: The Prevalence of HIV by Ethnic Group Is Correlated with HSV-2 and Syphilis Prevalence in Kenya, South Africa, the United Kingdom, and the United States. Interdiscip Perspect Infect Dis. 2014c; 2014: 284317. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKenyon CR, Schwartz I, Buyze J: Did AIDS mortality decrease the number of lifetime sexual partners in Kenya: an ecological analysis? Epidemiol Infect. 2016a; 144(3): 556–559. PubMed Abstract \| Publisher Full Text\n\nKenyon CR, Tsoumanis A, Schwartz IS: HIV Prevalence Correlates with High-Risk Sexual Behavior in Ethiopia's Regions. PLoS One. 2015b; 10(10): e0140835. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKenyon CR, Tsoumanis A, Schwartz IS: A population's higher-risk sexual behaviour is associated with its average sexual behaviour-An ecological analysis of subpopulations in Ethiopia, Kenya, South Africa, Uganda and the United States. Epidemics. 2016b; 15: 56–65. PubMed Abstract \| Publisher Full Text\n\nKirby D: Changes in sexual behaviour leading to the decline in the prevalence of HIV in Uganda: confirmation from multiple sources of evidence. Sex Transm Infect. 2008; 84 Suppl 2: ii35–ii41. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLaumann EO, Youm Y: Racial/ethnic group differences in the prevalence of sexually transmitted diseases in the United States: a network explanation. Sex Transm Dis. 1999; 26(5): 250–261. PubMed Abstract \| Publisher Full Text\n\nLe LC, Vu LT: Audio computer-assisted self interview compared to traditional interview in an HIV-related behavioral survey in Vietnam. MEDICC Rev. 2012; 14(4): 26–31. PubMed Abstract \| Publisher Full Text\n\nLeclerc-Madlala S: Cultural scripts for multiple and concurrent partnerships in southern Africa: why HIV prevention needs anthropology. Sex Health. 2009; 6(2): 103–110. PubMed Abstract \| Publisher Full Text\n\nMermin J, Musinguzi J, Opio A, et al.: Risk factors for recent HIV infection in Uganda. JAMA. 2008; 300(5): 540–549. PubMed Abstract \| Publisher Full Text\n\nMishra V, Medley A, Hong R, et al.: Levels and spread of HIV seroprevalence and associated factors: evidence from national household surveys. 2009. Reference Source\n\nMorris M, Epstein H, Wawer M: Timing is everything: international variations in historical sexual partnership concurrency and HIV prevalence. PLoS One. 2010; 5(11): e14092. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMorris M, Kurth AE, Hamilton DT, et al.: Concurrent partnerships and HIV prevalence disparities by race: linking science and public health practice. Am J Public Health. 2009; 99(6): 1023–1031. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMorris M, Vu L, Leslie-Cook A, et al.: Comparing Estimates of Multiple and Concurrent Partnerships Across Population Based Surveys: Implications for Combination HIV Prevention. AIDS Behav. 2014; 18(4): 783–790. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMulawa M, Yamanis TJ, Hill LM, et al.: Evidence of social network influence on multiple HIV risk behaviors and normative beliefs among young Tanzanian men. Soc Sci Med. 2016; 153: 35–43. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOluoch T, Mohammed I, Bunnell R, et al.: Correlates of HIV Infection Among Sexually Active Adults in Kenya: A National Population-Based Survey. Open AIDS J. 2011; 5(1): 125–34. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPerneger TV: What's wrong with Bonferroni adjustments. BMJ. 1998; 316(7139): 1236–1238. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRamsuran V, Kulkarni H, He W, et al.: Duffy-null-associated low neutrophil counts influence HIV-1 susceptibility in high-risk South African black women. Clin Infect Dis. 2011; 52(10): 1248–1256. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRichardson ET, Collins SE, Kung T, et al.: Gender inequality and HIV transmission: a global analysis. J Int AIDS Soc. 2014; 17(1): 19035. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRothman KJ: No adjustments are needed for multiple comparisons. Epidemiology. 1990; 1(1): 43–46. PubMed Abstract \| Publisher Full Text\n\nSawers L, Stillwaggon E: Concurrent sexual partnerships do not explain the HIV epidemics in Africa: a systematic review of the evidence. J Int AIDS Soc. 2010; 13(1): 34. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSusser M: The logic in ecological: I. The logic of analysis. Am J Public Health. 1994; 84(5): 825–829. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nThe Positive Deviance Approach in Female Genital Mutilation Eradication. Retrieved from Washington, DC. 1999.\n\nWeiss HA, Buvé A, Robinson NJ, et al.: The epidemiology of HSV-2 infection and its association with HIV infection in four urban African populations. AIDS. 2001; 15 Suppl 4: S97–108. PubMed Abstract \| Publisher Full Text\n\nWellings K, Collumbien M, Slaymaker E, et al.: Sexual behaviour in context: a global perspective. Lancet. 2006; 368(9548): 1706–1728. PubMed Abstract \| Publisher Full Text\n\nXu F, Sternberg MR, Kottiri BJ, et al.: Trends in herpes simplex virus type 1 and type 2 seroprevalence in the United States. JAMA. 2006; 296(8): 964–973. PubMed Abstract \| Publisher Full Text\n\nYamanis TJ, Fisher JC, Moody LW, et al.: Young Men's Social Network Characteristics and Associations with Sexual Partnership Concurrency in Tanzania. AIDS Behav. 2016; 20(6): 1244–1255. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "42625", "date": "09 Jan 2019", "name": "Nico Nagelkerke", "expertise": [ "Reviewer Expertise Biostatistics", "Epidemiology and Infectious disease modeling" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nKenyon and his coauthors explore whether self reported sexual \"risk\" taking on an aggregate level is associated with HIV prevalence. They present cogent arguments, supported by an extensive theoretical literature on sexual networks and propagation of STIs along such networks, why this is a valuable approach. I agree, but before reading their paper would have been pessimistic about the outcome because of the well-known misreporting of sexual behavior. That they nonetheless find some moderate to strong associations between (aggregated) self-reported risk behavior and HIV prevalence is not just a (methodological) justification of their approach but also (finally perhaps) a rationale for asking such behavior questions (in DHS and other surveys).\n\nThe paper is well written with appropriate figures and tables. All potential weaknesses were - as far as I could tell - recognized, acknowledged, and addressed by the authors. As regards to the statistical methods: although alternative approaches were possible, their approach was not flawed, and the choice of methods seemed appropriate.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "43673", "date": "06 Feb 2019", "name": "Brian G. Williams", "expertise": [ "Reviewer Expertise Mathematics", "Statistics", "Epidemiology", "HIV", "TB" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors carry out a detailed analysis of the association between risky sex and HIV prevalence at the regional level. They show that the prevalence of HIV is associated with risky sex. While this result is not unexpected it is important to quantify the relationship and to examine how the prevalence of HIV and risky sex vary among regions within countries.\nHowever, in my view a critical factor in determining the timing and peak prevalence of HIV epidemics depends critically on how people move. It is trivial to observe that sexually transmitted infections depend on direct, intimate contact between two individuals. If people don't move then, by definition, the epidemic cannot spread.\n\nDHS studies do not, of course, capture information on migration, and in particular oscillating migration, as is common in South Africa and its neighbouring countries and territories, and a direct consequence of the needs of the mining industry for a cheap supply of very large numbers of unskilled men rooted in the system of Apartheid. I would recommend reading this paper in parallel with the paper by John Hargrove \"Migration, mines and mores: the HIV epidemic in southern Africa\" published in the South African Journal of Science1.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1879
https://f1000research.com/articles/7-1878/v1	02 Dec 18	{ "type": "Method Article", "title": "Developing fit-for-purpose self-report instruments for assessing consumer responses to tobacco and nicotine products: the ABOUT™ Toolbox initiative", "authors": [ "Christelle Chrea", "Catherine Acquadro", "Esther F. Afolalu", "Erica Spies", "Thomas Salzberger", "Linda Abetz-Webb", "Stefan Cano", "Benoit Arnould", "Nelly Mainy", "Jed Rose", "Rolf Weitkunat", "Catherine Acquadro", "Esther F. Afolalu", "Erica Spies", "Thomas Salzberger", "Linda Abetz-Webb", "Stefan Cano", "Benoit Arnould", "Nelly Mainy", "Jed Rose", "Rolf Weitkunat" ], "abstract": "Background. Determining the public health impact of tobacco harm reduction strategies requires the assessment of consumer perception and behavior associated with tobacco and nicotine products (TNPs) with different exposure and risk profiles. In this context, rigorous methods to develop and validate psychometrically sound self-report instruments to measure consumers’ responses to TNPs are needed. Methods. Consistent with best practice guidelines, including the U.S. Food and Drug Administration’s “Guidance for Industry Patient-Reported Outcome Measures: Use in Medical Product Development to Support Labeling Claims,” scientifically designed, fit-for-purpose, reliable, and valid instruments are now being applied to tobacco regulatory research. Results. This brief report presents the ABOUT™ Toolbox (Assessment of Behavioral OUtcomes related to Tobacco and nicotine products) initiative. This communication: (1) describes the methodological steps followed for the development and validation of the measurement instruments included in the ABOUT™ Toolbox, (2) presents a summary of the high-priority tobacco-related domains that are currently covered in the ABOUT™ Toolbox (i.e., risk perception, dependence, product experience, health and functioning, and use history), and (3) details how the measurement instruments are made accessible to the scientific community. Conclusions. By making the ABOUT™ Toolbox available to the tobacco research and public health community, we envision a rapidly expanding knowledge base, with the goals of (1) supporting consumer perception and behavior research to allow comparisons across a wide spectrum of TNPs, (2) enabling public health and regulatory communities to make better-informed decisions for future regulation of TNPs, and (3) enhancing surveillance activities associated with the impact of TNPs on population health.", "keywords": [ "Modified risk tobacco products", "Reduced risk products", "Self-report instruments", "Behavior", "Consumer perception", "Best measurement practices" ], "content": "List of abbreviations\n\nABOUT: Assessment of Behavioral OUtcomes related to Tobacco and nicotine products;\n\nFDA: U.S. Food and Drug Administration; ICF: International Classification of Functioning, Disability, and Health; mCEQ: Modified Cigarette Evaluation Questionnaire MRTP: Modified risk tobacco product; PRO: Patient-reported outcomes; PROQOLID: Patient-Reported Outcome and Quality of Life Instruments Database; RMM: Rasch measurement methods; TA: Translatability assessment; TNP: Tobacco and nicotine product; WHO: World Health Organization.\n\n\nIntroduction\n\nMany stakeholders have recognized that there is a risk continuum for tobacco and nicotine products (TNPs)1,2. On this continuum, combustible products, cigarettes in particular, present the most risk, because burning tobacco creates the vast majority of the harmful and potentially harmful constituents implicated in the development of smoking-related diseases3. Cessation is at the lower end of the continuum, as it is the best way for smokers to lower their risk4. Alternative noncombustible TNPs (sometimes referred to as alternative nicotine delivery systems) that avoid combustion lie somewhere between these anchoring points of the continuum1. Tobacco harm reduction is an approach recognized by the U.S. Food and Drug Administration (FDA), the Institute of Medicine, and the World Health Organization (WHO) as part of a solution to more rapidly reduce the burden of preventable deaths and smoking-related diseases3,5–7. In the U.S., this has given rise to a regulatory framework for manufacturers to market modified-risk tobacco products (MRTPs), defined by the FDA as “any tobacco product that is sold or distributed for use to reduce harm or the risk of tobacco-related disease associated with commercially marketed tobacco products”8. However, to implement this approach successfully, consistent, transparent, and evidence-based science on the reduced risk potential of alternative products is paramount2.\n\nIn alignment with the FDA’s draft guidance on MRTPs, consumer perception and behavior assessments are key components of assessing the full public health impact of tobacco harm reduction8. Valid and reliable self-report measures are needed to assess consumer responses to MRTPs in comparison with other commercially available TNPs9–11. Although this has been acknowledged for quite some time10, the field of tobacco regulatory research falls short of specifically developed measures due to the lack of adherence to measurement best practices and specific guidelines that would facilitate standardization and harmonization of measures across studies (e.g., see a recent review on risk perception measurement in tobacco control research11). Some measurement and standardization initiatives have recently been proposed, the most predominant ones being the Patient-Reported Outcomes Measurement Information System Smoking Initiative12–17 and PhenX measures for Tobacco Regulatory Research (https://www.phenxtoolkit.org/collections/trr). It is worth noting that both initiatives focus primarily on combustible tobacco products, with their development based on legacy measures developed solely for cigarettes15. It is crucial, however, to attempt further efforts to develop new measurement instruments that would be “fit-for-purpose” to compare combustible and noncombustible products on the same risk continuum in order to better inform the public health decisions.\n\nWe present an ongoing collaborative effort to develop fit-for-purpose measurement instruments (i.e., concept-driven instruments providing interpretable outcomes for the purpose intended) to enhance the scientific framework of harm reduction. This new initiative has resulted in the creation of the ABOUT™ Toolbox (Assessment of Behavioral OUtcomes related to Tobacco and nicotine products).\n\nThe objectives of this communication are (1) to describe the methodological steps followed for the development and validation of the measurement instruments included in the ABOUT™ Toolbox, (2) to present a summary of the high-priority tobacco-related domains that are to be covered in the ABOUT™ Toolbox, and (3) to detail how the measurement instruments are to be made accessible to the scientific community.\n\n\nMethods\n\nThe ABOUT™ Toolbox has been developed using best measurement development practices and as the manifest of an underpinning behavioral conceptual model for TNPs.\n\nSeveral guidelines, including the FDA’s “Guidance for Industry Patient-Reported Outcome Measures: Use in Medical Product Development to Support Labeling Claims”18, have been used as the foundation for the creation of the ABOUT™ Toolbox initiative19,20. These guidelines provide the scientific basis for the development, modification, and validation of patient-reported outcome (PRO) measures in support of medical care research. Although not specifically designed for the tobacco regulatory research field, these recommendations are essential in outlining a wide range of development considerations, such as (1) defining the context of use and identifying the concept(s) of interest, (2) generating items that best capture the concept(s) of interest as expressed by the population of interest, (3) choosing the right response options and recall period, (4) evaluating the content validity of the instrument, and (5) assessing psychometric measurement properties for construct validity. These recommendations also cover the considerations around the adaptation of a self-report measure (e.g., for a different context of use, target population, or cultural group [see below the paragraph “cross-cultural equivalence”]).\n\nThe application of these best practices requires the use of mixed-methods research, which can be defined as “research in which the investigator collects and analyzes data, integrates the findings, and draws inferences using both qualitative and quantitative approaches, and methods in a single study or program of inquiry”21. As noted in a recent paper on rare disease patients22, qualitative methods alone are unable to inform us about the extent to which concepts are measurable. Conversely, quantitative methods alone cannot inform us about which concepts should be measured.\n\nBy applying these best measurement practices for the development of the ABOUT™ Toolbox (see Figure 1), the initiative further enhances tobacco regulatory research by combining five major components (described in the paragraphs below) that are paramount for rigorous instrument development.\n\n1. Generation of a conceptual framework\n\nThe development of each instrument starts with the generation of a conceptual framework, which is grounded in theory and supported by the triangulation of evidence data from literature reviews, consumer input, and expert opinions. Furthermore, the development of each measurement instrument in the ABOUT™ Toolbox has been or is being conducted in close partnership with scientific experts from academic and commercial organizations with expertise in the fields of nicotine addiction, motivational aspects of consumer perception, and relevant areas on approaches to measurement (e.g., PROs, cross-cultural adaptation, psychometrics, regulatory submissions). The role of these experts is pivotal not only to provide input during the development of the conceptual framework but also to assist in the objective consensus-building process throughout the entire development of the instrument.\n\n2. Evaluation of content validity\n\nThe main goal is to evaluate whether the instrument represents the concepts of interest, and the instructions and item content are appropriate, relevant, comprehensive, and understandable to the target population. This evaluation is performed in accordance with current good research practices23,24. The content validity of the instruments included in the ABOUT™ Toolbox is supported by the execution of qualitative research with consumers (e.g., 25–27).\n\n3. Use of an appropriate psychometric model\n\nThe psychometric assessment of most of the ABOUT™ instruments is based on the use of Rasch measurement methods (RMM)28, supplemented by diagnostic evidence of the dimensionality and reliability properties rooted in classical test theory29,30. RMM particularly investigate to what extent (1) the items are targeted for the type and range of issues to be measured31, (2) the items work well together as a set forming a unidimensional scale31–33, (3) the response options are working as intended34, and (4) the items do not show differential item functioning across various population groups or TNPs35. This evaluation provides the necessary evidence that the conceptual framework has been converted successfully into a list of items, the responses to which can be summed to form a statistically sufficient total score (i.e., comprising all available information). The score can then be transformed into a linear measure that is comparable across population groups or TNPs, conceptually meaningful and substantively interpretable (e.g., 36,37). The final outcome of the application of RMM is the development of a calibrated scoring table that transfers sum scores to logit measures, which are mapped to a 0–100 scale for ease of interpretability. The conversion is a simple, linear transformation that changes the logit mean of 0 to 50 and converts most extreme measures to 0 and 100, respectively (e.g., 37). The purpose, description, administration, and scoring (including the calibrated scoring table) of the final validated version of an instrument is documented in a user manual that accompanies all the instruments of the ABOUT™ Toolbox.\n\n4. Cross-cultural equivalence of the ABOUT™ instruments\n\nIn the context of the globalization of tobacco regulatory research, measures appropriate for use in different cultures are crucial. The demonstration of cross-cultural equivalence requires investigating that the instrument measures the same concepts in a comparable way across different languages and cultures38–40. The first step toward cross-cultural equivalence is to ensure that rigorous translation procedures are followed. In the field of health outcomes research, recommendations to support the use of self-reported measures in multinational contexts have been provided41,42. These recommendations apply to measures developed in one language and subsequently translated for use in other countries and cultures. In line with this guidance, the instruments included in the ABOUT™ Toolbox follow a thorough linguistic validation process consisting of two forward translations to the targeted new language, one back translation to the source language, and qualitative cognitive debriefing interviews with participants in the targeted language to ensure that the translations are understood (e.g., 43 and Figure 2)44.\n\nThe investigation and demonstration of the cross-cultural equivalence of the ABOUT™ instruments are completed by quantitative steps (e.g., 37,45), such as the evaluation of the psychometric properties of the translations and differential item functioning46.\n\nTranslatability assessment (TA) is a technique that will be applied to all future ABOUT™ instruments. It is defined as the review of an original measure, preferably during the development stage, prior to its use, in order to determine its suitability for future translations in multilingual studies47. TA can be viewed as the very first step toward ensuring measurement equivalence between the original measure and its future translations.\n\n5. Appropriate access and use of the validated instruments (original and translations)\n\nEasy, centralized, and appropriate access to original instruments and their translations is often a prerequisite to efficient research. With a unique point of access endorsed by the developers of instruments, researchers are able to access and use the original instrument and its translations48. The centralization and control of access also enables the integrity of the each measurement instrument (original and translations) to be respected49. Once the instruments are fully developed and validated (Steps 1 and 2 on Figure 1), they are made publicly available through the Patient-Reported Outcome and Quality of Life Instruments Database (PROQOLID™)50, managed by Mapi Research Trust and part of the ePROVIDE™ web platform (https://eprovide.mapi-trust.org/).\n\nTo get access to the instruments distributed on ePROVIDE™, each new user has to complete a free registration form at https://eprovide.mapi-trust.org/login (click to the Free Registration button). The following link leads to a tutorial to help navigating through the whole registration process: https://eprovide.mapi-trust.org/tutorials/registering-for-free. Once the registration is completed, the register button will lead the user to his/her free ePROVIDE™ account. From there, any retrieval of instruments can be performed.\n\nInstruments and their respective user manuals can be retrieved through the main search engine of the database by using either the full name or parts of the name (e.g., ABOUT™–Perceived Risk, Perceived, or Risk) or within the “Our Catalog” section of ePROVIDE™ (https://eprovide.mapi-trust.org/catalog) by filtering through the Behavior and Behavior Mechanisms category.\n\nTo be able to use any of the ABOUT™ Toolbox instruments, all users (whether they are commercial or non-commercial users) will have to accept the conditions of a license/user agreement and complete the appropriate form. This license, issued by Mapi Research Trust, which is the official distributor of the instruments, specifies the terms under which each instrument should be used: Special terms, i.e., specific to each instrument, and General terms: https://eprovide.mapi-trust.org/user-agreement-general-terms.\n\nTutorials are available on how to submit requests (commercial users) or download copies directly (non-commercial users) (see https://eprovide.mapi-trust.org/faq). All requests are handled by the PROVIDE team of Mapi Research Trust.\n\n\nResults\n\nThe ABOUT™ Toolbox currently comprises five measurement instruments, which are at various stages of development and validation. Table 1 presents a summary of the relevant domains, concepts of interest, and context of use to be covered by the instruments. The initiative is to be expanded by additional domains as they are identified. The initial inclusion of these current instruments were informed extensively by existing research and domains that have been prioritized based on public health impact and issues of key importance to tobacco regulatory research.\n\nPROQOLID™, Patient-Reported Outcome and Quality of Life Instruments Database; TNPs, Tobacco and Nicotine Products.\n\n1. ABOUT™−Perceived Risk\n\nAssessment of risk perception is an important domain of tobacco-related behaviors and influences any tobacco harm reduction strategies aimed at getting individuals to switch to less harmful alternatives to cigarette smoking1,11.\n\nThe Perceived Risk Instrument was developed as a multiscale instrument intended to assess the perceived risks associated with use across a range of different of TNPs, relative to other products, cessation aids, and quitting all tobacco products. The health risk and addiction risk scales have been calibrated in three countries (U.S., Italy, and Japan) for two perception types: risk to the individual respondent (personal risk) and risk to users of the product in general (general risk)25,37. The validation of further scales for perceived social and practical risk is currently under development.\n\n2. ABOUT™−Use History\n\nOne of the key aims underlying tobacco control and harm reduction is to reduce the burden of smoking-related diseases through the implementation of monitoring strategies for tobacco consumption and characterization of key patterns and trends in tobacco use3. The Smoking Questionnaire, included in the ABOUT™ Toolbox, was developed to provide a core set of questions that cover the major dimensions of TNP use and is consistent with criteria used for defining smoking history and status51. It assesses frequency and intensity of current and past TNP use behavior, initiation, and cessation and demonstrates good test-retest reliability52,53.\n\n1. ABOUT™−Product Experience\n\nProduct experience encompasses a range of self-reported expressions of an individual’s experience using a TNP and is a key predictive measure of short-term preference and long-term TNP use54. The modified Cigarette Evaluation Questionnaire (mCEQ) has been endorsed by regulatory and public health bodies to use in the context of MRTP assessment9,55. The ABOUT™ Toolbox includes a measure consisting of two multi-item scales and three single-item scales arising from an adaptation and rewording of the mCEQ56 and the Product Evaluation Scale57. The scales focus on satisfaction, psychological reward, craving reduction, aversion, and enjoyment of respiratory tract sensation. Psychometric testing and validation have been carried out for the use of the measure to assess cigarettes and heat-not-burn products45, and further validation is ongoing for a variety of TNPs (e.g., e-cigarettes, cigars/cigarillos, smokeless products) and different recall periods.\n\n2. ABOUT™−Dependence\n\nNicotine dependence has been shown to be a primary driver of smoking and TNP use behaviors58. However, dependence on nicotine has generally focused on cigarette smoking. Evidence suggests that measuring dependence with a common set of symptoms across different TNP use groups is feasible and may better reflect the dynamics of dependence across a range of products59–61. Work is ongoing to develop such a fit-for-purpose dependence instrument for inclusion in the ABOUT™ Toolbox. The proposed instrument is rooted in a conceptual framework of dependence that identifies lack of control (e.g., urgency to use upon waking up, difficulty to cease using, self-awareness of dependence) as the core construct. Recommended practices in PRO development have been used to generate a draft instrument from a range of qualitative research steps, including literature review, concept elicitation, and cognitive debriefing interviews with different groups of adult tobacco users27. The draft instrument is currently undergoing quantitative field testing to assess psychometric properties and produce a final validated measure.\n\n3. ABOUT™−Health and Functioning\n\nHealth and functioning is a relevant dimension for the evaluation of TNP impact on public health and requires further investigation8,9,58. There are no established measures specific to tobacco-related health outcomes to date. In addition, existing generic health and functioning instruments do not capture the small, but potentially important, concepts that may change when a smoker switches to an MRTP. Currently, efforts are ongoing to develop a new outcome measure for inclusion in the ABOUT™ Toolbox that would accurately reflect the health and functioning status of individuals who use TNPs, with a particular focus on healthy adult smokers who switch to MRTPs. The measure’s development will be underpinned by conceptual frameworks including, but not limited to, the WHO’s International Classification of Functioning, Disability, and Health (ICF)62 and the Revised Wilson and Cleary Model for Health-Related Quality of Life63. The ICF is particularly well suited to serve as a guiding framework, as it conceptualizes a person’s level of functioning as a dynamic interaction between body structure and functions, health conditions, social participation, personal factors, and environmental factors. The Wilson and Cleary Model is included to address subjective dimensions of health and functioning, such as well-being and health-related quality of life. Literature review, an expert panel, and qualitative research with a wide range of consumers are planned for the initial phase of the project.\n\n\nDiscussion and Conclusion\n\nIn the present paper, we have presented a new initiative aiming at enhancing the scientific framework of harm reduction and promoting the establishment of consensus and standardized tools to be used across tobacco regulatory research studies. Within the new regulatory MRTP pathway, FDA can issue an order authorizing the marketing of MRTPs. To do so, data must demonstrate that use of the new product (1) would present a significantly lower risk of harm to the individual user and (2) would reduce the incidence of harm in the population as a whole (e.g., with evidence that marketing the new product as a less risky alternative to cigarettes would not increase use of the new product by non-smokers), as described in section 911 of the Federal Food Drug and Cosmetic Act64. Since the inception of the MRTP pathway in 2011, FDA has received 35 MRTP applications and granted zero Modified Risk Orders (https://www.fda.gov/TobaccoProducts/Labeling/TobaccoProductReviewEvaluation/ucm304465.htm#2). This suggests that (1) no manufacturer has yet successfully demonstrated the harm reduction potential of any new TNP to the FDA and (2) the FDA and manufacturers have not shared a common understanding of what classifies consistent, transparent, and evidence-based science. Therefore, it is of paramount importance to establish a common, well-defined understanding of the types and volume of research data that would demonstrate a candidate MRTP to be appropriate for the protection of the public health to the FDA and other regulatory authorities. As expressed by the National Institutes of Health Tobacco Regulatory Science Program and the FDA Center for Tobacco Products, “One way to accomplish this is to provide investigators with a common set of tools and resources to facilitate sharing, comparing, replication of findings, and integration of data from multiple sources”65.\n\nAs part of this effort, the current ABOUT™ Toolbox may facilitate progress toward consensus of domains to be assessed and consensus on how they should be measured and reported. With the development and dissemination of the ABOUT™ Toolbox, researchers will have access to instruments that are (1) developed and validated with state-of-the-science methods to be psychometrically sound, straightforward to implement in clinical and population-based studies, and easy to interpret; (2) created to be relevant and applicable across the whole spectrum of TNPs and across various population groups; and (3) designed to enhance standardization and comparison of data on perception and behaviors toward MRTPs across academic, industry, and public health research communities.\n\nThe current measurement instruments highlighted in the ABOUT™ Toolbox fit within what can be described as a broader behavioral conceptual model, designed to understand TNP switching or transition behaviors, which encompasses several levels of assessment (i.e., individual, product, and environment [Spies et al., manuscript in preparation]). This conceptual model evolved from the review of several existing frameworks that propose explanations of and factors associated with TNP use10,66,67 and was complemented by literature on principles of behavioral changes, taking a socio-ecological approach to the conceptualization of TNP switching or transition behaviors68. Each of the individual, product, and environment levels includes several categories for which concepts and variables are defined. For instance, the individual level includes individual traits (e.g., dependence), attitudes and beliefs toward products (e.g., perceived risk), response to the product (e.g., satisfaction), self-reported product behavior (e.g., consumption changes), and functional health and quality of life. Instruments within the ABOUT™ Toolbox are intended to be used within this conceptual model to measure each of these concepts of interest in a standardized way.\n\nBy making the ABOUT™ Toolbox available to the tobacco research and public health communities through PROQOLID™, we envision a rapidly expanding knowledge base, with the goals of not only advancing further the interpretation of consumer perception data comparing a large spectrum of TNPs but also enabling public health and regulatory communities to make better-informed decisions for future regulation of MRTPs and to enhance surveillance activities associated with smoking-related disease. The ABOUT™ Toolbox launches a dialogue on new perspectives required to develop standards and best practices in the spirit of current guidance for self-reported measures and may encourage the creation of a consortium to work on standard measures across the industry.\n\n\nData availability\n\nAccess to the measurement instruments, as well as further information on the original versions and translations is freely available for non-commercial use on ePROVIDE™ (https://eprovide.mapi-trust.org).", "appendix": "Grant information\n\nThe work presented in this communication was sponsored by Philip Morris International.\n\n\nAcknowledgements\n\nThe authors would like to thank the following individuals for their participation in the development of the ABOUT™ Toolbox:\n\n• Patricia Binggeli; Emilie Clerc, B.S.; Sophie Gallot, M.S.; Frank Lüdicke, M.D.; Pierpaolo Magnani; Antonio Ramazzotti; and Steve Roulet at Philip Morris International\n\n• Vanessa Brunel; Jeremy Lambert, Ph.D.; and Celine Marcuccilli at Mapi, an ICON plc Company\n\n• Thomas Alfieri, Ph.D.; Mark Atkison, M.S.; Stephen Crawford, Ph.D.; and Marie Martin, Ph.D., at Covance Market Access Services, Inc.\n\n• Karl Fagerström, Ph.D., at K. Fagerström Consulting AB\n\n\nReferences\n\nAbrams DB, Glasser AM, Pearson JL, et al.: Harm Minimization and Tobacco Control: Reframing Societal Views of Nicotine Use to Rapidly Save Lives. Annu Rev Public Health. 2018; 39: 193–213. PubMed Abstract \| Publisher Full Text\n\nBerman ML, Connolly G, Cummings KM, et al.: Providing a Science Base for the Evaluation of Tobacco Products. Tob Regul Sci. 2015; 1(1): 76–93. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nU.S. Department of Health and Human Services: The Health Consequences of Smoking: 50 Years of Progress. A Report of the Surgeon General. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, 2014; Last Accessed 2 October 2018. Reference Source\n\nU.S. Department of Health and Human Services: The Health Benefits of Smoking Cessation. A Report of the Surgeon General. Rockville (MD): U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control, Center for Chronic Disease Prevention and Health Promotion Office on Smoking and Health. DHHS Publication No. (CDC) YO-K-116. 1990; Last accessed 2 October 2018. Reference Source\n\nZeller M, Hatsukami D, Strategic Dialogue on Tobacco Harm Reduction Group: The Strategic Dialogue on Tobacco Harm Reduction: a vision and blueprint for action in the US. Tob Control. 2009; 18(4): 324–332. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nInstitute of Medicine: Clearing the Smoke: Assessing the Science Base for Tobacco Harm Reduction. Washington, DC: The National Academies Press, 2001; Last accessed 2 October 2018. Publisher Full Text\n\nWorld Health Organization: Monograph: Advancing Knowledge on Regulating Tobacco Products. Geneva: World Health Organization, 2009; Last accessed 2 October 2018. Reference Source\n\nU.S. Department of Health and Human Services: Guidance for industry - Modified Risk Tobacco Product Applications - Draft Guidance. Silver Spring (MD): U.S. Department of Health and Human Services, FDA, Center for Tobacco Products, 2012; Last accessed 2 October 2018. Reference Source\n\nInstitute of Medicine: Scientific standards for studies on modified risk tobacco products. Washington, DC: The National Academies Press, 2012; Last accessed 2 October 2018. Publisher Full Text\n\nRees VW, Kreslake JM, Cummings KM, et al.: Assessing consumer responses to potential reduced-exposure tobacco products: a review of tobacco industry and independent research methods. Cancer Epidemiol Biomarkers Prev. 2009; 18(12): 3225–3240. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKaufman AR, Persoskie A, Twesten J, et al.: A review of risk perception measurement in tobacco control research. Tob Control. 2018; pii: tobaccocontrol-2017-054005. PubMed Abstract \| Publisher Full Text\n\nEdelen MO: The PROMIS smoking assessment toolkit--background and introduction to supplement. Nicotine Tob Res. 2014; 16 Suppl 3: S170–174. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHansen M, Cai L, Stucky BD, et al.: Methodology for developing and evaluating the PROMIS smoking item banks. Nicotine Tob Res. 2014; 16 Suppl 3: S175–189. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEdelen MO, Stucky BD, Hansen M, et al.: The PROMIS Smoking Initiative: initial validity evidence for six new smoking item banks. Nicotine Tob Res. 2014; 16 Suppl 3: S250–260. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNiaura R: Delivering on its promises: the PROMIS Smoking Initiative item banks. Nicotine Tob Res. 2014; 16 Suppl 3: S261–262. PubMed Abstract \| Publisher Full Text\n\nEdelen MO, Huang W, Stucky BD: Additional validity evidence for the PROMIS Smoking Assessment Toolkit. Addict Behav. 2016; 58: 80–84. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStucky BD, Huang W, Edelen MO: The Psychometric Performance of the PROMIS Smoking Assessment Toolkit: Comparisons of Real-Data Computer Adaptive Tests, Short Forms, and Mode of Administration. Nicotine Tob Res. 2016; 18(3): 361–365. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nU.S. Department of Health and Human Services: Guidance for Industry - Patient-Reported Outcome Measures: Use in Medical Product Development to Support Labeling Claims. Rockville (MD): U.S. Department of Health and Human Services FDA Center for Drug Evaluation and Research, Center for Biologics Evaluation and Research, Center for Devices and Radiological Health, 2009; Last accessed 2 October 2018. Reference Source\n\nMokkink LB, Terwee CB, Knol DL, et al.: The COSMIN checklist for evaluating the methodological quality of studies on measurement properties: a clarification of its content. BMC Med Res Methodol. 2010; 10: 22. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAaronson N, Alonso J, Burnam A, et al.: Assessing health status and quality-of-life instruments: attributes and review criteria. Qual Life Res. 2002; 11(3): 193–205. PubMed Abstract \| Publisher Full Text\n\nTashakkori A, Creswell JW: Editorial: Exploring the nature of research questions in mixed methods research. J Mix Methods Res. 2007; 1(3): 207–211. Publisher Full Text\n\nMorel T, Cano SJ: Measuring what matters to rare disease patients - reflections on the work by the IRDiRC taskforce on patient-centered outcome measures. Orphanet J Rare Dis. 2017; 12(1): 171. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPatrick DL, Burke LB, Gwaltney CJ, et al.: Content validity--establishing and reporting the evidence in newly developed patient-reported outcomes (PRO) instruments for medical product evaluation: ISPOR PRO Good Research Practices Task Force report: part 2--assessing respondent understanding. Value Health. 2011; 14(8): 978–988. PubMed Abstract \| Publisher Full Text\n\nPatrick DL, Burke LB, Gwaltney CJ, et al.: Content validity--establishing and reporting the evidence in newly developed patient-reported outcomes (PRO) instruments for medical product evaluation: ISPOR PRO good research practices task force report: part 1--eliciting concepts for a new PRO instrument. Value Health. 2011; 14(8): 967–977. PubMed Abstract \| Publisher Full Text\n\nSalzberger T, Chrea C, Cano SJ, et al.: Perceived risks associated with the use of tobacco and nicotine-containing products: Findings from qualitative research. Tobacco Science and Technology. 2017; 50(Suppl 1): 32–42. Reference Source\n\nAtkinson M, Chrea C, Salzberger T, et al.: Factors Influencing Consumers' Perceived Risk Of Tobacco Products. Value Health. 2015; 18(7): A505. PubMed Abstract \| Publisher Full Text\n\nChrea C, Salzberger T, Abetz-Webb L, et al.: Development of a tobacco and nicotine products dependence instrument.Poster presented at the Society for Research on Nicotine and Tobacco 2018 Annual Meeting, Baltimore, MD USA. Last accessed 2 October 2018. Reference Source\n\nAndrich D: Rasch models for measurement. Beverly Hills: Sage Publications; 1988. Reference Source\n\nCronbach LJ: Coefficient alpha and the internal structure of tests. Psychometrika. 1951; 16(3): 297–334. Publisher Full Text\n\nDeVellis RF: Scale development: theory and applications. USA: Sage Publications; 2003. Reference Source\n\nAndrich D: Rating scales and Rasch measurement. Expert Rev Pharmacoecon Outcomes Res. 2011; 11(5): 571–585. PubMed Abstract \| Publisher Full Text\n\nAndrich D: An elaboration of Guttman scaling with Rasch models for measurement. Sociol Methodol. 1985; 15: 33–80. Publisher Full Text\n\nAndrich D, de Jong JHAL, Sheridan BE: Diagnostic opportunities with the Rasch model for ordered response categories. In Applications of latent trait and latent class models in the social sciences. Edited by Rost J, Langeheine R: Waxmann Publishing Co.; 1997; 59–70. Reference Source\n\nAndrich D: A rating formulation for ordered response categories. Psychometrika. 1978; 43(4): 561–573. Publisher Full Text\n\nAndrich D, Hagquist C: Real and artificial differential item functioning. J Educ Behav Stat. 2012; 37(3): 387–416. Publisher Full Text\n\nChrea C, Cano S, Salzberger T, et al.: Using Rasch measurement to quantify the perceived risks associated with the use of tobacco and nicotine-containing products. Value Health. 2017; 20(9): A765–A766. Publisher Full Text\n\nCano S, Chrea C, Salzberger T, et al.: Development and validation of a new instrument to measure perceived risks associated with the use of tobacco and nicotine-containing products. Health Qual Life Outcomes. 2018; 16(1): 192. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHerdman M, Fox-Rushby J, Badia X: 'Equivalence' and the translation and adaptation of health-related quality of life questionnaires. Qual Life Res. 1997; 6(3): 237–247. PubMed Abstract \| Publisher Full Text\n\nHerdman M, Fox-Rushby J, Badia X: A model of equivalence in the cultural adaptation of HRQoL instruments: the universalist approach. Qual Life Res. 1998; 7(4): 323–335. PubMed Abstract \| Publisher Full Text\n\nRegnault A, Hamel JF, Patrick DL: Pooling of cross-cultural PRO data in multinational clinical trials: how much can poor measurement affect statistical power? Qual Life Res. 2015; 24(2): 273–277. PubMed Abstract \| Publisher Full Text\n\nWild D, Grove A, Martin M, et al.: Principles of Good Practice for the Translation and Cultural Adaptation Process for Patient-Reported Outcomes (PRO) Measures: report of the ISPOR Task Force for Translation and Cultural Adaptation. Value Health. 2005; 8(2): 94–104. PubMed Abstract \| Publisher Full Text\n\nWild D, Eremenco S, Mear I, et al.: Multinational trials-recommendations on the translations required, approaches to using the same language in different countries, and the approaches to support pooling the data: the ISPOR Patient-Reported Outcomes Translation and Linguistic Validation Good Research Practices Task Force report. Value Health. 2009; 12(4): 430–440. PubMed Abstract \| Publisher Full Text\n\nAcquadro C, Belay A, Popielnicki A, et al.: Linguistic validation of the Perceived Risk Instrument (PRI) into French, German, Italian, Japanese, Polish and Russian.Poster presented at the Society for Research on Nicotine and Tobacco 2018 Annual Meeting, Baltimore, MD, USA. Last accessed 2 October 2018. Reference Source\n\nMear I, Giroudet C: Chapter 1: Linguistic Validation Procedures. In Linguistic Validation Manual for Health Outcome Assessments. Edited by Acquadro C, Conway K, Giroudet C, et al. Lyon, France: Mapi Institute; 2012; 15–117.\n\nSalzberger T, Cano S, Mainy N, et al.: Psychometric evaluation of the mCEQ applied to cigarettes and heat-not-burn products in the US and Japan.Poster presented at the Society for Research on Nicotine and Tobacco 2018 Annual Meeting, Baltimore, MD, USA. Last accessed 2 October 2018. Reference Source\n\nRegnault A, Herdman M: Using quantitative methods within the Universalist model framework to explore the cross-cultural equivalence of patient-reported outcome instruments. Qual Life Res. 2015; 24(1): 115–124. PubMed Abstract \| Publisher Full Text\n\nAcquadro C, Patrick DL, Eremenco S, et al.: Emerging good practices for Translatability Assessment (TA) of Patient-Reported Outcome (PRO) measures. J Patient Rep Outcomes. 2017; 2(1): 8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAnfray C, Emery MP, Conway K, et al.: The advantages of a centralized dissemination stratefy for health outcomes instruments and their translations: A case exemple with the Zarit Burden Interview (ZBI). Value Health. 2009; 12(7): A399.\n\nAnfray C, Emery MP, Conway K, et al.: Questions of copyright. Health Qual Life Outcomes. 2012; 10: 16. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEmery MP, Perrier LL, Acquadro C: Patient-reported outcome and quality of life instruments database (PROQOLID): frequently asked questions. Health Qual Life Outcomes. 2005; 3: 12. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWorld Health Organization: Monitoring tobacco use. In Guidelines for Controlling and Monitoring the Tobacco Epidemic. World Health Organization; 1998; 76–101.\n\nWeitkunat R, Coggins CRE, Sponsiello-Wang Z, et al.: Assessment of cigarette smoking in epidemiologic studies. Beiträge zur Tabakforschung International/Contributions to Tobacco Research. 2013; 25(7): 638–648. Publisher Full Text\n\nSponsiello-Wang Z, de La Bourdonnaye G, David M, et al.: Accuracy of the smoking questionnaire. Beiträge zur Tabakforschung International/Contributions to Tobacco Research. 2017; 27(8): 224–239. Publisher Full Text\n\nO'Connor RJ, Lindgren BR, Schneller LM, et al.: Evaluating the utility of subjective effects measures for predicting product sampling, enrollment, and retention in a clinical trial of a smokeless tobacco product. Addict Behav. 2018; 76: 95–99. PubMed Abstract \| Publisher Full Text\n\nHanson K, O'Connor R, Hatsukami D: Measures for assessing subjective effects of potential reduced-exposure products. Cancer Epidemiol Biomarkers Prev. 2009; 18(12): 3209–3224. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCappelleri JC, Bushmakin AG, Baker CL, et al.: Confirmatory factor analyses and reliability of the modified cigarette evaluation questionnaire. Addict Behav. 2007; 32(5): 912–923. PubMed Abstract \| Publisher Full Text\n\nHatsukami DK, Zhang Y, O'Connor RJ, et al.: Subjective responses to oral tobacco products: scale validation. Nicotine Tob Res. 2013; 15(7): 1259–1264. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNational Academies of Sciences, Engineering, and Medicine: Public Health Consequences of E-Cigarettes. Washington, DC: The National Academies Press, 2018; Last accessed 2 October 2018. Reference Source\n\nFagerstrom K, Eissenberg T: Dependence on tobacco and nicotine products: a case for product-specific assessment. Nicotine Tob Res. 2012; 14(11): 1382–1390. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStrong DR, Messer K, Hartman SJ, et al.: Measurement of multiple nicotine dependence domains among cigarette, non-cigarette and poly-tobacco users: Insights from item response theory. Drug Alcohol Depend. 2015; 152: 185–193. PubMed Abstract \| Free Full Text\n\nStrong DR, Pearson J, Ehlke S, et al.: Indicators of dependence for different types of tobacco product users: Descriptive findings from Wave 1 (2013-2014) of the Population Assessment of Tobacco and Health (PATH) study. Drug Alcohol Depend. 2017; 178: 257–266. PubMed Abstract\n\nWorld Health Organization: International Classification of Functioning, Disability and Health: ICF. Geneva: World Health Organization, 2001; Last accessed 2 October 2018. Reference Source\n\nFerrans CE, Zerwic JJ, Wilbur JE, et al.: Conceptual model of health-related quality of life. J Nurs Scholarsh. 2005; 37(4): 336–342. PubMed Abstract \| Publisher Full Text\n\nSection 911 of the Federal Food, Drug, and Cosmetic Act - Modified Risk Tobacco Products. Last accessed 2 October 2018. Reference Source\n\nNotice Announcing Availability of Common Data Elements for Tobacco Regulatory Research via the PhenX Toolkit. Last accessed 9 October 2018. Reference Source\n\nFong GT, Cummings KM, Borland R, et al.: The conceptual framework of the International Tobacco Control (ITC) Policy Evaluation Project. Tob Control. 2006; 15 Suppl 3: iii3–11. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHyland A, Ambrose BK, Conway KP, et al.: Design and methods of the Population Assessment of Tobacco and Health (PATH) Study. Tob Control. 2017; 26(4): 371–378. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMcLeroy KR, Bibeau D, Steckler A, et al.: An ecological perspective on health promotion programs. Health Educ Q. 1988; 15(4): 351–377. PubMed Abstract \| Publisher Full Text" }	[ { "id": "43288", "date": "12 Feb 2019", "name": "Christopher Proctor", "expertise": [ "Reviewer Expertise Tobacco regulatory science" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors tackle an important issue in tobacco regulatory science - how to integrate data on the potential impact of the introduction of potentially reduced risk tobacco and nicotine products on the population. It draws on approaches to labelling claims on medical products, and as such may miss some important elements of what may influence behaviours and use of what are in most jurisdictions consumer products. Some of these may become apparent as the tool is used in a multi-national context. For example, the cost of the devices of vaping or 'heat-not-burn' may have a major influence on the likely uptake in countries with low disposable income. Other cultural influences may also become important, and so rather than a translation followed by back translation, additional items may need to be added in some countries.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Partly\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? No source data required\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Partly", "responses": [] }, { "id": "44881", "date": "08 Mar 2019", "name": "Michael Balls", "expertise": [ "Reviewer Expertise Our experience as toxicologists is primarily in the development", "validation", "acceptance and application of non-animal toxicity test procedures and integrated test systems", "and in particular", "those which involve human cells and tissues." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article deals with an important issue facing society today – how to encourage tobacco harm reduction without introducing new and undefinable hazards, and to make possible the perception of risk by individuals using a range of devices. While not wishing to question the integrity of the authors or the value of the thought and effort they have put into developing ABOUTTM Toolbox, we think it essential that wider questions related to the development of safer smoking materials and alternative sources of nicotine are considered, before turning to specific aspects of the ABOUTTM Toolbox proposal itself.\nIn the Introduction, the authors refer to the continuum between cigarette smoking at one extreme and cessation at the other. However, it is not clear whether “cessation” refers only to tobacco products, including non-combustible TNPs, or to all forms of voluntary uptake of nicotine. Our own position, as toxicologists, is clear. Nicotine is a very toxic and addictive substance, so its use should be actively discouraged. Therefore, the increasing use of electronic cigarettes (ECs) is a matter of great public health concern.\nAnother continuum is that between heavy cigarette smokers and individuals who have never smoked at all. Encouraging heavy smokers to switch to ECs in order to reduce their risk of tobacco harm can be a helpful strategy, but encouraging non-smokers to use ECs can only be seen as highly undesirable.\nPublic Health England (PHE), an executive agency of the Department of Health and Social Care, has stated that ECs are 95% safer than tobacco cigarettes. Such a statement is without scientific justification, as there are no compatible quantitative measures of the risks involved in smoking cigarettes or ECs, which would permit such a comparison between them to be made. The statement is based on the conclusions of a small number of experts who advise PHE.\n\nSafety is related to risk, which is the product of hazard and exposure. However, there is no satisfactory body of scientific evidence concerning the hazard represented by the vapours produced by ECs. The effects of exposure to EC vapour must also be highly variable, as they will depend on a variety of factors, including the type of EC used, the sources and levels of its nicotine and other additives, puff volume, puff frequency, the extent of inhalation, and the age, previous smoking history, health conditions affecting the respiratory system, such as asthma, and genetic polymorphism in relation to, for example, nicotine biotransformation enzymes, and nicotinic-acetylcholinergic receptor subunit structures, which affect binding affinity for nicotine. In combination, these and other factors must mean that there is also a continuum of individuals subject to high to low risks from exposure to EC vapour, whatever those risks may be.\nAn important principle in society is that individuals should have the freedom to make their own decisions about what they do, bearing in mind the risks involved, as they are able to perceive them. However, given the lack of evidence about the safety of alternatives to cigarette smoking, how can members of the general public be expected to make their own decisions about what, if anything, to smoke, and with what limitations?\nThe article refers to a triangulation of evidence from literature reviews, consumer input and experts opinions, but we would argue that none of these are a satisfactory basis for the self-assessment of risk. Such as it is, the literature is sparse and conflicting, and both consumers and policy makers are strongly influenced by the media, where reporters are easily led by soundbites, such as the PHE “95% safer” slogan. In addition, finding experts without conflicts of interest is not easy. There is a continuum from individuals who lack sufficient expertise to those who have a great deal of expertise, but those at the latter end of the spectrum are likely to have conflicts of interest as a result of the ways in which their experience was gained. We note that the list of experts mentioned in relation to the generation of a conceptual framework did not include toxicologists. Having a sufficient basis for perceiving the potential toxic effects of tobacco and other nicotine products is surely as essential to ABOUTTM Tool as is the use of an appropriate psychometric model.\nThe development of systems such as ABOUTTM Toolbox raises many important questions, including how, and by whom, they are designed and managed, the results are analysed and reported, and their outcomes are used in making policy decisions. This must involve high degrees of independence and transparency, and there is also a continuum in terms of motivation. At one end are those responsible for advising on public health, while at the other end are those with commercial interests in developing and maintaining markets and dividends. These various motivations are defensible, and those with one kind of responsibility should recognise the responsibilities of others placed elsewhere in the continuum.\nThis leads us to other matters of particular concern. First, the move from cigarettes smoking to the use of ECs should only be a stage in a longer process, leading to the avoidance of all forms of nicotine, however, consumed, in order to protect the delicate tissues of the respiratory system, in particular, from avoidable harm resulting from exposure to toxic and addictive substances. There is growing evidence that that is not a scenario that the tobacco industry wants, as the industry appears to be aiming to replace one kind of profitable dependence on addiction to nicotine with another one, on the grounds that it is less harmful. Secondly, tobacco harm results from repeated and long-term exposure to tobacco smoke, but there is no available evidence concerning the long-term effects of repeated exposure to EC vapours. Here, then, the precautionary principle should apply, and, ideally, the health of cohorts of EC users and non-users should be monitored over several years, before any statements about safety are made. Thirdly, we are all taken in by the tactic of pricing of goods at $10.99, which seems lower than $11.00. Similarly, we tend to think that, if something is 95% safer than something else, it is absolutely safe. That is not true for ECs. This tendency may be useful to the short-term ambitions of PHE and to the commercial hopes of the producers of ECs, but cannot possibly be a satisfactory and acceptable basis for a public health policy. Fourthly, there must be a clear distinction between policies aimed and assisting cigarette smokers to switch to something less harmful and those aimed at individuals who have never smoked at all. Regrettably, there are campaigns aimed at attracting young people, including children, to use ECs, and there is evidence that individuals who begin to take in nicotine via ECs, later switch to cigarettes.\n\nWe hope that the authors may wish to consider our points and perhaps make some minor modifications to their article.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? No source data required\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1878
https://f1000research.com/articles/7-1277/v1	14 Aug 18	{ "type": "Research Article", "title": "Neuromodulation influences synchronization and intrinsic read-out", "authors": [ "Gabriele Scheler" ], "abstract": "Background: The roles of neuromodulation in a neural network, such as in a cortical microcolumn, are still incompletely understood. Methods: (a) Neuromodulation influences neural processing by presynaptic and postsynaptic regulation of synaptic efficacy. Synaptic efficacy modulation can be an effective way to rapidly alter network density and topology. We show that altering network topology and density, will affect spike synchronization. Fast synaptic efficacy modulation may therefore influence the amount of correlated spiking in a network. (b) Neuromodulation also affects ion channel regulation for intrinsic excitability, which alters the neuron’s activation function. Results: We show that synchronization in a network influences the read-out of these intrinsic properties. Highly synchronous input drives neurons, such that differences in intrinsic properties disappear, while asynchronous input lets intrinsic properties determine output behavior. Thus, altering network topology can alter the balance between intrinsically vs. synaptically driven network activity. Conclusion: We conclude that neuromodulation may allow a network to shift between a more synchronized transmission mode and a more asynchronous intrinsic read-out mode. This has significant implications for our understanding of the flexibility of cortical computations.", "keywords": [ "neuromodulation", "synchronization", "network topology", "synaptic efficacy", "intrinsic excitability", "asynchronous", "activation function" ], "content": "Introduction\n\nIn this paper we present a realistic network model, akin to a cortical microcolumn1–4, and investigate its properties under the assumption of fast synaptic and intrinsic modulation as evidenced by neuromodulation5. We hypothesize that rapid synaptic efficacy changes allow a network to operate with different topologies, and that network topology is a decisive factor towards creating and sustaining synchronized inputs vs. producing asynchronous input.\n\nWe have previously shown for a conductance-based neural model of striatal medium spiny neurons that neuronal heterogeneity expressed by the contribution of individual ion channels (such as delayed rectifier potassium channels or GIRK channels) may still result in uniform responses, if the neurons are driven with highly correlated synaptic input. If the same neurons are driven by more asynchronous, distributed synaptic input, the heterogeneity is manifest in the response patterns, i.e. the spike rates and the timing of the spikes (see 6). These results were achieved using conductance-based point neurons6. Here we use two-dimensional neural models7 to further investigate the effect and determine its significance in the context of a cortical neural network.\n\nDue to Hebbian learning8,9, under normal conditions synaptic weights follow a lognormal distribution, which results in graphs with a heavy tail degree distribution. Degree modification by rapid synaptic efficacy changes would not only allow for alterations to the density, but also the topology of the connecting graph. In this paper we examine the hypothesis that such changes in network topology actually occur, driven by neuromodulatory effects on presynaptic release or postsynaptic response5,10–12. We analyze this situation with two example graphs, and we also perform further analysis to show that there is a continuum of graphs which can be reached by rapid synaptic changes.\n\n\nMethods\n\nThe conductance-based neural model of a striatal medium spiny neuron is described in detail in 6. The membrane voltage Vm is modeled using the equation\n\n\n\nwhere the Ii are the currents, induced by the individual ion channels. Variability of the neuron is modeled by modifications to µi. This model includes ion channels for Na (INa), K (IK), slow A-type K channels (IAs), fast A-type K channels (IAf), inward rectifying K channels (IKir), L-type calcium channels (ICaL), and the leak current (Ileak). The definition of all parameters and the dynamics of the ion channels can be found in 6.\n\nFor the experiments in this paper, we only focus on variability induced by changes in the strength of the slow A-type K channels. The total current contribution for this channel is µIAs where µ was selected between 1.0 and 1.5.\n\nIn order to illustrate the variability in neuron behavior, we excited the neuron model by input signals, resembling two kinds of synaptic input: uncorrelated and correlated. These signals were generated by superposition of excitatory and inhibitory spikes from individual Poisson-distributed spike trains (50 excitatory and 10 inhibitory), and biased Gaussian background noise.\n\nThe amount of pairwise correlation in these spike trains governs the type of input signal. A high correlation factor was used in order to generate sequences which have short periods (10–15ms) of high activity.\n\nIn order to do large-scale simulation we needed to employ a simple, computationally tractable neuron model. We used a two-dimensional model of a neural oscillator (cf. 7), and employed an instantiation of the model with parameters fitted to the general properties of cortical pyramidal neurons13 as a generic model (g). The model consists of an equation for the membrane model v (Equation 2), fitted to experimental values for cortical pyramidal neurons, and an equation for a gating parameter u (Equation 3).\n\nv˙ = 0.04v2 + 5v + 140 − u − Isyn (2)\n\nu˙ = a(bv − u)\n\nb = 0.2 (3)\n\na = 0.02\n\nWhen the neuron fires a spike (defined as v(t) = 30mV), v is set back to a low membrane potential v := c; c = −65.8mV and the gating variable u is increased by a fixed amount d (u := u + d; d = 8) (cf. 13). This formulation allows for a very simple neuron model, which avoids the explicit modeling of the downslope of the action potential, and rather resets the voltage. Time-dependence is modeled by the gating variable u.\n\nNeuronal heterogeneity is achieved by systematic variation of inactivation parameters. By varying d, we can vary the inactivation dynamics of the model after a spike, by varying a we vary the inactivation dynamics throughout the computation. In this way, we can attempt to model neuronal variability in activation/inactivation dynamics, which is sufficient to model frequency-selectivity as a stored intrinsic property. The parameters used in this paper for different neuron types are listed in Table 1.\n\nWe created graphs of N (= 210) excitatory neurons, and K (∼ 1900) excitatory connections. For the excitatory neurons, we use randomly connected graphs (N,K) with different dispersion σ. This corresponds to normal (Gaussian) to lognormal graphs with different widths and length of the heavy tail. We model inhibition by Poisson-distributed inhibitory synaptic input directly onto excitatory neurons.\n\nWe use specific instantiations of these graphs (RG, LG1) for the simulations. Table 2 shows global graph characteristics for the Gaussian graph (RG), the lognormal graphs (LG1), and intermediate graphs LG2, LG3, and LG4.\n\nWe define synchronization s in a network by pairwise correlations: for each neuron ni , we count, for each other neuron nj, the number of spikes which occur within a window W (W = 10ms) of ni’s spike events, divided by the total number of spikes for ni. More precisely, for each neuron, we bin all firing events into 5ms bins. We then count the number of other neurons, which fire in a 10ms window around the (start) of the bin. The synchronization s is then the average over all neuron pairs in the network:\n\n\n\nwhere Bij is the number of spikes that neurons i and j have in common within a moving window of W = 10ms during the entire measuring time. Sj is the number of spikes of neuron j during the entire measuring time.\n\nThe rewiring algorithm used to change the properties of a graph G is a greedy algorithm, which iteratively selects the node s with the highest degree. One of its edges is then rewired to random nodes with lower degrees, decreasing σ. The algorithm terminates, when the value of σ* falls below a given threshold.\n\nAll simulations were performed with the software tool CNeuroSim, which is implemented in Matlab (R2016b) and C, and available at 14.\n\n\nResults\n\nWe show how we can model gain as a stored intrinsic property, defined as a spike rate in response to constant or fluctuating input of fixed strength (A/Hz). We use a full ion channel based model (the MSN model6), with variation in the slow A-type potassium channel (IAs).\n\nIn Figure 1A, we show the response of MSN model neurons with a scaling of µIAs = 1.0, 1.3, 1.5 to a noisy signal, derived from uncorrelated Poisson-distributed synaptic input. The top panel shows the development of the membrane potential, Vm, over time for all neurons. The middle panel shows the spike-train for each neuron with the mean ISI and its standard deviation; the total number of spikes is shown on the right. The total number of spikes includes bursts, which were excluded from ISI calculation. The bottom panel shows the synaptic input. The dots correspond to the spiking events for a single neuron 3 (µIAs = 1.5). The resulting mean ISIs are 25, 37, and 45 ms. With a standard deviation of 6, 11, and 8, they are clearly distinguishable. This is also shown by the Gaussian distribution for the mean ISIs for each neuron type (Figure 1B).\n\nA. Frequency response of 3 conductance-based MSN model neurons with variable scaling of IAs to uncorrelated input B. Probability distributions of ISIs. We see a clear separation of frequency responses.\n\nThis model shows frequency-specificity as read-out of the relative contribution of the slow A-type potassium channel, indicated by the scaling factor µIAs. The relative contribution of an ion channel corresponds to its density or distribution on the somato-dendritic membrane, or in some cases its specific localization at dendritic branch points. Experimental evidence has shown that this is a plastic feature for neurons.\n\nWe then employ highly correlated synaptic input, defined as in 6 (see Methods). We stimulate the same neurons with the correlated input and observe the spike pattern (Figure 2). We can show that the frequency-specificity of the neuron disappears. Instead we see a time-locked spike pattern which is expressed by a similar spike frequency (Figure 2A) and an overlap of the mean ISIs (Figure 2B). What this experiment showed is that a stored intrinsic property, the gain, is available to the processing network in a conditional manner. The property is continually expressed, the differences in ion channel density persist. Depending on the mode of stimulation, however, this property is manifested as intrinsic gain, or it is obscured when a neuron is driven by strongly correlated input.\n\nA. Frequency response of the same MSN model neurons as in Figure 1 to correlated input. B. Probability distribution of ISIs. We see overlapping of frequency responses.\n\nTo continue with exploring this property of model neurons in a networked context, we switched to a simplified model neuron7 and created a set of variations for this model (see Methods). We show the response of two-dimensional model neurons to asynchronous input in Figure 3, and to regular, synchronous input in Figure 4. In the first case, we have clearly separated frequencies, and in the second case, the ISIs are nearly identical with a narrow distribution. When we stimulate the neurons with irregular, but synchronous input, the ISIs become identical, but with a wider distribution to reflect the different duration of pauses between the synchronous stimulation (Figure 5).\n\nA. Spike response of the two-dimensional dynamic model neurons (1,2,3) to asynchronous input. B. Gaussian distributions of ISIs for six model neurons (1,2,3,4,5,6) to asynchronous input as in A. We see a clear separation of frequency responses for model neurons.\n\nA. Spike response of the two-dimensional dynamic model neurons (1,2,3) to regularly timed, correlated input. B. Distributions of ISIs for six model neurons (1,2,3,4,5,6) to the same input as in A. We see strong overlapping of frequency responses, at about 50ms ISI, in accordance with the input. We notice that neuron 6 fires at lower frequencies than the input, it probably has a longer reset period, as seen in Figure 3B.\n\nA. Spike response of the two-dimensional dynamic model neurons (1,2,3) to irregularly timed, correlated input. B. Gaussian distributions of ISIs for six model neurons (1,2,3,4,5,6). We see strong overlap of frequency responses.\n\nWe may also consider the question of whether a neuron can simultaneously respond to an input and read out its stored spike frequency. If there are single synchronous events, which interrupt ongoing spiking, can we recover the intrinsic properties for each neuron? In Figure 6 it is shown that this is possible. Figure 6A shows the input and the synchronous responses, and in Figure 6B we still see a clear separation of frequencies.\n\nA. Multiplexed input and response of different model neurons (1–6). Three synchronous events are clearly represented in the spike pattern of all neurons. B. Mean ISIs for each neuron type. The separation of frequencies is kept. Compared to Figure 3B the standard deviation is somewhat higher because of the additional spikes caused by strong synchronous input.\n\nWe conclude that we can multiplex asynchronous and synchronous input. It is also apparent that there needs to be a lower limit on the intervals between synchronous events that can be processed without disrupting intrinsic properties. This interval needs to be defined as functionally dependent on the intrinsic frequencies. In this case, it is 3/s for the synchronous events, with 10Hz for the slowest neuron.\n\nThe simplified model neurons allow the creation of large networks of heterogeneous neurons and exploration of different topologies. We hypothesized that a lognormal graph, because of its hierarchical topology and the existence of hub neurons would lead to synchronization of action potentials – even with heterogeneous neurons – while a Gaussian topology would support asynchronous spiking behavior15,16. We define synchronization s in a network by pairwise correlation (Methods). The spike frequency for each neuron type is assessed by the mean and standard deviation for ISIs, as before.\n\nWe first use a Gaussian connected graph (RG) with N = 210 and K = 1800 and use 7 different neuronal types (1–6, plus the generic neuron g) with 30 Neurons each (Methods). Figure 7A shows an excerpt of the graph structure. We can see that the graph is connected such that all neurons have a comparable number of connections. This is also apparent in Figure 8, where we can see a (narrow) normal distribution for connectivity for the Gaussian graph RG. Table 2 contains the graph characteristics for both graphs.\n\nA. Part of a Gaussian Graph (RG), here for 50 neurons B. Part of a Lognormal Graph (LG1), for 50 neurons. The more regular, lattice-like structure of the Gaussian graph and the higher clustering and the appearance of highly connected ’hub’ neurons in the lognormal graph is apparent.\n\nThe LG has more neurons with few connections. It also has a heavy tail of neurons with 20 and more connections (’hubs’), which are lacking in the Gaussian graph.\n\nN = 210 is about the size of a minicolumn or ensemble unit within a larger network with presumably dense interconnections17. The maximal density d = K/(N×N−1) in a cortical microcircuit is estimated at 0.1 for 104 neurons, and 107 synaptic connections,18. With ˜50% of synapses internal to the network, d = 0.04 − 0.07 is a realistic value for internal connectivity17. There is also a small background inhibition to all neurons present, implemented by 10% inhibitory neurons with Poisson-distributed firing and complete connectivity to excitatory neurons.\n\nWe now stimulate the graph by an initial stimulation to 10 excitatory neurons (for about 1s). In Figure 9A, we see highly asynchronous neuronal activity after 1s of stimulation. The pairwise correlation values s is low (s = 0.11). Figure 9B shows that each neuronal type retains its own frequency, i.e., has its own typical ISI, separated from other neuronal types. We also notice that some neurons fire with low frequencies (5Hz) and others with higher frequencies (20Hz). Very low firing neurons (2Hz) which are typical for cortex are not represented in this model.\n\nA. Asynchronous behavior in the Gaussian graph with variable neuron types. Groups of neuronal types are apparent in the rasterplot. Some structure is probably due to background inhibition. B. Average spikes/s with probability distributions for all neuronal types, with clear separation by frequency. Pairwise correlation is s = 0.11.\n\nNext we changed the topology of the network to a lognormal graph (LG1), with N = 210 and K = 1924 and used the same neurons as before.\n\nFigure 7B shows an excerpt of a lognormal graph structure. The connectivity structure seems much denser, because of ’hub’ neurons in the center of the graph. In Figure 8, we can see a much wider distribution of degrees for the lognormal graph (blue), with a number of nodes with high connectivity. Presumably, those nodes are capable of synchronizing the network, because they can reach many neurons simultaneously. What is the effect on the presence of neural heterogeneity?\n\nFigure 10 shows that a high amount of synchronization can be achieved in spite of heterogeneity of intrinsic frequency of model neurons. The rasterplot (Figure 10A) shows the activity in LG1 with the same neurons and the same stimulation as before. The overall correlation, defined by pairwise correlation of neurons, is much higher (s=0.32). The distribution of ISIs in this case is strongly overlapping (Figure 10B), similar to Figure 4B, where neurons were explicitly driven by highly synchronous input. This means that synchronization is dependent on the network topology, and a lognormal graph exhibits a higher tendency for pairwise synchronization. Also, that neuronal heterogeneity is apparent in an asynchronous network mode but is repressed in a synchronous firing mode.\n\nA. Synchronization in a heavy-tail graph (LG1) with variable neuron types. The rasterplot shows that different neuronal types respond uniformly. B. Frequency distributions. High overlap between neuronal types is apparent. Pairwise correlation in the graph is high with s = 0.32.\n\nWe could show that differences in intrinsic properties appear or become more prominent when there is less synchronicity in a network. In our model, the pairwise synchronicity s is dominated by the network topology, more precisely by the width of the degree distribution (dispersion) ranging from Gaussian to lognormal with a heavy tail.\n\nTo confirm this observation we used a number of intermediate graphs and mapped the pairwise synchronization dependent on the dispersion σ* (Figure 11). The graphs RG and LG1 that we used have values of σ* = 1.44 and σ* = 2.89 (Methods). They have the same density, i.e., the same number of connections and neurons (0˜.05). Additionally, we analysed the dependence of synchronicity on the density of the graph between 0.01 and 0.1 (Figure 11).\n\nThe experimentally attested dispersion for weights in cortical tissue8 is σ* = 2 − 3.5, with a mean at 3. We achieve higher synchronization in the lognormal region, also dependent on density, but no synchronization in the Gaussian (low dispersion region) (σ* < 1.5), except close to maximal connectivity. Black dots signify actual measurements. There seem to be no abrupt transitions.\n\nThere is higher synchronization in the lognormal region, especially with higher σ* > 2.5 but no synchronization for Gaussian graphs. For heavy-tail graphs, synchronization depends linearly on the density between d = 0.03 − 0.08 (s = 0.2 − 0.5).\n\nHow are the different graphs related? We hypothesized that fast synaptic switching19 by neuromodulation could change the network topology sufficiently to switch from a synchronous to an asynchronous regime. In Figure 12, we plot the number of edges that were changed to achieve different dispersions σ* of a graph. The algorithm used was a simple greedy algorithm (Methods), which is suboptimal, i.e. overestimates the number of edges required. It appears that 30–50% of edges changed would be sufficient.\n\n\nDiscussion\n\nWe employ a parameterizable two-dimensional neural oscillator model to encode different intrinsic excitability manifested by different frequency responses to constant input. What the experiments show is that a stored intrinsic property, the gain, is available to the processing network in a conditional manner: The gain is continually present, the differences in ion channel density persist. Depending on the mode of stimulation, however, this property is manifested as spike rate, or it is obscured when a neuron is driven by strongly correlated input. This is interesting because it shows a property of memory that synaptic plasticity lacks: the memory is not always ’read-out’ in any processing step. It is conditional, it can be accessed or ignored depending on the state of the network. This seems to be an essential property of memory in any intelligent system.\n\nDifferent statistical properties of synaptic input can be modeled by a variability in the correlation properties of input neurons. In a network model, this means that the overall correlation in the network determines what input a neuron receives. With a Gaussian degree distribution topology, correlation is low and neurons fire irregularly with their own preferred frequency. With a heavy-tailed, lognormal topology, correlation is higher, and neurons fire when they receive correlated input, irrespective of intrinsic properties. I.e. driving neurons by correlated vs. uncorrelated input leads to uniform spiking behavior vs. read-out of stored differences in ion channel conductances.\n\nA restriction of the present model with respect to a biological simulation model is the simplified treatment of inhibition. However, experimental work shows that cortical parvalbumin-expressing (PV+), fast-spiking interneurons have no connection specificity to pyramidal neurons, rather they present as an ’unspecific, densely homogeneous matrix covering all nearby pyramidal cells’ (20, p. 13260), which corresponds to our model.\n\nConditions for neuronal read-out may include the activity of inhibitory neurons. Inhibition and excitation are tightly linked by feedback interaction. 21 suggested that the close coupling of inhibition and excitation in cortical tissue cancels out purely input-dependent, i.e. not network generated synchrony. 13 suggested that with highly correlated input, both inhibitory and excitatory, the neuron may receive less input which allows it to be driven only by strong synaptic input, with distributed input it may receive a barrage of excitatory and inhibitory inputs where the membrane voltage remains close to firing threshold and the neuron fires continuously. In our sense, it is reading out stored properties. Both inhibitory and excitatory synaptic input will conform to be asynchronous or synchronous, to drive neurons or it causes a neuron to emit spikes according to its own stored intrinsic properties.\n\nHowever, neuromodulation has effects on inhibitory neurons as well22,23, which we have not modeled. Further simulations will show whether the I-E coupling is altered during enhanced neuromodulation, or whether the effects are synergistic to the present results.\n\nNeuromodulation influences both intrinsic properties and synaptic connectivity5, e.g. acetylcholine, (via nucleus basalis stimulation), noradrenaline (via LC stimulation) or dopamine (via VTA stimulation)5,24. Experimental estimates on the distribution of synaptic neuromodulatory receptors are at approximately 30%–50% of connections19. That is sufficient to transform the topological properties of a graph, such as the dispersion of its degree distribution from heavy-tailed graph to a more Gaussian, less clustered graph without requiring tight optimization for the positions of neuromodulatory receptors (Figure 12).\n\nNeuromodulation disables or enhances various ion channels, such as Sk-channels which guide reset times after a spike, or A-type potassium channels which influence latency to spike6. In this way, neuromodulation influences intrinsic properties. If neuromodulation reduces synchrony by acting at synaptic receptors, it uncovers intrinsic heterogeneity, and induces a mode of processing that allows read-out and storing of intrinsic properties. Depending on the neuromodulator used, and the amplitude and duration of the signal, different soma-dendritic ion channel profiles would emerge.\n\nThe algorithm is not optimized (Methods) and overestimates the number of edges that have to be changed.\n\nIn the synchronous mode, intrinsic heterogeneities are reduced in the presence of tightly correlated input which drives neurons reliably. This invariance of neuronal intrinsic properties in synchronous mode allows synaptic transmission and information processing independent of neuronal heterogeneity.\n\nThe idea of introducing synchronous events by common input to an asynchronous background, and in this way use reliable synaptic transmission without affecting the state of the system (multiplexing) has also been documented in experimental results. For instance, (25, Figure 4A) shows a case of multiplexing in response to behavioral stimuli. In this case, intrinsic read-out can continue, and single events are transmitted reliably through driven activations.\n\nWhy should synchronization properties be switched by neuromodulation? Increased correlation in the network supports population-coded information to be propagated effectively26. Turning on neuromodulation would decorrelate an area and increase the capacity for information coding in an ensemble or a cortical microcolumn27. This area would become an information source to surrounding areas. When turned off, increased correlation would allow this area to transmit information and to disregard the stored neural memory.\n\nBasal forebrain stimulation, which results in increased acetylcholine release and muscarinic/nicotinic receptor activation, decreases correlation between cortical neurons (28–30 (Figure 3.C)). Likewise, 31 (Figure 3 and Figure 4) shows reduced noise (internal) correlations with cholinergic stimulation, while inactivation of the basal forebrain caused more synchronized activity. 32 shows reduction of correlation for task-relevant perception, where presumably task-relevance causes neuromodulatory activity. 33 provides evidence for the involvement of noradrenaline in desynchronization of cortical state and the enhancement of sensory coding.\n\nThere is considerable evidence34–37 showing that several neuromodulators, including at least noradrenaline and acetylcholine, modulate pairwise spike correlation, such that strongly synchronized states (anesthesia, slow wave sleep) have high correlation and low neuromodulation, while asynchronous states (normal waking), with higher neuromodulation, have lower pairwise correlation.\n\n37 observed intrinsic fluctuations in synchronization of cortical networks during wakefulness which correlated with the amount of encoded perceptual information and perceptual performance. Their results showed a mean decrease in correlations from synchronized to desynchronized state corresponding to perceptual performance by approximately 20%, similar to values observed during attention38, and after adaptation39. We have shown (Figure 11) that correlation changes are continuous with network topology and a 20% correlation change is well within the range of the current simulations. Importantly, the results in 37 point to fluctuations in synchronization that reflect local changes in network activity rather than just global cortical state dynamics which have traditionally been associated with central neuromodulatory release.\n\nThe role of presynaptic neuromodulation in suppressing cortical connections11,12 and changing attractor states40, as well as allowing rapid synaptic weight changes19 has previously been assessed. Theoretical work has also emphasized the connection between correlations and information content27,41–43.\n\nHere we bring these observations together to suggest that neuromodulation of synapses may alter network topology and in this way bring about an increased decorrelation of spiking, and a more asynchronous state, with a higher informational capacity. It may provide a general explanation (a) on how fluctuations in synchrony can be engineered rapidly and in small cortical areas and (b) why intrinsic memory may be conditional, accessible only at certain times and in a localized fashion.\n\n\nConclusion\n\nWe created a number of different parameterized neuron models to capture neuronal heterogeneity. This affects the properties of the neuron such that it has less or more intrinsic excitability, leading to different firing rates when stimulated in an asynchronous way. Under synchronous stimulation the differences are greatly reduced.\n\nWe also suggested that synaptic neuromodulation can be an effective way of rapidly altering network topology. We investigated changes in network topology along the dimensions of Gaussian vs. heavy-tailed degree distributions. We hypothesized that heavy-tailed graphs produce more globally synchronized behavior than comparable Gaussian graphs. In accordance with the hypothesis, we find that in a heavy-tailed graph, because of high population synchrony, the difference between neuronal intrinsic properties is minimized, while a Gaussian graph allows read-out of neuronal intrinsic properties. Thus, altering network topology can alter the balance between intrinsically determined vs. synaptically driven network activity.\n\n\nData availability\n\nUnderlying data for this study is available from Zenodo. Dataset 1. gscheler/CNeuroSyn: initial version, https://doi.org/10.5281/zenodo.116409614.\n\nData is available under a Creative Commons CC BY-NC 4.0 license.\n\n\nSoftware availability\n\nThe source code for the model is available from GitHub: https://github.com/gscheler/CNeuroSyn/tree/V1.0/src/analysis\n\nArchived source code at time of publication is available from Zenodo https://doi.org/10.5281/zenodo.116409614.\n\nSoftware is available under GNU GPL v2.0 license.", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nDouglas RJ, Martin KA: Neuronal circuits of the neocortex. Annu Rev Neurosci. 2004; 27(1): 419–451. PubMed Abstract \| Publisher Full Text\n\nDouglas RJ, Martin KA: Recurrent neuronal circuits in the neocortex. Curr Biol. 2007; 17(13): R496–500. PubMed Abstract \| Publisher Full Text\n\nCarlo CN, Stevens CF: Structural uniformity of neocortex, revisited. Proc Natl Acad Sci U S A. 2013; 110(4): 1488–1493. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKunze T, Peterson ADH, Haueisen J, et al.: A model of individualized canonical microcircuits supporting cognitive operations. PLoS One. 2017; 12(12): e0188003. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nScheler G: Regulation of neuromodulator receptor efficacy--implications for whole-neuron and synaptic plasticity. Prog Neurobiol. 2004; 72(6): 399–415. PubMed Abstract \| Publisher Full Text\n\nScheler G: Learning intrinsic excitability in medium spiny neurons [version 2; referees: 2 approved]. F1000Res. 2014; 2: 88. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIzhikevich EM: Which model to use for cortical spiking neurons? IEEE Trans Neural Netw. 2004; 15(5): 1063–1070. PubMed Abstract \| Publisher Full Text\n\nScheler G: Logarithmic distributions prove that intrinsic learning is Hebbian [version 2; referees: 2 approved]. F1000Res. 2017; 6: 1222. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKoulakov AA, Hromádka T, Zador AM: Correlated connectivity and the distribution of firing rates in the neocortex. J Neurosci. 2009; 29(12): 3685–94. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVarela JA, Hirsch SJ, Chapman D, et al.: D1/D5 modulation of synaptic NMDA receptor currents. J Neurosci. 2009; 29(10): 3109–19. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOhshima M, Itami C, Kimura F: The α2A -adrenoceptor suppresses excitatory synaptic transmission to both excitatory and inhibitory neurons in layer 4 barrel cortex. J Physiol. 2017; 595(22): 6923–6937. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKobayashi M, Kojima M, Koyanagi Y, et al.: Presynaptic and postsynaptic modulation of glutamatergic synaptic transmission by activation of alpha1- and beta-adrenoceptors in layer V pyramidal neurons of rat cerebral cortex. Synapse. 2009; 63(4): 269–281. PubMed Abstract \| Publisher Full Text\n\nIzhikevich EM, Gally JA, Edelman GM: Spike-timing dynamics of neuronal groups. Cereb Cortex. 2004; 14(8): 933–944. PubMed Abstract \| Publisher Full Text\n\nScheler G: gscheler/CNeuroSyn: initial version. Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1164096\n\nBarrat A, Barthélemy M, Vespignani A: Dynamical Processes on Complex Networks. Cambridge University Press, 2008. Publisher Full Text\n\nArenas A, Diaz-Guilera A, Kurths J: Synchronization in complex networks. Physics Reports. 2008; 469(3): 93–153. Publisher Full Text\n\nKlinshov VV, Teramae JN, Nekorkin VI, et al.: Dense neuron clustering explains connectivity statistics in cortical microcircuits. PLoS One. 2014; 9(4): e94292. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLansner A, Diesmann M: Virtues, Pitfalls, and Methodology of Neuronal Network Modeling and Simulations on Supercomputers. Springer Netherlands, Dordrecht, 2012; 283–315. Publisher Full Text\n\nScheler G, Schumann J: Presynaptic modulation as fast synaptic switching: State-dependent modulation of task performance. In Proceedings of the International Joint Conference on Neural Networks. 2003; 1. Publisher Full Text\n\nPacker AM, Yuste R: Dense, unspecific connectivity of neocortical parvalbumin-positive interneurons: a canonical microcircuit for inhibition? J Neurosci. 2011; 31(37): 13260–13271. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGraupner M, Reyes AD: Synaptic input correlations leading to membrane potential decorrelation of spontaneous activity in cortex. J Neurosci. 2013; 33(38): 15075–15085. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKruglikov I, Rudy B: Perisomatic GABA release and thalamocortical integration onto neocortical excitatory cells are regulated by neuromodulators. Neuron. 2008; 58(6): 911–924. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStringer C, Pachitariu M, Steinmetz NA, et al.: Inhibitory control of correlated intrinsic variability in cortical networks. eLife. 2016; 5: pii: e19695. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGuedj C, Monfardini E, Reynaud AJ, et al.: Boosting Norepinephrine Transmission Triggers Flexible Reconfiguration of Brain Networks at Rest. Cereb Cortex. 2017; 27(10): 4691–4700. PubMed Abstract \| Publisher Full Text\n\nGutnisky DA, Beaman C, Lew SE, et al.: Cortical response states for enhanced sensory discrimination. eLife. 2017; 6: pii: e29226. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZylberberg J, Pouget A, Latham PE, et al.: Robust information propagation through noisy neural circuits. PLoS Comput Biol. 2017; 13(4): e1005497. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSompolinsky H, Yoon H, Kang K, et al.: Population coding in neuronal systems with correlated noise. Phys Rev E Stat Nonlin Soft Matter Phys. 2001; 64(5 Pt 1): 051904. PubMed Abstract \| Publisher Full Text\n\nGoard M, Dan Y: Basal forebrain activation enhances cortical coding of natural scenes. Nat Neurosci. 2009; 12(11): 1444–1449. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEdeline JM: Beyond traditional approaches to understanding the functional role of neuromodulators in sensory cortices. Front Behav Neurosci. 2012; 6: 45. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLee SH, Dan Y: Neuromodulation of brain states. Neuron. 2012; 76(1): 209–222. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMinces V, Pinto L, Dan Y, et al.: Cholinergic shaping of neural correlations. Proc Natl Acad Sci U S A. 2017; 114(22): 5725–5730. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJeanne JM, Sharpee TO, Gentner TQ: Associative learning enhances population coding by inverting interneuronal correlation patterns. Neuron. 2013; 78(2): 352–363. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFazlali Z, Ranjbar-Slamloo Y, Adibi M, et al.: Correlation between Cortical State and Locus Coeruleus Activity: Implications for Sensory Coding in Rat Barrel Cortex. Front Neural Circuits. 2016; 10: 14. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHarris KD, Thiele A: Cortical state and attention. Nat Rev Neurosci. 2011; 12(9): 509–523. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchölvinck ML, Saleem AB, Benucci A, et al.: Cortical state determines global variability and correlations in visual cortex. J Neurosci. 2015; 35(1): 170–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRenart A, de la Rocha J, Bartho P, et al.: The asynchronous state in cortical circuits. Science. 2010; 327(5965): 587–590. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBeaman CB, Eagleman SL, Dragoi V: Sensory coding accuracy and perceptual performance are improved during the desynchronized cortical state. Nat Commun. 2017; 8(1): 1308. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMitchell JF, Sundberg KA, Reynolds JH: Spatial attention decorrelates intrinsic activity fluctuations in macaque area V4. Neuron. 2009; 63(6): 879–888. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGutnisky DA, Dragoi V: Adaptive coding of visual information in neural populations. Nature. 2008; 452(7184): 220–4. PubMed Abstract \| Publisher Full Text\n\nKanamaru T, Fujii H, Aihara K: Deformation of attractor landscape via cholinergic presynaptic modulations: a computational study using a phase neuron model. PLoS One. 2013; 8(1): e53854. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCohen MR, Kohn A: Measuring and interpreting neuronal correlations. Nat Neurosci. 2011; 14(7): 811–819. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCohen MR, Newsome WT: Context-dependent changes in functional circuitry in visual area MT. Neuron. 2008; 60(1): 162–173. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPoulet JF, Petersen CC: Internal brain state regulates membrane potential synchrony in barrel cortex of behaving mice. Nature. 2008; 454(7206): 881–885. PubMed Abstract \| Publisher Full Text" }	[ { "id": "37164", "date": "28 Aug 2018", "name": "Joel Zylberberg", "expertise": [ "Reviewer Expertise Computational neuroscience", "dynamical systems", "sensory systems", "memory", "learning" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis study asks when the intrinsic biophysical properties of neurons do or do not affect their spiking responses to synaptic inputs. I denote the main claims and results of the paper as I,II,and III below, each of which is discussed individually thereafter.\nI. The author begins by noting that, in the presence of strong synchronous inputs, essentially all neurons are driven to quickly spike, regardless of their intrinsic properties. In the presence of less-synchronous inputs, differences in intrinsic properties do lead to noticeable changes in spiking outputs. This is (presumably) because differences in integration time, and excitability (effective spiking threshold) determine how much input is needed to get the cell to spike, but sufficiently strong synchronous inputs always suffice to generate spikes. This is a neat observation, somewhat obvious in hindsight, but nevertheless made me glad to read the paper.\nII. Next, they investigate how different network structures (quantified by dispersion of the graph's out-degree distribution) affect the synchrony in the inputs to individual neurons: networks with high dispersion (e.g., small numbers of critical \"hubs\" in the network) lead to higher synchrony. Thus, in those networks, the neurons' intrinsic properties are less influential than in networks with lower dispersion.\nIII. Finally, the authors speculate that, because neuromodulators can change effective connectivity, they could switch the network between different graph structures, thereby either enabling, or disabling, the influences of neurons' intrinsic properties. In that way, the network can \"multiplex\" between operational modes, on of which (the asynchronous one) is more affected by neurons' intrinsic plasticity, and thus more affected by memory and experience.\nBelow, I summarize first the relation to prior studies for each claim (I-III), and then comment on some technical and presentation issues that, to me, hinder somewhat the readability of the paper. My hope is that addressing these concerns will help the author to increase the impact of their work.\nComments on relation to other studies:\nResult I. This is a neat observation. It is somewhat foreshadowed by some older work by Salinas and Sejnowski1,2, showing that synchronous inputs are good at driving spikes in basically any neuron model. It would be useful to differentiate the findings of this paper from their older work.\nResult II. Closely related results were obtained in a pair of studies by Yu Hu, James Trousdale, and colleagues3,4. There, they computed the correlation structure (closely related to the author's synchrony measure) in networks with different connectivity motifs, using a neat analytical approach involving motif cumulants. It would be worthwhile (again) differentiating the results here from that prior work, so that readers know what's new and significant about this paper vs. the prior state of the field.\nResult III. I'm puzzled by the claim that the network topology is strongly affected by neuromodulators (which is the key to the switching property that's discussed through the paper). While I understand that neuromodulators can strengthen or weaken synapses, that mechanism will change the weights within the graph describing network connectivity, but not change the underlying unweighted graph describing which neurons are connected to each other (regardless of the strength of that connection). To change the network topology, the neuromodulators would need to either a) add new synapses (or \"unsilence\" previously silenced synapses) selectively to some neurons, or b) remove (or completely silence) synapses selectively from some neurons. I think that, in order for the author to make claims about neuromodulators changing network topology, they should include some discussion (ideally based on the experimental literature) of when and how neuromodulators actually change network topology. This literature could of course exist: there's lots that I don't know about neuromodulators. And in that case, I think the author's claims would be much strengthened by referencing and discussing it.\nComments on presentation and specific technical issues:\nMethods, 3rd paragraph. Why only focus on modulators affecting slow A-type K channels? Are those the most important ones? The ones most affected by common neuromodulators? Which neuromodulators are those?\n\nMethods, 4th paragraph. It would help me if you would describe explicitly the method used to generate either correlated or uncorrelated synaptic input traces. For example, do these incoming spike trains (generating the synaptic inputs) come from a process like SIP or MIP? And specifically, how were the \"correlation factors\" quantified and manipulated?\n\nSimilar to (2) it would help to describe explicitly the synapse model used. For example, are the synapses conductance based (e.g., convolve incoming spike trains with a double exponential to get conductance traces)? I assume not, given that in Figs. 1-6, the same synaptic current trace is shown for all neurons. If the synapses were conductance-based, the differences in membrane potentials of those neurons would lead to differences in synaptic currents.\n\nThe definition of the synchronization is fine to me but doesn't seem to match the description in the preceding paragraph. Specifically, the formula doesn't seem to \"count the number of other neurons which fire in a 10ms window. . . \" so much as \"count the number of spikes emitted by each other neuron in a 10 ms window\".\n\nI'd describe the rewiring process in the \"graph properties\" section of the Methods and not the \"Definitions\" section. That way, when you introduce the dispersion \\sigma, you can explain the process used to manipulate it in your simulations.\n\nThe figures generally could use much more annotation, to help the readers parse what all the different curves mean. For example, in Figs. 1 and 2, I'd color code the curves (distributions) in panel B, and have those same colors used for the different spike trains in panel A, so that readers know which distribution goes with which spike train goes with which model. The same idea could be applied to the other figures: help the reader out so that, at a quick glance (which is how people read papers now) they can tell what's what and what the main point is.\nSimilarly, in Fig. 11, I'd label the color bar, so that people know what variable the colors are describing.\n\nThe formula for density on p. 6 (bottom left) seems incorrect, as d=K/(NxN-1). That would be K/(N^2 -1), whereas I think you mean K/[Nx(N-1)].\n\nP. 6, upper right, description of stimulating the graph by driving 10 excitatory neurons. How was this done? Did you inject current into those model neurons? If so, how much?\n\nSome consistency between the network sizes in the simulations would help a lot. Specificlaly, in Fig. 10, you study a 1924 neuron network, whereas Fig. 9 uses 1800 neurons. This leave the interpretation muddier than it needs to be: readers might wonder whether some of the differences could be attributed to changes in network size, and not just to changes in connectivity. In general, I'd suggest keeping all but one key variable constant, so that the comparisons can be made as cleanly as possible.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] }, { "id": "38207", "date": "24 Oct 2018", "name": "Guillaume Drion", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper compares (i) the response of heterogeneous neurons to two different types of synaptic inputs (uncorrelated or correlated), and (ii) the response of neuronal populations having two different types of synaptic weight distribution (Gaussian or Lognormal). The main results can be summarized as follows:\nThe chosen neuronal heterogeneity induces variability in mean spiking frequency. This variability persists under the application of uncorrelated synaptic inputs, but is masked by the application of correlated synaptic inputs. Stimulating a subpopulation of neurons in a neuronal population induces an asynchronous response when the distribution of synaptic weights is Gaussian, whereas it induces a synchronous response when this distribution is Lognormal.\nThen, it is speculated that switches between asynchronous and synchronous states can be controlled by neuromodulators, which would modify the network topology from Gaussian to Lognormal and conversely.\n\nAfter carefully reading this paper, I think that neither the content nor the form is in a sufficiently advanced stage for the manuscript to be indexed in its current form. A thorough revision is strongly suggested to boost the potential impact of the research. I provide specific comments below, in order of appearance in the text.\n\nTable and figure captions are not descriptive enough. One cannot fully comprehend the different figures, which include numerous notations, by reading the captions. Some notations are not consistent (in Fig. 1-6, ISI is noted I=*** on the left panels, and ISI on the right panels).\n\nTitle: The title is totally misleading. It states that “neuromodulation influences and intrinsic read-out”, but nothing of the sort is demonstrated in the manuscript. Neuromodulation is only speculated to be a plausible source to control the mechanisms highlighted in this manuscript, without any evidence or results to support this speculation. I have nothing against speculations and interpretations, quite the contrary, but they should not appear as grand truth in the title.\n\nAbstract: The “Methods” part of the abstract is solely composed of results. This part should summarize the methods used in this paper.\n\nMethods: “For the experiments in this paper, we only focus on variability induced by changes in the strength of the slow A-type K channels. The total current contribution for this channel is μIAs where μ was selected between 1.0 and 1.5.\" Please motivate this choice (both channel type and chosen range), because the model is composed of 6 voltage-gated channels that could be used to model variability. Also, I would not use “experiments” but “simulations”, although this is a minor comment.\n\n“Time-dependence is modeled by the gating variable u.” What does this sentence mean? Both ODEs model time-dependent evolution.\n\n“By varying d, we can vary the inactivation dynamics of the model after a spike, by varying a we vary the inactivation dynamics throughout the computation.” I do not see the why variables a and d correspond to inactivations. a is the time constant of activation/deactivation of u (u indeed activates/deactivates, the slope b being always positive), and d is the increase in u after a reset, which represents how much a spike activates u.\n\n“For the excitatory neurons, we use randomly connected graphs (N,K) with different dispersion sigma*. This corresponds to normal (Gaussian) to lognormal graphs with different widths and length of the heavy tail.”. This dispersion coefficient, and how it “corresponds to normal (Gaussian) to lognormal graphs with different widths and length of the heavy tail “, should be defined properly, because it is an important parameter in the results.\n\n“The rewiring algorithm used to change the properties of a graph G is a greedy algorithm, which iteratively selects the node s with the highest degree”. Does “s” refer to a node, or is it a measure of network synchronization?\n\nResults: “We show how we can model gain as a stored intrinsic property, defined as a spike rate in response to constant or fluctuating input of fixed strength (A/Hz)” This sentence is unclear. What kind of gain is it referring to? What is a “stored” intrinsic property? What does “fixed strength” mean for a fluctuating input?\n\n“In Figure 1A, we show the response of MSN model neurons with a scaling of μIAs = 1.0, 1.3, 1.5 to a noisy signal, derived from uncorrelated Poisson-distributed synaptic input.” From what I understood, neurons are not connected to each other in this case. It should be explicitly mentioned, because at this stage it is unclear whether we discuss isolated neurons or neurons within a network.\n\n“The total number of spikes includes bursts, which were excluded from ISI calculation.” How were bursts defined? Why were they excluded from ISI calculation? But then why excluding them from one measure but not from the other?\n\nResults related to Fig. 1 and 2: Only one example of uncorrelated/correlated input is shown, and conclusion is drawn from this example. This part requires further investigation. For instance, the increase in firing rate that we see in Fig. 2 could potentially induce overlapping of spiking rate, because it moves the neurons to a flatter portion of the f-I curve. It is important to rule out this effect to make sure that the difference is indeed due to the correlated/uncorrelated nature of the inputs, for instance by showing that increasing the firing rate with uncorrelated synaptic inputs does not induce the same effect. I also do not find surprising that inputs with high amplitude, low frequency fluctuations would affect the firing more than an almost constant input, but I guess that it is more subjective. You also talk about “a time-locked spike pattern”. How do you define it in the context of this paper? If is it synonymous of synchronization, having clearly separated frequencies does not mean neurons cannot be synchronized: one neuron could have a frequency that is twice the other and be fully synchronized, yet the frequencies would be separated. On the other hand, two neurons could fire at a same frequency and coefficient of variation, and yet not be synchronized.\n\n“To continue with exploring this property of model neurons in a networked context, we switched to a simplified model neuron7 and created a set of variations for this model (see Methods).” Again, I guess that the results of this section are provided on isolated neurons. If it is the case, this sentence is misleading.\n\nResults connected to Fig. 3, 4 and 5: My comment connects to the one relates to Fig. 1 and 2. The spikes of Fig. 3A look much more “time-locked” than the ones of Fig. 2A, although the conclusions that are drawn from these two figures are opposite. Is the “clear separation of ISIs” such a good measure in this context? In Fig. 4 and Fig. 5, most events look like bursts. Are bursts excluded from ISI calculation like in the previous case? If not why? Again, how is bursting defined in this case as compared to the previous case? Finally, in Fig. 5 all neurons seem to burst, yet the distribution of ISIs is a Gaussian, which does not make sense, unless you are only plotting the distribution for the interburst intervals.\n\n“We first use a Gaussian connected graph (RG) with N = 210 and K = 1800 and use 7 different neuronal types (1–6, plus the generic neuron g) with 30 Neurons each (Methods). Figure 7A shows an excerpt of the graph structure. We can see that the graph is connected such that all neurons have a comparable number of connections. This is also apparent in Figure 8, where we can see a (narrow) normal distribution for connectivity for the Gaussian graph RG. Table 2 contains the graph characteristics for both graphs.” Only one graph is mentioned in the paragraph, and Table 2 sows 5 different cases. Both graph probably mean “Gaussian” and “Lognormal”, but that should be clarified.\n\n“We now stimulate the graph by an initial stimulation to 10 excitatory neurons (for about 1s).” How were these 10 neurons selected? Where they selected randomly, or based on their degree of connectivity? Are they chosen to correspond to nodes with a high connectivity degree in the lognormal case? If so, then what is happening if we either stimulate neurons with lower connectivity degrees in the lognormal case, or add a few hub neurons in the Gaussian case? Is the difference in synchronization really due to the whole distribution of the synaptic weights, or simply due to the existence of hub neurons that would be specifically targeted by synaptic inputs? This question needs to be answered in order to support the conclusion “In our model, the pairwise synchronicity s is dominated by the network topology, more precisely by the width of the degree distribution (dispersion) ranging from Gaussian to lognormal with a heavy tail.” In Figure 11, are the “experimental conditions” similar than in Figures 9 and 10? Which 10 neurons are stimulated?\nDiscussion “Neuromodulation disables or enhances various ion channels, such as SK-channels which guide reset times after a spike, or A-type potassium channels which influence latency to spike.” References that support these claims are needed.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [] } ]	1	https://f1000research.com/articles/7-1277
https://f1000research.com/articles/7-1866/v1	29 Nov 18	{ "type": "Research Article", "title": "Attitudes of women in Cambodia towards child physical abuse", "authors": [ "Koustuv Dalal", "Reshma Parvin Nuri", "Ming Shinn Lee,", "Chao Kuang Lin", "Mervyn Gifford", "Gainel Ussatayeva", "Animesh Biswas", "Koustuv Dalal", "Reshma Parvin Nuri", "Ming Shinn Lee,", "Chao Kuang Lin", "Mervyn Gifford", "Gainel Ussatayeva" ], "abstract": "Background: This study attempted to explore the women’s attitude towards child physical abuse in relation to the respondent’s background factors, personal issues and autonomy. Methods: This was a cross-sectional study of 18,749 women of reproductive age (15-49 years) using 2010 Cambodia Demographic and Health Survey. Chi-square tests and bivariate analyses were performed. Results: A significant proportion of women supported beating physically abusing sons (69.2%) and daughters (67.2%). Rural, non-Buddhist, those with no or primary education, poverty, seasonal or occasional employment seem to be risk factor for supporting child physical abuse by women (in bivariate analysis). Age, education and household economic status of the women are significantly relevant for child physical abuse (in bivariate analysis). Women who came from male-headed households more often supported beating their children. Female autonomy is an important factor for child physical abuse. Women who justify physical abuse towards wives were also generally supportive of child physical abuse. Conclusions: The current study provides knowledge about maternal factors such as age, education, economic status, rural/urban dwelling, two or more lifetime partners and autonomy in the supporting of beating sons and daughters. Further attention needs to be paid to increasing women’s education and autonomy in Cambodian family life.", "keywords": [ "children", "abuse", "physical punishment", "autonomy", "mother", "Cambodia" ], "content": "Introduction\n\nChild abuse is a major public health problem1. Child abuse is now alarmingly widespread in East Asia and the Pacific region, and it has immediate, long lasting and devastating effects on children2. Among other forms of abuse, physical abuse is more common among Asian and Pacific families than families from other parts of the world3–5. The overall prevalence rates of physical abuse range from 10 to 30.3% in the East Asia region. Boys are more exposed of physical abuse than girls, which can include physical punishment and severe physical contact violence2. In Cambodia over half of all children have experienced physical abuse6.\n\nBoth abuse and neglect negatively affect the development of a child’s brain7. It is well established that physical abuse has negative impact on physical, psychological and social and behavioral health of children8. This can ultimately affect the child’s long-term health-related quality of life and lead to problems such as depression, substance abuse, anxiety and suicidal behavior, increased risk of sexually transmitted diseases, and even cancer9–12. Adolescents who have been physically abused are more likely to show increased of externalizing problems and criminal offending13,14. In addition, adolescents and adults who had experienced physical and/or sexual abuse in their childhood are four times more likely to have suicidal ideation, than their peers who had not experienced such abuses2. Furthermore, among adults who had been physically abused as children there is a higher degree of aggression compared with adults who had not been abused13,14. Physical punishment of children are very frequent in East Asia and the Pacific region including Cambodia2. A study of low- and middle-income countries indicated that more than half of children were subjected to some kind of violent physical punishment, as the adults believe that such physical abuses of the children are not harmful15. However, the experts have strongly argued that child physical punishment such as beating should not be believed to be a minor form of physical abuse, as it causes several physical and psychological health problems16.\n\nMany studies have explored the parental risk factors of child abuse. Such risk factors include poverty, low family income and socioeconomic status. Lower level of parental education, large families, younger parental age, substance abuse and mental illness in parents are also contributing factors of child abuse. Adults having been maltreated in their childhood, intimate partner violence, parents’ divorce, or violence between other family members are identified risk factors for child abuse1,4,13,17. Furthermore, low parental self-esteem, depression, psycho-pathology, and social isolation are also positively related to child physical abuse18,19. Moreover, a recent study suggested that men who abused their wives were also frequently abusive to their children20. Therefore, parents’ background and personal experiences can influence the likelihood of child abuse. However, we have no specific information how Cambodian parents are perpetrating, though child abuse prevalence is very high2,16. African studies and Indian study have indicated that autonomy is important factor for child abuse, while we have no information from Cambodia1,16,18. Therefore, in this context, considering the mother’s role within family and towards her children, the current study has focused on women’s background factors, personal issues and autonomy for household decision-making in relation to child physical abuse.\n\nIt has been narrated that mothers are twice more likely to psychologically or physically abuse their children than fathers21. Cambodian refugee mothers in America who suffered from depression frequently abused their children22. Previous research found that maternal stress has a direct role in the physical abuse of children23. Another study conducted among Korean immigrant mothers found that the amount of time spent with children, experience of corporal punishment as a child, children’s gender and age, family acculturation conflicts, mothers’ age, and length of time in US are the macro-level variables that affect Korean immigrant mothers’ attitudes toward child physical abuse24. Existing studies of abusing Cambodian children are mainly focused on immigrant families in high income countries4. To better understand the situation, a national representative study is highly warranted1. Cambodia has a very high prevalence of child physical abuse, including child punishment at home. UNICEF has strongly advocated for preventing child abuse at home16. Women are leading factors in household issues, especially in Southeast Asian countries, such as Cambodia. However, to our knowledge, no research has been conducted to reveal women’s attitude towards child abuse at home in Cambodia4. Therefore, the objective of this study was to explore women’s attitude towards child physical abuse in relation to the respondent’s background factors, personal issues and autonomy.\n\n\nMethods\n\nThe study employed 2010 Cambodia Demographic and Health Survey (CDHS-2010) data25. In total, 18,749 women of reproductive age (15–49 years) were interviewed between 23 July 2010 and 20 January 2011 in 19 national sampling domains throughout Cambodia (14 individual provinces: Banteay Mean Chey, Kampong Cham, Kampong Chhnang, Kampong Speu, Kampong Thom, Kandal, Kratie, Phnom Penh, Prey Veng, Pursat, Siem Reap, Svay Rieng, Takeo, and Otdar Mean Chey. Five groups of provinces: Battambang and Pailin, Kampot and Kep, Preah Sihanouk and Koh Kong, Preah Vihear and Steung Treng, and Mondol Kiri and Rattanak Kiri).\n\nThe sampling frame used two stratified stages (more details available in CDHS-2010)25. Initially, stratification was made by separating 19 domains into urban and rural areas. Then from each of 19 sampling domains one rural area and one urban area were considered totaling 38 sampling strata. Initially, from these 38 sampling strata, 611 enumeration areas (EAs) were selected with a probability proportional to size (PPS) based on the 2008 Cambodia General Population Census. In total, 191 EAs from the urban areas, and 420 EAs from the rural areas were selected. Each household was then listed within each selected EA. From the list, 24 households in each urban EA and 28 households in each rural EA were randomly selected, totaling, 16,344 households.\n\nHowever, 15,829 households had potential respondents during data collection. In these households 19,237 women of reproductive age (15–49 years) who were the usual residents of the selected households, or visitors who had spent the previous night before the survey, were identified as being eligible for individual interview. Finally, 18,749 women responded to the interview questionnaires (response rate 97.5%).\n\nIn total, 109 field staff received 22 days training and 4 days field practice. Each field team was comprised of a team leader, a field editor, three female interviewers, and one male interviewer. Data processing personnel had three data processing supervisors, 10 office editors/coders, 19 data entry operators. In total 37 data entry staff had received necessary classroom training. To minimize data entry error, all questionnaires were entered twice in to the data entry system.\n\nIn the survey, three questions were asked regarding the women respondents’ justification towards parental beating of sons and three questions addressed the parental beating of daughters. “In your opinion, is a parent justified in hitting or beating his son/ daughter for the following reasons:\n\ni) disobeying them: yes/no\n\nii) being impolite: yes/no\n\niii) doing something embarrassing to the family: yes/no\n\nIn the current study, main variables were constructed merging all three reasons “justified beating of son by parents” and “justified beating of daughter by parents”.\n\nAs independent variables, respondents background factors, personal issues and their autonomy issues are considered in the study. Background factors include: age (seven age groups: 15–19, 20–24, 25–29, 30–34, 35–39, 40–45 and 45–49 years); residency (rural/urban); religion (Buddhist/non-Buddhist); education (No-education, primary, secondary and higher); economic status (poorest, poorer, middle, richer and richest); employment status (all round the year, seasonal and occasional) and sex of household head (female or male) and if the respondent was covered by health insurance (yes/no). Personal issues of the respondents consisted of eight variables: exposure to media, such as reading newspapers, listening to radio or watching television (yes/no); sons/daughters who lived at home; sons/ daughters who have died; husband lived with the respondent (living with her/lives elsewhere); justified wife beating (yes/no); number of partners (one/two or more). Autonomy of the respondents was measured by five questions: Who decided on spending money; decision making on healthcare; decision making on large household purchases; decision making on visits to family or relatives; decision making on what to do with the money the husband earns. All questions have had three options: respondent alone; respondent and husband/partner jointly or partner or other person without the respondent.\n\nWe used proportions and chi-squared tests to explore the cross relationships between attitudes towards beating sons/daughters and the independent variables. Bivariate logistic regressions were performed to study the potential associations between the justification of child beating by parents and the respondents’ socioeconomic factors, personal issues and autonomy. IBM SPSS version 22 was used for the data analysis.\n\nThis study used secondary data and hence does not need any ethical permission.\n\n\nResults\n\nOf the 18,749 participants, a significant proportion of women supported physically abusing sons and daughters (69.2% and 67.2%). There is an association between age and supporting beating children. Women who lived in rural area, who had no education, and who were poorest, most often supported beating sons and daughters (p<0.001). This was also more prevalent among non-Buddhist women. In addition, full- or part-time employment of women was associated with supporting beating sons and daughters (p<0.001). Finally, women who came from male-headed households more often supported beating their children (Table 1).\n\nSupport for beating sons and daughters was more prevalent among those parents for whom health care service costs were covered by insurance (p<0.001). A higher proportion of women who had no exposure to media supported beating sons and daughters. A signification proportion of women with at least one dead son/daughter supported beating children. In addition, a higher proportion of women who justified wife beating more often supported beating their child (p<0.001). Finally, women who had two or more lifetime partners more often supported beating sons and daughters (p<0.001) (Table 2).\n\nA significant proportion of women who could not make decisions about their own healthcare, large household purchases, visits to family or relatives or how to spend their husband’s money most often supported beating sons and daughters (p<0.001) (Table 3).\n\nCompared with the oldest age group, women between 20–34 years were more likely to support beating sons and daughters. Uneducated mothers were more likely to support beating of sons (odds ratio (OR) 2.8187, CI 1.936-4.099; P<0.0001) and to support beating daughters (OR 2.644, CI 1.825-3.829; P<0.0001) compared to higher educated women. In addition, compared with the richest women, the poorest mothers were more likely to support child beating (son: OR 1.53, CI 1.205-1.943; daughter: OR 1.432, CI 1134-1.809). Moreover, women who did not justify wife beating were less likely to support child beating (son: OR 0.194, CI 0.172-0.219; daughter: OR 0.206, CI 0.184-0.232). Furthermore, women who had the autonomy to take decision about their own healthcare, visiting family or relatives, and on spending husbands’ money were less likely to support beating sons and daughters (p<0.0001) (Table 4).\n\nReference categories are denoted as Ref. *p<0.001; p<0.01.\n\n\nDiscussion\n\nThis study investigated women’s attitudes towards child physical abuse by parents in relation to the women’s background factors, personal issues and autonomy. The current work provides some new insight in the research concerning the women’s demographic factors and supporting beating sons and daughters.\n\nThe study found that a significant proportion of women supported beating sons and daughters, and that it was more prevalent among uneducated rural dwellers and the poorest mothers. Women who came from male-headed households more often supported beating their children. Women without media exposure proportionally experienced more physical violence than their peers with media exposure. The first group of women (without media exposure) supports more child physical abuse than the other group (with media exposure). Also, women having two or more lifetime partners more often supported beating sons and daughters.\n\nPrevious literature demonstrates a robust connection between poverty and child physical abuse26. The current study also has same findings. Compared with other factors, poverty and low socioeconomic status is consistently associated with child maltreatment and the most severe abuse cases were found among the poorest people27,28. We have found that younger women were more likely to support beating children than older women. In addition, younger parental age, low education, and parental physical-mental health are significant predictors of child maltreatment29. The results of a previous study indicates that families with low socioeconomic status or lower level of income, parental mental illness augments the risk of child abuse and maltreatment30. Our findings have supported the same results in the context of Cambodia. It was also found that mental stress because of inadequate housing, overcrowding, lack of social support, and social isolation all increase the risk of abuse among Cambodian refugee families in America4. Poor families face challenges in providing adequate child care and supervision.\n\nAlthough most of the previous studies found little differences in child abuse incidence rates between urban and rural areas, we found that women who live in rural areas supported beating their children more than in urban areas22–24,31. Similarly, another study found that physical and sexual abuse was common among rural dwellers than non-rural dwellers in the USA32.\n\nThe present study indicates that women’s higher education may be a protective factor for child physical abuse, which is contrary to the existing research, where higher parental education was a risk factor for child physical abuse13. Previous studies also indicated that lack of income to meet the family’s needs as well as no or lower education augments the risk of child physical abuse in both low- and high-income countries33,34.\n\nWe also found that women who experience physical violence more often supported child beating. Previous studies indicated that domestic violence (DV) enhances the risk of child physical abuse17,31. There is evidence that exposure to physical DV is independently associated with an increased risk of mothers using violent methods to correct child behavior1. Children from families with a history of DV are at increased risk of physical abuse, resulting a cycle of violence between spouse and child physical abuse31,35.\n\nThis study found in the bivariate analysis that women who came from male-headed households often supported beating their child. In the logistic regression this was not significant. Similarly, a prior study found that the rate of female-headed households was negatively associated with abuse rates in African American neighborhoods36.\n\nThe results of this study indicate that increased autonomy for women in relation to their own health care and permitting visiting family or relatives and spending husband’s money may reduce son and daughter beating. In fact, World Health Organization (WHO) violence prevention meetings have highlighted female empowerment as a way of tackling violence within the family37. Our study has added materials in to that WHO call37 for women empowerment extending the protective factor from DV to child abuse. However, the current study along with other reproductive age group, includes young women aged 15–19 years who are still regarded as minors in many countries. In Cambodia, girls are often married early. In Cambodia, the majority of the teenaged girls have household autonomy such as own monetary spending, health care, visiting friends/relatives and a level of control over husband/partners income25. Previous studies found that maternal stress, psychopathology and depression are closely linked with child abuse26,37. Less autonomy increases maternal stress which in turn increases the likelihood of child physical abuse. There is evidence that personal autonomy and social capital reduces the onset of depression in women, which ultimately reduces child abuse38.\n\nCompared to other national surveys, the CDHS has some important advantages. CDHS is nationally representative; therefore, findings and conclusions are applicable to the whole of Cambodia. CDHS, like other DHS studies, has high response rates. CDHS has used very high-quality interviewer training and standardized data collection procedures. The sampling methodology is well tested in Cambodia and in other low and middle income countries. The study has some limitations. DHS data from many countries advocate that DHS surveys underestimate the extent of violence and abuse. On the other hand, other surveys like multi-country surveys from WHO or UNICEF better estimate the extent of violence and abuse39. Therefore, the current findings might underestimate the extent of violence and abuse. In many cases, more than one respondent from the same household was interviewed. However, each respondent was interviewed separately and privately to minimize response bias31. The cross-sectional analysis does not allow for causal inference. Therefore, the current study is suffering from the same problem indicating the necessity of longitudinal studies to firmly establishing the causality. This study contributed to the literature by identifying relevant women’s characteristics in relation to attitude towards child physical abuse. Fathers’ characteristics were not assessed in this study. Future studies may include fathers in order to make a full evaluation of familial disciplinary attitudes and behaviors. Also, qualitative studies exploring the attitude towards child physical abuse could be of interest.\n\nThe current work provides knowledge about maternal factors such as age, education, economic status, rural/urban dwelling, two or more lifetime partners and autonomy in the supporting of beating sons and daughters. Further attention needs to be paid to increasing women’s education and autonomy in their family life in Cambodia.\n\n\nData availability\n\nData for this study were obtained from the DHS Program (Cambodia: Standard DHS, 2010). Access to the dataset requires registration, and is granted to those that wish to use the data for legitimate research purposes. A guide for how to apply for dataset access is available at: https://dhsprogram.com/data/Access-Instructions.cfm.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nDalal k, Lawoko S, Jansson B: The relationship between intimate partner violence and maternal practices to correct child behavior: a study on women in Egypt. J Inj Violence Res. 2010; 2(1): 25–33. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nUNICEF: A statistical snapshot on child abuse in East Asia and the Pacific. 2012; cited 2017 Dec 5. Reference Source\n\nChang J, Rhee S, Weaver D: Characteristics of child abuse in immigrant Korean families and correlates of placement decisions. Child Abuse Negl. 2006; 30(8): 881–891. PubMed Abstract \| Publisher Full Text\n\nChang J, Rhee S, Berthold SM: Child abuse and neglect in Cambodian refugee families: characteristics and implications for practice. Child Welfare. 2008; 87(1): 141–60. PubMed Abstract\n\nZhai F, Gao Z: Child maltreatment among Asian Americans: characteristics and explanatory framework. Child Maltreat. 2009; 14(2): 207–24. PubMed Abstract \| Publisher Full Text\n\nMinistry of women’s affairs(MOWA), UNICEF Cambodia, US Centers for Disease Control and Prevention: Findings from Cambodia’s violence against children survey 2013. Cambodia: ministry of women’s affairs, 2014. Reference Source\n\nPerry BD: The neurodevelopmental impact of violence in childhood. In: Schetky D, Benedek EP(Ed). Textbook of child and adolescent forensic psychiatry. Washington DC: American psychiatric risk of mother-reported child abuse in the first three years of life. Child abuse and neglect; 2001. Reference Source\n\nCorcoran J: Family interventions with child physical abuse and neglect: a critical review. Child Youth Serv Rev. 2000; 22(7): 563–591. Publisher Full Text\n\nDraper B, Pfaff JJ, Pirkis J, et al.: Long-term effects of childhood abuse on the quality of life and health of older people: results from the Depression and Early Prevention of Suicide in General Practice Project. J Am Geriatr Soc. 2008; 56(2): 262–271. PubMed Abstract \| Publisher Full Text\n\nJewkes RK, Dunkle K, Nduna M, et al.: Associations between childhood adversity and depression, substance abuse and HIV and HSV2 incident infections in rural South African youth. Child Abuse Negl. 2010; 34(11): 833–41. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNorman RE, Byambaa M, De R, et al.: The long-term health consequences of child physical abuse, emotional abuse, and neglect: a systematic review and meta-analysis. PLoS Med. 2012; 9(11): e1001349. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFuller-Thomson E, Brennenstuhl S: Making a link between childhood physical abuse and cancer: results from a regional representative survey. Cancer. 2009; 115(14): 3341–50. PubMed Abstract \| Publisher Full Text\n\nLueger-Schuster B, Knefel M, Glück TM, et al.: Child abuse and neglect in institutional settings, cumulative lifetime traumatization, and psychopathological long-term correlates in adult survivors: The Vienna Institutional Abuse Study. Child Abuse Negl. 2018; 76: 488–501. PubMed Abstract \| Publisher Full Text\n\nWorld Health Organization: Preventing child maltreatment: a guide to action and generating evidence. Geneva: world health organization and international society for prevention of child abuse and neglect; 2006. Reference Source\n\nStraus M, Gelles R, Steinmetz S: Physical violence in the American family. Beverly Hills, CA: Sage Publications. 1990; 622. Reference Source\n\nUNICEF: Child disciplinary practices at home: evidence from a range of low- and middle-income countries. UNICEF: New York, 2010. Reference Source\n\nStraus MA: Beating the devil out of them: corporal punishment by parents and its effects on children. New york: Lexington books. 1994; 297. Reference Source\n\nZhu Y, Dalal K: Childhood exposure to domestic violence and attitude towards wife beating in adult life: a study of men in India. J Biosoc Sci. 2010; 42(2): 255–69. PubMed Abstract \| Publisher Full Text\n\nStoltenborgh M, Bakermans-Kranenburg MJ, van Ijzendoorn MH: The neglect of child neglect: a meta-analytic review of the prevalence of neglect. Soc Psychiatry Psychiatr Epidemiol. 2013; 48(3): 345–55. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChan KL, Brownridge DA: Personality characteristics of Chinese male batterers: an exploratory study of women's reports from a refuge sample of battered women in Hong Kong. Am J Mens Health. 2008; 2(3): 218–28. PubMed Abstract \| Publisher Full Text\n\nAl Dosari MN, Ferwana M, Abdulmajeed I, et al.: Parents' perceptions about child abuse and their impact on physical and emotional child abuse: A study from primary health care centers in Riyadh, Saudi Arabia. J Family Community Med. 2017; 24(2): 79–85. PubMed Abstract \| Free Full Text\n\nChan KL: Co-occurrence of intimate partner violence and child abuse in Hong Kong Chinese families. J Int Per Viol. 2011; 26(7): 1322–1342. PubMed Abstract \| Publisher Full Text\n\nDouki ZE, Esmaeili MR, Vaezzadeh N, et al.: Maternal child abuse and its association with maternal anxiety in the socio-cultural context of iran. Oman Med J. 2013; 28(6): 404–409. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPark MS: The factors of child physical abuse in Korean immigrant families. Child Abuse Negl. 2001; 25(7): 945–58. PubMed Abstract \| Publisher Full Text\n\nCDHS. 2010. cited 2018 Jan 7. Reference Source\n\nHahm HC, Lee Y, Ozonoff AI, et al.: The impact of multiple types of child maltreatment on subsequent risk behaviors among women during the transition from adolescence to young adulthood. J Youth Adolesc. 2010; 39(5): 528–540. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFourth National Incidence Study of Child Abuse and Neglect (NIS-4): Report to Congress. 2010. cited 2017 Nov 26. Reference Source\n\nBywaters P, Bunting L, Davidson G, et al.: The relationship between poverty, child abuse and neglect: an evidence review. York: Joseph Rowntree Foundation. 2016. Reference Source\n\nHolt S, Buckley H, Whelan S: The impact of exposure to domestic violence on children and young people: a review of the literature. Child Abuse Negl. 2008; 32(8): 797–810. PubMed Abstract \| Publisher Full Text\n\nPrinz RJ: Parenting and family support within a broad child abuse prevention strategy: Child maltreatment prevention can benefit from public health strategies. Child Abuse Negl. 2016; 51: 400–406. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDalal K, Lindqvist K: A national study of the prevalence and correlates of domestic violence among women in India. Asia Pac J Public Health. 2012; 24(2): 265–77.35. PubMed Abstract \| Publisher Full Text\n\nFreisthler B, Merritt DH, LaScala EA: Understanding the ecology of child maltreatment: a review of the literature and directions for future research. Child Maltreat. 2006; 11(3): 263–80. PubMed Abstract \| Publisher Full Text\n\nIsaranurug S, Nitirat P, Chauytong P, et al.: Factors relating to the aggressive behavior of primary caregiver toward a child. J Med Assoc Thai. 2001; 84(10): 1481–1489. PubMed Abstract\n\nDalal K, Rahman F, Jansson B: The origin of violent behaviour among child labourers in India. Glob Public Health. 2008; 3(1): 77–92. PubMed Abstract \| Publisher Full Text\n\nStraus M, Gelles R, Steinmetz S, et al.: Behind closed doors. Garden City, Ny: Anchor Press. 1980.\n\nGarbarino J, Kostelny K: Child maltreatment as a community problem. Child Abuse Negl. 1992; 16(4): 455–464. PubMed Abstract \| Publisher Full Text\n\nWHO: Inspire: Seven Strategies For Ending Violence Against Children. Geneva: WHO. 2016. Reference Source\n\nBojorquez-Chapela I, Manrique-Espinoza BS, Mejía-Arango S, et al.: Effect of social capital and personal autonomy on the incidence of depressive symptoms in the elderly: evidence from a longitudinal study in Mexico. Aging Ment Health. 2012; 16(4): 462–71. PubMed Abstract \| Publisher Full Text\n\nCorsi DJ, Neuman M, Finlay JE, et al.: Demographic and health surveys: a profile. Int J Epidemiol. 2012; 41(6): 1602–13. PubMed Abstract \| Publisher Full Text" }	[ { "id": "54537", "date": "07 Oct 2019", "name": "Mina Golestani", "expertise": [ "Reviewer Expertise Injury research", "of course I have also published an article in this area" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIt seems very important to choose the subject of child abuse and research in this area; in this way, a few items seemed vague that needed correction:\nIt's not better to look at social status or relationships with friends in relation to confounding factors. It also seems to be effective in child abuse.\n\nIn Table 2, the authors noted that \"media exposure\" had a significant relationship with the child abuse. Please explain this clearly.\n\nThe authors stated that \"families with daughters/sons who died\" supported beating the child - this is due to many factors regardless of the cause of the child's loss. Please explain the reason for this in the paper.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "66337", "date": "03 Jul 2020", "name": "John D Melville", "expertise": [ "Reviewer Expertise Child Abuse Pediatrics" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe work presents a cross sectional study of attitudes of Cambodian women regarding corporal punishment.\n\nA few criticisms:\nThe question asked, when translated into English, combines both \"hitting\" and \"beating\" which, at least in the American South, considers different degrees of severity. It is unclear to me what behaviors the women considered to be acceptable.\n\nThe risk factors identified reflect well known risk factors for physical discipline, and the ORs are quite modest. Many of these small distinctions are statically significant due to the large sample size.\n\nThe reporting of the logistic regression is unclear. The methods section describes \"bivariate logistic regression\" which in the results section is titled \"multivariate analysis.\" Logistic regression can be used for both bivariate and multivariate analysis but the interpretation of the resulting numbers is very different. The article should clarify how the analysis was performed.\n\nThe second paragraph of the introduction has little relevance to the paper presented and should be omitted.\n\nA somewhat surprising statement: \"It has been narrated that mothers are twice more likely to psychologically or physically abuse their children than fathers\" is cited to Al Dosari (2017). Upon reviewing the cited paper it is not immediately clear that the citation supports the claim.\n\nThe authors suggest that increasing female autonomy will decrease child physical abuse. While this may be true, it is inappropriate to infer causation in a cross-sectional survey such as this one.\nIn summary this paper reports a secondary analysis of a significant effort to survey a representative sample of Cambodian women. The article is of limited significance, only because the results are strongly concordant with prior research. The paper could be improved with a better description of the multivariate analysis and a more focused hypothesis and conclusion.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [] } ]	1	https://f1000research.com/articles/7-1866
https://f1000research.com/articles/7-1863/v1	29 Nov 18	{ "type": "Research Article", "title": "Tracking the timely dissemination of clinical studies. Characteristics and impact of 10 tracking variables", "authors": [ "Daniel Strech", "Sören Sievers", "Stefanie Märschenz", "Nico Riedel", "Susanne Wieschowski", "Jörg Meerpohl", "Holger Langhof", "Stephanie Müller-Ohlraun", "Ulrich Dirnagl", "Sören Sievers", "Stefanie Märschenz", "Nico Riedel", "Susanne Wieschowski", "Jörg Meerpohl", "Holger Langhof", "Stephanie Müller-Ohlraun", "Ulrich Dirnagl" ], "abstract": "Background: Several meta-research studies and benchmarking activities have assessed how comprehensively and timely, academic institutions and private companies publish their clinical studies. These current “clinical trial tracking” activities differ substantially in how they sample relevant studies, and how they follow up on their publication. Methods: To allow informed policy and decision making on future publication assessment and benchmarking of institutions and companies, this paper outlines and discusses 10 variables that influence the tracking of timely publications. Tracking variables were initially selected by experts and by the authors through discussion. To validate the completeness of our set of variables, we conducted i) an explorative review of tracking studies and ii) an explorative tracking of registered clinical trials of three leading German university medical centres. Results: We identified the following 10 relevant variables impacting the tracking of clinical studies: 1) responsibility for clinical studies, 2) type and characteristics of clinical studies, 3) status of clinical studies, 4) source for sampling, 5) timing of registration, 6) determination of completion date, 7) timeliness of dissemination, 8) format of dissemination, 9) source for tracking, and 10) inter-rater reliability. Based on the description of these tracking variables and their influence, we discuss which variables could serve in what ways as a standard assessment of “timely publication”. Conclusions: To facilitate the tracking and consequent benchmarking of how often and how timely academic institutions and private companies publish clinical study results, we have two core recommendations. First, the improvement in the link between registration and publication, for example via institutional policies for academic institutions and private companies. Second, the comprehensive and transparent reporting of tracking studies according to the 10 variables presented in this paper.", "keywords": [ "clinical studies", "trials", "registries", "follow-up", "trial tracking", "university medical centers", "private companies" ], "content": "List of abbreviations\n\nCD, completion date; CENTRAL, Cochrane Central Register of Controlled Trials; CSR, clinical study reports; EU, European Union; FDA, Food and Drug Administration; FDAAA, Food and Drug Administration Amendments Act; GCP, Good Clinical Practice; ICH, International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use; ICMJE, International Committee of Medical Journal Editors; NCT, national clinical trial; PCD, primary completion date; UMC, University medical center\n\n\nBackground\n\nThe results of clinical trials and observational studies form the basis of evidence-based decision making in health care, coverage decisions, the planning and funding of new research studies, ethics reviews, and research quality assessment1. Over the past three decades, several meta-research projects have demonstrated that the results of clinical studies are often not reported at all, or the reporting is incomplete, biased, or inconsistent with what was planned at the protocol stage2–5.\n\nRecent studies have compared the proportion of trials/studies with published/disseminated results across academic institutions and for-profit companies. For instance, Chen et al. compared the publication rates of completed trials conducted by 51 academic medical centres across the US6. They found that overall, the results were published for 67% of 4,347 trials; however, the results of only 36% of trials were published within two years of study completion. They identified wide variation in the proportion of trials being published in a timely manner across the 51 medical centres. The TrialsTracker project, which started in 2016, automatically identifies completed trials on Clinicaltrials.gov, and links them with results from automated searches for published results to present data on publication rates at the individual study centre level (see TrialsTracker). In the newest version of TrialsTracker, approximately 65% of all applicable trials were found to be reported (accessed 21 March 2018). A third approach to this line of research is the “Good Pharma Scorecard”, which takes Food and Drug Administration (FDA)-approved drugs as a starting point, and provides benchmarks for transparency in industry-run clinical drug research and development7,8. In this initiative, the authors combined data from various sources (e.g., Clinicaltrials.gov, PubMed, Google Scholar) and different forms of results dissemination (summary reports, clinical study reports, peer-reviewed publications). This study revealed that of 505 trials relating to 31 drugs approved by the FDA in 2014, the results for a median of 68% of trials per drug were publicly available. The authors also found that in 233 of the 505 trials that were conducted with patients (in contrast to healthy participants), a median of 96% of trials, per drug, were publicly available.\n\nThe above-mentioned studies differed in many ways regarding their methods for sampling relevant clinical studies, and following up on their publication. What are the definition and operationalization of a “completed trial”? Should publication follow-up be applied only for prospectively registered trials, or also for trials that were registered many months or years after the study started? What databases are searched for publications? What publication formats and contents count as a “publication”? What amount of time to publication is considered “timely”? Should only clinical trials or also observational studies be followed up on? Should only completed studies or also discontinued ones be followed up on? Different decisions on the various tracking variables and decisions on other methodologically relevant issues lead to different results for how academic institutions and for-profit companies perform overall and on comparative rankings or benchmarks.\n\nThe comprehensive and timely publication of clinical study results affects the reputation of academic institutions and private companies. Such data impact public trust and the willingness of foundations to fund research. To provide a basis for informed policy and decision making through publication assessment and benchmarking, this paper aims to identify and characterize the different variables affecting the results of tracking whether and how timely results from clinical studies are published.\n\n\nMethods\n\nThe selection of tracking variables presented in this paper was initially driven by expert knowledge and discussions within the group of authors. To validate the completeness of our set of tracking variables, we then conducted i) an explorative review of studies that followed up on the publication of clinical studies and ii) an explorative follow-up study of registered clinical trials of three leading German university medical centres (UMCs): Berlin, Freiburg, and Hannover.\n\nThe explorative review started with a set of eight follow-up studies that we were aware of6,8–14. All references of these (and the later-included) studies were evaluated for additional follow-up studies. Altogether, we identified 34 follow-up studies. All identified studies were read in full to extract reported methods for sampling and follow-up. The extracted content was checked for methodological details that our initial expert-driven set of tracking variables did not contain, and these were added to our set accordingly. The detailed results of this review of follow-up studies will be published elsewhere.\n\nThe methods for our explorative follow-up study have been published as a preregistered study protocol (Extended data15). We used an R script (see Software availability) to combine all relevant datasets from clinicaltrials.gov and search criteria needed to retrieve clinical trials from all 36 German UMCs and to extract their study characteristics. For each of the included studies, a results publication was searched independently by two researchers in a 3-step process in i) the registry, ii) in Pubmed, and iii) in Google Scholar. In the meantime, we conducted our study based on this protocol, evaluating all 36 German UMC as specified by the German Medical Faculty Association (MFT). The results of this comprehensive follow-up study will be published elsewhere. Our explorative follow-up study helped to further clarify how the tracking variables described in this paper influence the results of follow-up studies.\n\n\nResults\n\nBased on the above-mentioned expert discussion, as well as the review of existing follow-up studies, and the insights from our explorative follow-up study of three German UMC, we were able to distinguish 10 variables influencing the design and results of follow-up studies of clinical studies. Table 1 categorizes the 10 variables into two broad areas: “sampling of studies” and “follow-up of studies”.\n\nBelow, we describe the tracking variables, potential challenges in their operationalization, and their current influence on publication assessments. In the Discussion section, conceptual and normative issues for all 10 variables will be explored, and we will provide practical recommendations for future study tracking.\n\nFrom a legal perspective, a specific university and its clinical study investigators are responsible for the dissemination of study results only if they are the “responsible party”, that is, if they are either the sponsor and/or the principal investigator. The International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) Guideline for Good Clinical Practice (GCP), for example, defines in section 1.53 that the sponsor is “an individual, company, institution, or organization which takes responsibility for the initiation, management, and/or financing of a clinical trial”. The new European Union (EU) clinical trials regulation 536/2014, which is expected to come into force in 2019, also refers to the ICH GCP. According to that legal perspective, a specific university and its investigators who are “only” cooperating as a recruiting site are “not responsible” for whether and how timely the trial results are published. From an ethical perspective, however, physician investigators should only recruit participants for clinical trials that generate social value. Every recruiting physician investigator should therefore feel responsible for facilitating the timely publication of study results.\n\nSeveral follow-up studies explicitly sampled clinical trials according to this legal definition of responsibility6. Other follow-up studies sampled all trials approved by local ethics committees, but did not further specify the issue of responsibility10,12. Our explorative study demonstrated that only one third of trials conducted at the three German UMC are trials for which the university is the “legally” responsible party.\n\nClinical studies are broadly differentiated into interventional trials and observational studies, with different legal requirements. Therefore, studies following up on clinical studies for a specific university must decide whether to include both or only one study type or only subgroups, such as prospective cohort studies. ClinicalTrials.gov currently (21 July 2018) lists 221,251 interventional trials and 55,875 observational studies.\n\nMost follow-up studies we reviewed focused on interventional trials. One exception is the study by Ross et al. that followed up on a sample of both types of clinical studies that were registered in ClinicalTrials.gov and found that the publication rate was 56% for interventional, placebo-controlled trials and 42% for observational studies11. In a recent study, Spelsberg et al. were able to follow up on a full sample of 558 observational post-marketing studies on adverse drug reactions16. They could not find any results reported for the 558 studies in the drug regulator’s public adverse drug reactions database. A peer-reviewed journal publication was found for five (1%) post-marketing studies16.\n\nAnother relevant characteristic of clinical studies is whether they are subject to mandatory reporting requirements. Several more recent follow-up studies excluded all trials that do not fall within the mandatory reporting rules according to the Food and Drug Administration Amendments Act (FDAAA) 8017,11,14,17.\n\nMost follow-up studies focus on the analysis of how timely completed studies are published, e.g.,11,17,18. These follow-up studies therefore must exclude all terminated, discontinued, and withdrawn trials. Knowledge, however, is also gained from discontinued trials that, for example, report recruitment barriers or unanticipated adverse effects. A follow-up study by Pica et al. explicitly included all clinical trials, irrespective of whether they were completed, terminated, withdrawn, or suspended13. Anderson et al. also included both completed and terminated studies14. Others further specified the exclusion of non-completed studies. Miller et al., for example, stated that they excluded any trials that were terminated without participant enrolment7, but they do not explicitly state how they addressed other types of terminated, withdrawn, or suspended trials. Kasenda et al. showed that for 1,017 clinical trials approved at one of six institutional review boards (IRBs) in Germany, Switzerland and Canada, 25% were discontinued12. Discontinued trials were more likely to remain unpublished than completed trials (55% vs 34%).\n\nIRB archives should be the most sensitive and least biased source for sampling all clinical studies conducted at one specific university, at least in countries where ethics reviews of clinical studies became mandatory sometime after the 1975 Tokyo revision of the Declaration of Helsinki19. The previously mentioned study by Kasenda et al., for example, found that 10 years after the IRB approval of 1,017 clinical trials, the results of 56% (n=567) were published as a full journal article12.\n\nIn practice, however, most newer studies following up on clinical trials refer to public registries, as this is more convenient and practicable. Registries, however, rarely include all clinical studies conducted at any given university. The prospective registration of clinical trials in a public registry became legally binding under FDAAA 801 in the USA for all phase II/III trials in 2007. Another strong incentive to register clinical trials was the International Committee of Medical Journal Editors (ICMJE) policy issued in 2015, which recommends that all medical journals require prospective registration as a condition of consideration for publication20. In the EU, prospective registration for clinical trials in the public EU Clinical Trials Database will become a legal requirement with EU regulation 536/2014 (Article 67). However, the same Article 67 still allows certain pieces of information not to be published, i.a. also recognizing the “legitimate economic interests” of sponsors. All clinical studies that do not fall under these laws (observational studies, trials on psychotherapy, surgery studies) can nevertheless be registered, even retrospectively. Van den Bogert et al. found publication rates of 75% for prospectively registered IRB-approved clinical trials and 48% for non-registered trials21.\n\nThe inclusion of all registered trials by studies following up on the publication of results might overestimate publication rates for another reason. Even trials with mandatory registration are sometimes registered after the sponsor plans to publish the results, and then realizes that this requires registration. Although the above-mentioned ICMJE policy explicitly refers to prospective registration, many journals also allow retrospective registration22,23.\n\nHow to determine the cut-off between prospective and retrospective registration is complicated by two points. First, legal obligations for registration allow a time window. The FDAAA 801, for example, requires trial registration within 21 days after the enrolment of the first participant. Second, several “timings” for registration can be distinguished. The ISRCTN registry, for example, reports both the date that researchers submitted their request for registration and the date that trial registration was completed.\n\nRecent follow-up studies have differed in how they deal with retrospectively registered trials and how they define “retrospective”. Pica et al., for example, excluded all trials that were registered more than 60 days after the study started13. Others, such as Chen et al. and TrialsTracker, did not exclude retrospectively registered trials at all and did not allow for subgroup analyses6,24.\n\nIn our explorative follow-up study, we found that the proportion of trials with published results was 68% for all prospectively registered trials, 73% for trials registered more than 60 days after the trial started, and 82% for trials registered after study completion. We also found 16 trials registered after the date of the first results publication, which of course have a 100% publication rate.\n\nOne can broadly distinguish three ways to determine the completion date of an interventional trial: A) the “primary completion date” (PCD), which, according to ClinicalTrials.gov, is the “date on which the last participant in a clinical study was examined or received an intervention to collect final data for the primary outcome measure”, B) the later “completion date” (CD), which is defined as the date of the last participant’s last visit to collect final data, and thus also includes secondary outcome measures and adverse events, or C) the “estimated” study completion date, which is the study completion date expected by the researchers. According to the ClinicalTrials.gov glossary, the “estimated” date is perceived in the same way as the CD.\n\nAt present, follow-up studies using registries are heterogeneous with regard to the CD definition to which they refer. Some refer to the PCD14,18, some refer to the CD11, and some do not specify25. A follow-up study that sampled clinical study protocols archived at IRBs regarded a study as completed if the data collection was terminated or if the study results were published10.\n\nIn our pilot follow-up study, we found, unsurprisingly, that the proportion of published trials is higher for all three German UMC 24 months after the CD (35% of trials published) than 24 months after the PCD (28% published).\n\nOf course, the proportion of published studies also varies considerably with regard to what one accepts as “timely”. Both the FDA and EU legal frameworks allow 12 months for the publication of “summary results”, which form part of the respective registry entries. The FDAAA has mandated results reporting since 200826. The EU Commission introduced similar requirements in 2014 but is still facing implementation barriers27. However, no offical standards exist for how timely more detailed and contexualized results should be published in peer-reviewed journals.\n\nOngoing and past follow-up studies have dealt differently with how and where to set adequate time frames. TrialsTracker reports the proportion of trials for each sponsor published within 24 months after the PCD9,24. Ross et al. reported the publication rates for a timeframe of 30 months after the CD. Chen et al. focused their results reporting on 24 months after the PCD but also illustrated how the publication ratio changed with alternative timeframes6. Miller et al. compared the publication ratio at FDA approval, and 3 months and 6 months post-approval8. Many other follow-up studies have assessed the all-time publication ratio, often allowing more than 10 years between the completion and publication of the trial10,12.\n\nIn our pilot study with three German UMC, we found publication rates (including summary results) of 16% for 12 months after the CD, 35% for 24 months after the CD, and 71% for all-time follow-up.\n\nWhat should count as a relevant publication or other dissemination format? Peer-reviewed publications are the typical dissemination format and were accordingly accepted by almost all of the above-mentioned follow-up studies. Another newer, but increasingly accepted, dissemination format is that of the previously mentioned summary results6,9. Miller et al. further included clinical study reports (CSRs)8. Other formats such as theses, conference proceedings, books, or data uploaded on data-sharing platforms might also reveal important and sufficient information on trial results. The 34 follow-up studies we are aware of did not include any of the latter publication formats in their searches.\n\nIrrespective of the dissemination format, one might count only those result publications that report the essential information needed. Guidance exists on the essential content to be included in summary results (see ClinicalTrials.gov28 definitions), journal publications29, and CSRs30. However, appraising each identified publication for its comprehensiveness and appropriateness requires considerable time. Some follow-up studies invested this time. Kasenda et al., for example, checked all full texts of retrieved publications and demonstrated that they often do not report on all predefined outcomes or deviate in other ways from registered information on the study design31.\n\nTime to publication also strongly depends on where publications are sought. A first obvious search can be performed on the registry itself. The ClinicalTrials.gov database highlights for each trial the “first results received”, if available. However, the “first results received” only illustrate when summary results for the trial were reported. The registry entry for each trial further automatically indexes all publications listed in PubMed that mention the trial-specific NCT identifier in the abstract (see NIH page of registry numbers)32. Furthermore, ClinicalTrials.gov allows sponsors and principal investigators to manually index publications for registered trials.\n\nAnother obvious search strategy is to search PubMed for the NCT number. As the automatic indexing at ClinicalTrials.gov does not work in all cases, this approach might reveal additional result publications. In our explorative follow-up study of German UMC, we identified an additional 3% of result publications via this approach. Chen et al. and TrialsTracker also checked for NCT identifiers in PubMed6,24.\n\nFurther search strategies become more time intensive and require more interpretive judgements. Bluemle et al. searched the Cochrane Central Register of Controlled Trials (CENTRAL)10. They contacted the applicants of all included protocols by personal letter. For each submitted protocol, they asked individually about the current project status, the verification of already-identified publications, and references of additional publications they may have missed. Their literature searches identified 138 full publications, and the survey identified an additional 72. However, Bluemle et al. did not report how much these additional publications changed the overall publication rate, which turned out to be 48%10.\n\nNone of the 34 follow-up studies we are aware of searched publications for trials in general search engines such as Google, Bing, or Yahoo. In our follow-up of trials from three German UMC, we identified result publications for 48% of all trials by searching summary results and indexed publications at the respective registry entries, combined with PubMed searches for NCT identifiers. However, additional manual searches in Google by two independent searchers yielded result publications for another 27% of trials. The manual search, therefore, increased the overall proportion of result publications to 75% for the three academic institutions.\n\nFinally, some private drug and medical device companies operate company-specific databases that might list publications from completed trials. In response to results from TrialsTracker, a blogger, Adam Jacobs, argued that the 45% of trials that TrialsTracker found to be undisclosed shrink to 21% if one searches in these company-specific databases (see The Stats Guy blog)33. However, each of these company-owned databases functions in a different way, and many companies do not publish such databases.\n\nExpertise, ideally from more than one rater, is required to determine whether a publication matches a specific trial, and can thus be considered ‘published’. This certainly applies if manual searches in databases or internet search engines are added to automated registry and database checks. In Kasenda et al., for example, two investigators working independently, and in duplicate determined whether identified publications matched the corresponding protocol12.\n\nAs we also applied manual searches in our explorative follow-up study, we had at least two researchers independently search for publications of registered trials. Although all researchers had a background in systematic review methodology and were trained in identifying result publications for clinical trials, we faced high inter-rater differences. For 16% of all trials, a publication was found by only one person, and for 10% of all trials, two different publications were found.\n\n\nDiscussion\n\nIn this paper, we identified and characterized 10 variables influencing the tracking of whether and how timely clinical studies from academic institutions and private companies are disseminated. We further demonstrated the current opportunities and challenges of using these variables in tracking studies. Some of these variables need further conceptual and normative clarification.\n\nResponsible party for clinical studies: First, we see the need to revisit our understanding of who is “responsible” for the timely (and unbiased) publication of trial results. All physicians functioning as investigators and their universities should not feel responsible only for trials where they are the legally defined “responsible party”. Investigators (both principal and co-investigators) and the hosting academic institutions are ethically obliged to proactively force the timely publication of all trials that recruited “their” patients, even if they only recruited as a cooperating partner. The risks and burdens for patients participating in clinical trials are justifiable only if the study generates social value in terms of knowledge gains. Therefore, the timely and unbiased publication of trial results is to be understood as a basic promise each physician investigator gives to participants recruited for a specific study.\n\nCurrent efforts to benchmark universities should consider the current legal and ethical perspectives on “responsibility” for timely publication. This requires following up on all clinical studies that recruit patients from a given UMC, irrespective of whether the university was the sponsor/principal investigator or only a cooperating partner/facility. Data on the publication ratios of the two samples could be reported separately.\n\nTypes and characteristics of clinical studies: Observational clinical studies are less regulated than interventional drug and device trials, and they do not face mandatory registration or reporting policies. From an ethical and economic perspective, however, those conducting observational studies have the same duties to increase value and reduce waste in biomedical research. For pragmatic reasons, follow-up studies currently focus on clinical studies that face mandatory registration and reporting, but future activities should also aim to shed more light on the registration and publication practices for other types of clinical studies.\n\nStatus of clinical studies: Reporting on the results or relevant barriers of discontinued, withdrawn, or early terminated studies is governed by the same ethical guidelines as completed clinical studies. Reporting does not necessarily require peer-reviewed publications but could also include reporting at registry websites and data-sharing platforms. The reporting of discontinued trials is a relevant measure to benchmark universities’ and companies’ contibutions to increasing value and reducing waste in research. The extent of trial discontinuation itself, however, does not serve as an appropriate measure for benchmarking activities. Academic institutions or companies conducting many complex or high-risk trials where discontinuation might be more probable than in simple trial designs should not be censured or discouraged. Furthermore, stigmatizing the discontinuation of trials might result in the inappropriate continuation of trials.\n\nSources for sampling: As long as study registration is not mandatory for all clinical studies, follow-up studies sampling at the registry level will most likely overestimate the true proportion of published clinical studies. To better understand the reporting performance of individual academic institutions and companies for non-registered clinical studies, follow-up studies must sample at the IRB level. Another way to improve opportunities to evaluate the reporting of, for example, observational studies would be legal or institutional policies requiring the prospective registration of all clinical studies.\n\nTiming of registration: Little is known about how much the timing of registration affects the likelihood of results publication of clinical studies. Our above-reported results from a pilot study indicate a higher proportion of results publication in retrospectively registered trials. Follow-up studies interested in benchmarking the timely publication of trial results for universities or companies, therefore, should either focus on samples of prospectively registered trials to avoid bias or at least report the subgroup results for all prospectively and retrospectively registered trials.\n\nDetermination of completion date: Another normative and policy-oriented question is whether the appropriate definition of “completion” should refer to the PCD or the CD (see definitions above). Should certain trials be labeled as “not timely published” with reference to a 24-month time window after the PCD, even if they published the primary and secondary outcomes within 24 months after the CD? In line with European law (536/2014 Art 37.4), we propose taking the CD as the start date when following up on the reporting of summary results and peer-reviewed publications.\n\nTimeliness of dissemination: We need a standard for what counts as “timely” publication. Several laws require the publication of summary results for certain types of studies within 12 months after the PCD (USA) or the CD (EU). Many other national laws, however, have no requirement in this regard. What should be seen as an ethically justifiable time to appropriately disseminate trial results via peer-reviewed publications or other dissemination formats? We propose 24 months after the CD as a potential normative standard for “timely publication” of peer-reviewed publications or similarly comprehensive and contextualized dissemination formats. Even for busy researchers, this 24-month period should allow enough time to publish. Publishing peer-reviewed results more than 24 months after the CD is, of course, better than not reporting them at all, but it should not be labeled as being “timely”. According to recent follow-up studies, less than 15% of all trials reported summary results within 12 months after the PCD14, and less than 30% produced peer-reviewed publications within 24 months after the CD6.\n\nFormat of publication/dissemination: Another complex issue is what types of publication or dissemination of trial results one should accept as appropriate. Most follow-up studies currently search for published summary results and/or peer-reviewed publications. But what about trial results provided via data-sharing platforms only or by industry-owned databases? Recent follow-up studies have demonstrated that accepting other publication formats, such as CSRs, yields higher publication rates8. It is problematic, however, that these publications are not more easily accessible. Sponsors or investigators should directly link all publications. irrespective of their publication format, to the relevant registry entry of the respective trial. Registries can thus become the one-stop shop for clinical trial stakeholders (e.g., physicians, patients, systematic review and clinical guideline groups, meta-research, oversight).\n\nSources for tracking: As indicated above, at present, the addition of manual searches for publications in internet search engines and for CSRs in industry-owned trial databases yields a much higher number of result publications. Follow-up studies should explicitly acknowledge these limitations if they apply less extensive searches. From a normative perspective, however, these limitations and complexities in searching for result publications are themselves a problem. The above-mentioned registry entry as a one-stop shop for clinical trial results could solve this problem. Academic institutions and private companies should develop policies and incentives for linking result publications with the registry entry of the relevant trial.\n\nInter-rater reliability: Our finding that even experts trained in searching for clinical trials relatively often found different publications for the same trials once again speaks in favour of the one-stop shop approach.\n\nOur study has the following limitations. First, although we reviewed more than 30 follow-up studies, and engaged in following up on clinical trials from German UMC, we might either have missed important tracking variables, or framed the 10 outlined tracking variables in a way that underplays certain aspects that deserve more attention. Second, in our outlines and the discussion of the 10 tracking variables, we only gave examples of how they are currently operationalized in follow-up studies. We will publish more detailed descriptions of the existing body of follow-up studies elsewhere.\n\n\nRecommendations\n\nFirst, registration and publication activities for specific trials should be better linked — from publication to registry entry and from registry entry to publication. This applies not only to peer-reviewed papers including the NCT but also to CSRs, PhD theses, and other formats for results publication. Consequently, registry entries for clinical studies should become a one-stop shop for clinical study stakeholders. If successful and broadly established, the threaded publications initiative might become an even more appropriate one-stop shop in this regard34,35.\n\nSecond, to reach this goal, academic institutions and private companies should develop policies and incentives to further improve i) the registration of all their clinical studies, ii) the timely publication of all their completed and discontinued studies, and iii) the effective linkage of registry entries and their respective publications.\n\nThird, future follow-up and tracking studies, as well as consequent rankings/benchmarking for academic institutions and private companies, should be transparent on how they specified the 10 tracking variables outlined in this paper and why they did so with regard to their follow-up objectives.\n\nRecent efforts at more systematic, comprehensive, and sustained follow-up/tracking activities around the timely publication of clinical studies in combination with recent announcements from leading funders to require and evaluate result publication suggest an intensified public interest in the topic. Therefore, ranking and benchmarking academic institutions and private companies according to their registration and publication efforts seems to be a logical consequence. We hope that our clarification of relevant tracking variables will help make these new assessment activities more valid, effective, and efficient from the start.\n\n\nSoftware availability\n\nThe R script used to as part of this study is available from Open Science Framework https://doi.org/10.17605/OSF.IO/FH42615\n\nLicence: MIT License\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\nData generated in the pilot study as well as the subsequent follow-up study of all German UMCs will be published in a separate publication that will be linked in the Open Science Framework registration of the project.\n\nOSF: Dataset 1: IntoValue. https://doi.org/10.17605/OSF.IO/FH42615\n\nThe data is available under a CC0 1.0 Licence\n\n\nExtended data\n\nThe methods for our explorative follow-up study have been published as a preregistered study protocol, available from Open Science Framework.\n\nOSF: Extended data: IntoValue. https://doi.org/10.17605/OSF.IO/FH42615\n\nAvailable under a CC0 1.0 licence", "appendix": "Grant information\n\nThis study was supported by intramural funding of Charité – Universitätsmedizin Berlin, Hannover Medical School, and University of Freiburg.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe thank the members of the IntoValue team, namely, Bob Siegerink, Katharina Kunzweiler, Christopher Schürmann, Hannes Kahrass, and Sean Kelley, for their contribution to the design and conduct of the pilot study with three German UMC.\n\n\nReferences\n\nMacleod MR, Michie S, Roberts I, et al.: Biomedical research: increasing value, reducing waste. Lancet. 2014; 383(9912): 101–4. PubMed Abstract \| Publisher Full Text\n\nChan AW, Song F, Vickers A, et al.: Increasing value and reducing waste: addressing inaccessible research. Lancet. 2014; 383(9913): 257–66. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSong F, Parekh S, Hooper L, et al.: Dissemination and publication of research findings: an updated review of related biases. Health Technol Assess. 2010; 14(8): iii, ix–xi, 1–193. PubMed Abstract \| Publisher Full Text\n\nDickersin K, Min YI, Meinert CL: Factors influencing publication of research results. Follow-up of applications submitted to two institutional review boards. JAMA. 1992; 267(3): 374–8. PubMed Abstract \| Publisher Full Text\n\nEasterbrook PJ, Berlin JA, Gopalan R, et al.: Publication bias in clinical research. Lancet. 1991; 337(8746): 867–72. PubMed Abstract \| Publisher Full Text\n\nChen R, Desai NR, Ross JS, et al.: Publication and reporting of clinical trial results: cross sectional analysis across academic medical centers. BMJ. 2016; 352: i637. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMiller JE, Korn D, Ross JS: Clinical trial registration, reporting, publication and FDAAA compliance: a cross-sectional analysis and ranking of new drugs approved by the FDA in 2012. BMJ Open. 2015; 5(11): e009758. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMiller JE, Wilenzick M, Ritcey N, et al.: Measuring clinical trial transparency: an empirical analysis of newly approved drugs and large pharmaceutical companies. BMJ Open. 2017; 7(12): e017917. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFDAAA Trialstracker: EBM DataLab. 2018. Reference Source\n\nBlumle A, Meerpohl JJ, Schumacher M, et al.: Fate of clinical research studies after ethical approval--follow-up of study protocols until publication. PLoS One. 2014; 9(2): e87184. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRoss JS, Mulvey GK, Hines EM, et al.: Trial publication after registration in ClinicalTrials.Gov: a cross-sectional analysis. PLoS Med. 2009; 6(9): e1000144. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKasenda B, von Elm E, You J, et al.: Prevalence, characteristics, and publication of discontinued randomized trials. JAMA. 2014; 311(10): 1045–51. PubMed Abstract \| Publisher Full Text\n\nPica N, Bourgeois F: Discontinuation and Nonpublication of Randomized Clinical Trials Conducted in Children. Pediatrics. 2016; 138(3): pii: e20160223. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAnderson ML, Chiswell K, Peterson ED, et al.: Compliance with results reporting at ClinicalTrials.gov. N Engl J Med. 2015; 372(11): 1031–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nProtocol for IntoValue at OSF. http://www.doi.org/10.17605/OSF.IO/FH426\n\nSpelsberg A, Prugger C, Doshi P, et al.: Contribution of industry funded post-marketing studies to drug safety: survey of notifications submitted to regulatory agencies. BMJ. 2017; 356: j337. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPrayle AP, Hurley MN, Smyth AR: Compliance with mandatory reporting of clinical trial results on ClinicalTrials.gov: cross sectional study. BMJ. 2012; 344: d7373. PubMed Abstract \| Publisher Full Text\n\nGordon D, Taddei-Peters W, Mascette A, et al.: Publication of trials funded by the National Heart, Lung, and Blood Institute. N Engl J Med. 2013; 369(20): 1926–34. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWorld Medical Association: Declaration of Helsinki: Ethical Principles for Medical Research Involving Human Subjects. Fortaleza, 2013. PubMed Abstract \| Publisher Full Text\n\nDe Angelis C, Drazen JM, Frizelle FA, et al.: Clinical trial registration: a statement from the International Committee of Medical Journal Editors. N Engl J Med. 2004; 351(12): 1250–1. PubMed Abstract \| Publisher Full Text\n\nvan den Bogert CA, Souverein PC, Brekelmans CT, et al.: Non-Publication Is Common among Phase 1, Single-Center, Not Prospectively Registered, or Early Terminated Clinical Drug Trials. PLoS One. 2016; 11(12): e0167709. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHarriman SL, Patel J: When are clinical trials registered? An analysis of prospective versus retrospective registration. Trials. 2016; 17: 187. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLoder E, Loder S, Cook S: Characteristics and publication fate of unregistered and retrospectively registered clinical trials submitted to The BMJ over 4 years. BMJ Open. 2018; 8(2): e020037. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPowell-Smith A, Goldacre B: The TrialsTracker: Automated ongoing monitoring of failure to share clinical trial results by all major companies and research institutions [version 1; referees: 2 approved]. F1000Res. 2016; 5: 2629. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShamliyan T, Kane RL: Clinical research involving children: registration, completeness, and publication. Pediatrics. 2012; 129(5): e1291–300. PubMed Abstract \| Publisher Full Text\n\nTse T, Williams RJ, Zarin DA: Reporting \"basic results\" in ClinicalTrials.gov. Chest. 2009; 136(1): 295–303. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEuropean Commission (2012/C 302/03): Commission Guideline — Guidance on posting and publication of result-related information on clinical trials in relation to the implementation of Article 57(2) of Regulation (EC) No 726/2004 and Article 41(2) of Regulation (EC) No 1901/2006. 2012. Reference Source\n\nUS National Library of Medicine: ClinicalTrials.gov Results Data Element Definitions for Interventional and Observational Studies. 2018. Reference Source\n\nSchulz KF, Altman DG, Moher D, et al.: CONSORT 2010 statement: updated guidelines for reporting parallel group randomised trials. PLoS Med. 2010; 7(3): e1000251. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEMA: ICH Topic E 3: Structure and Content of Clinical Study Reports. 1996. Reference Source\n\nKasenda B, Schandelmaier S, Sun X, et al.: Subgroup analyses in randomised controlled trials: cohort study on trial protocols and journal publications. BMJ. 2014; 349: g4539. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNIH: Clinical Trial Registry Numbers in MEDLINE/PubMed Records. Reference Source\n\nJacobs A: The trials tracker and post-truth politics. 2016. Reference Source\n\nAltman DG, Furberg CD, Grimshaw JM, et al.: Linked publications from a single trial: a thread of evidence. Trials. 2014; 15: 369. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChalmers I, Altman DG: How can medical journals help prevent poor medical research? Some opportunities presented by electronic publishing. Lancet. 1999; 353(9151): 490–3. PubMed Abstract \| Publisher Full Text" }	[ { "id": "44341", "date": "21 Feb 2019", "name": "Christiane Pauli-Magnus", "expertise": [ "Reviewer Expertise Methodological Research/ Research on Research in the Areas of clinical Research Quality (e.g. Monitoring", "Trial Registration)" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe paper by Strech et al. on tracking the timely dissemination and publication of clinical study results addresses a very relevant topic especially in academic clinical research. Since the metrics on published trials per institution/organisation are easily available through public databases such as the ‘trialstracker’ project, the ethical and scientific dimension of incomplete publication of clinical study results is gaining increasing public attention. Consequently, universities and university hospitals are asked to define measures on how to track and improve publication rates.\nBased on a literature review and an exploratory tracking of registered clinical trials conducted at three leading German university medical centres, the authors examine and discuss ten different tracking variables for clinical study publications. They finally come up with a set of recommendations on how academic institutions and private companies could increase trial publication.\nIn my opinion, the work is clearly presented and cites the relevant literature. It uses appropriate methods and gives sufficient detail to be replicated by others. However, the detailed results of the review of follow-up studies is not given in the paper as it is planned to be published elsewhere.\nWhile I agree with the overall methodological approach and the results, I do not agree with the entire conclusion:\nResponsible party for clinical studies: according to ICH-GCP, the overall responsibility for a clinical study including the publication lies with the study sponsor. Institutional follow-up on studies where the overall responsibility is with a third party (e.g. with a commercial sponsor or another academic institution in case of academic multicentre trials) will increase the bureaucratic burden of the institution without increasing publication rates. I would rather enforce that academic institutions establish policies to follow-up on the trials where they have sponsor-investigator responsibilities.\n\nSource of sampling: Sampling studies at the IRB level will not resolve the problem of incomplete/inaccurate sampling. For instance, a positive IRB vote does not necessarily mean that a study has been started and IRB data on study discontinuation are probably as incomplete as registration data in public registries. Furthermore, depending on national legislation, IRB data are not publically accessible. I therefore would strongly support to rather invest in institutional policies that support mandatory study registration in a public study registry.\n\nTimeliness of dissemination: I disagree that a 24-month period after CD should be enough time to publish even for busy investigators. This might be the case for studies, which are rapidly accepted for publication. In the case of studies with e.g. negative results, or results from discontinued/incomplete studies, substantial delay can derive from submission and resubmission processes. The clear definition of ‘timely’ therefore needs a publication platform that is equally accessible to all study sponsors, independent of study ‘success’.\nOn the other hand, I strongly support the overall recommendations that registration and publications activities should be better linked. I also think that registration and publication should be tracked for all prospective interventional AND observational clinical research and should include publication tracking for discontinued/incomplete studies. I like the idea that academic institutions should develop policies and incentives to improve registration and publication. By supporting clinical researchers during registration as part of a mandatory policy established at the University Hospital Basel in 2018, we could double the number of studies registered at clinicaltriasl.gov within one year. These are relatively easy measures with substantial local impact. However, on a national/international scale I am convinced that only a multi-stakeholder approach, including e.g. academic institutions, funders, IRBs and publishers will improve traceability and publication of study results.\n.", "responses": [] }, { "id": "47315", "date": "10 May 2019", "name": "Shelly Pranić", "expertise": [ "Reviewer Expertise Clinical trial reporting", "public health", "research assessment", "and peer review." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI agree with Reviewer 1 in that this study by Strech et al. provides a comprehensive list of variables to take into account when tracking the publication of clinical trial data. Now, with their applicable checklist, researchers can fine-tune their tracking of the data from trials. The paper is well-written, cites relevant literature, and describes the chosen variables well; however, since a part of this research involves a qualitative portion, it would be informative to describe the 'explorative review' clearer. Additionally, point 7 could be expanded. Please find comments below for your review, thank you.\nMethods:\nRegarding the 'explorative review,' it would be interesting to know what was the initial list of tracking variables compiled by the authors before arriving at 10. The cited methodology on OSF does not describe this step. It is unclear how the authors voted or decided on the tracking variables. This explorative review is qualitative in nature and the authors should describe the way they chose the variables with greater clarity so that this part of the study could be replicated by other researchers.\nDiscussion:\nTimeliness of the publication: I agree that researchers have limited time to publish, but as these tracking variables are also intended for assessing timeliness of companies' publications (including pharmaceuticals), the authors could also mention that there could be embargoes set forth by pharmaceutical industry officials and how embargoes could play a role in the timeliness of the publication of trial data irrespective of authors' available time to publish. I think this is a point worth mentioning in this section of the Discussion.", "responses": [] }, { "id": "47054", "date": "13 May 2019", "name": "Rabia Bashir", "expertise": [ "Reviewer Expertise Clinical Research Informatics and Software Engineering" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis manuscript presents a study based on 10 variables influencing the tracking of clinical studies and timely dissemination of their published results. The selection of these tracking variables is driven by expert knowledge and discussion among authors. To validate the completeness of these tracking variables, the authors used a sample of 34 studies that followed up on the publication of registered clinical trials of three leading German university medical centres (UMC). Based on a review of these studies, the authors make several recommendations. These include strengthening links between registry entry and publications and coordination among academic institutions and private companies to ensure the registration of clinical studies.\nThe study addresses an interesting topic by listing a range of factors that influence the tracking and dissemination of clinical evidence. The paper is generally well written and structured. It would be interesting to see the detailed results; which the authors are aiming to publish elsewhere.\nI liked the study idea and agree with most of the conclusions that the authors have drawn and the recommendations that the authors have made. I agree that observational studies need to be registered because they can be used to influence policy and provide basis for clinical decision making. Also, the finding that up to 75% of published results can be linked to registry entries by adding manual searches support the conclusions made in another systematic review (Bashir et al., 20171).\nHowever, I have a few minor suggestions for strengthening the research:\nStudies are likely to be a biased representation because they were selected from leading German university medical centres for selecting the studies. They find that 35% are published at 24 months after completion date. It is not clear how these results might generalise so it is important that the conclusions do not over-generalise.\n\nThe standard for “timely publication” was defined to be 24 months after the completion date. Using this as a standard for what counts as publication is arbitrary and while there are some details on mandates and policies the standard could have been better justified by the authors. A possible alternative might be to consider a model that estimates the median time to publication using a range of characteristics as factors, and avoid needing to choose a specific amount of time to judge what is timely and what is not.\n\nI agree with the authors’ recommendation that academic institutions and private companies need to play a role to improve the registration of clinical trials and their links with published results. However, journals are also often responsible for ensuring that trial registration information is provided. First, by following ICMJE guidelines on prospective registration. Second, they can ensure the identifier is included in the abstract and metadata they send to bibliographic databases. This will help to ensure that journals do not publish any clinical trials where trial identifier is not included in paper by authors or trial investigators.\n\nThe manuscript can get benefit from these references: Trinquart et al. (20182) and Bashir et al. (20171).", "responses": [] } ]	1	https://f1000research.com/articles/7-1863
https://f1000research.com/articles/7-1482/v1	18 Sep 18	{ "type": "Case Report", "title": "Case Report: Anesthetic management for Cesarean section in a parturient with unspecified inherited bleeding disorder", "authors": [ "Li Li", "Jill M. Johnsen", "Chau X. Doan", "Laurent A. Bollag", "Jill M. Johnsen", "Chau X. Doan", "Laurent A. Bollag" ], "abstract": "Neuraxial anesthesia, as the standard of care for Cesarean deliveries, is associated with decreased blood loss. However, parturients with inherited bleeding disorders are at increased risk for epidural hematomas. A small retrospective study has shown that parturients with known factor deficiencies can safely undergo neuraxial anesthesia once the specific factors are replenished. We present a patient who had a considerably increased risk of peripartum bleeding from an unspecified inherited bleeding disorder and was provided a successful neuraxial anesthetic without complications. We discuss the multidisciplinary approach among the surgeons, anesthesiologists, hematologist, and nursing staff to maximize patient safety and comfort.", "keywords": [ "Anesthesia", "Spinal", "Cesarean section", "Parturient", "Bleeding disorder" ], "content": "Introduction\n\nNeuraxial anesthesia, which encompasses both spinal and epidural anesthesia, has been established as a safe and effective regional anesthetic technique for the obstetric patient. Although considered the standard of care for Cesarean deliveries (CDs), this technique carries a small but potentially devastating risk of neuraxial bleeding complications (e.g. spinal epidural hematomas (SEH)) that may result in significant neurological injury or compromise. The risk of SEH in obstetric patients from neuraxial anesthesia is estimated to be 1:168,0001. Parturients with coagulopathy or bleeding diathesis are at increased risk of neuraxial hematomas2,3. Although a recent study has provided a management outline for patients with unclassified bleeding disorders undergoing procedures4, no specific recommendations were made for CDs, which is unique in that hematologic management directly impacts anesthetic management. In fact, although parturients with known hemostatic deficiencies can often safely undergo neuraxial anesthesia once the hemostatic defects are corrected5–7, due to the lack of recommendations, neuraxial anesthesia is often avoided in these patients, as in patients with other coagulopathies8. Here, we present the anesthetic management and safe utilization of neuraxial anesthesia for a CD in a patient with an unspecified inherited bleeding disorder.\n\n\nCase description\n\nA 39-year-old G6P3 with an unspecified inherited bleeding disorder presented at 37-6/7 weeks gestational age for a fourth repeat scheduled CD in the setting of fetal macrosomia. The patient had a diagnosis of an unspecified bleeding disorder. Prior to diagnosis, her bleeding history consisted of easy bruising, gingival bleeding, heavy menstrual bleeding, and multiple post-operative hemorrhagic complications after previous surgeries. These complications included hemorrhagic compartment syndrome after an ankle surgery and bleeding complications with all three prior deliveries including persistent vaginal bleeding after her first two CDs and intra-abdominal hemorrhage after her third CD. Although she reported having had epidurals during her prior pregnancies, these anesthetic records were unavailable. Her family history is positive for mucocutaneous bleeding including maternal heavy menstrual bleeding and death of her maternal grandmother secondary to postpartum hemorrhage. Prior to this pregnancy, she was referred for evaluation of a bleeding disorder in advance of cervical spine surgery. Her extensive coagulation laboratory evaluation was unremarkable, including normal complete blood count and smear, activated partial thromboplastin time, prothrombin time, thrombin time, von Willebrand factor (VWF) parameters, fibrinogen activity, platelet aggregation studies (including ristocetin-induced platelet aggregation), platelet function assay, factor XIII level, and rotational thromboelastography. There was good documentation of her prior bleeding and no suspicion for other disorders, such as endocrine or connective tissue disorders. She was diagnosed with an unspecified inherited bleeding disorder. The differential diagnosis for her hemostatic defect included rare congenital bleeding disorders such as undetected VWF qualitative dysfunction or undetected defects in fibrin, fibrinolysis, or platelet function. She received prophylactic fresh frozen plasma (FFP), cryoprecipitate, platelets, and anti-fibrinolytic treatment as prophylaxis for her cervical spine surgery and achieved good hemostasis without complication.\n\nUpon presentation for delivery, her laboratory values were unremarkable: hematocrit (Hct) 30%, platelets 169 × 103/ml, international normalized ratio (INR) 1.1, partial thromboplastin time (PTT) 30 s, fibrinogen activity 461 mg/dl, and a thromboelastogram (TEG) within normal parameters. Prior to her delivery, a multi-disciplinary care plan, including hematology, anesthesiology, and obstetric services was established and a delivery plan that balanced patient safety and birthing preferences agreed upon; contingency plans for transfusion and anatomic control of bleeding were in place in the event of obstetric hemorrhage. The risks and benefits of regional anesthesia were discussed at length with the patient, acknowledging the diagnostic challenges and lack of a useful laboratory monitoring test.\n\nThe patient prophylactically received 2 units of FFP, 10 units of cryoprecipitate within 4 hours prior to CD, and 2 units of platelets immediately before the placement of spinal anesthetic. Following the delivery of a healthy infant, intravenous aminocaproic acid was immediately started and routine oxytocin was infused. Poor uterine tone was noted after delivery, which improved after single dose of 200 ug IM methylergonovine and 800 mg buccal misoprostol administration. The estimated blood loss was 1.5 liters. In response to her higher-than-average blood loss, 2 additional units of platelets were given in the immediate postoperative period. Deep venous thrombosis prophylaxis consisted of sequential compression devices until ambulation; low molecular weight heparin was avoided. The patient was closely monitored. Her postoperative labs were notable for Hct 22%, platelets 175 x 103/ml, INR 1.1, PTT 28 s, and fibrinogen activity 462 mg/dl. Additionally, TEG results during and after the CD remained normal. The patient continued intravenous aminocaproic acid therapy postpartum, and transitioned to oral tranexamic acid for 5 days. She had an uneventful recovery and was discharged 3 days after surgery. The case management timeline and dosing regimen are shown in Figure 1.\n\nHematologic interventions for the planned Cesarean section are shown on top, whereas the intraoperative interventions to reduce bleeding risks are shown on the bottom of the arrow. Case events are shown in red.\n\n\nDiscussion\n\nThis patient’s unspecified inherited bleeding disorder placed her at considerably increased risk of peripartum bleeding with risks compounded by fetal macrosomia, multiple prior CDs, and a personal history of obstetric hemorrhage in the absence of prophylaxis. Typically, for a well-defined bleeding disorder prophylaxis can be used with appropriate repletion of the respective pathway components. For patients with unspecified bleeding disorders, it is not possible to know if the hemostatic defect has been adequately corrected by the prophylaxis treatment options available. However, in this case her spine surgery history could inform a prophylaxis regimen; good communication and detailed hematology recommendations regarding the type, dose, and timing of blood product administration were critical in the planning and delivery of a safe spinal anesthetic for this patient.\n\nDespite the potential for prolonged duration of surgery in the setting of a fourth CD, a spinal anesthetic was preferred over a combined spinal epidural to minimize the risk of an epidural catheter-associated hematoma9,10. General anesthesia, even with comparable safety in obstetrics11, was reserved for a failed spinal anesthesia or bleeding emergency to optimize maternal uterine tone12 and fetal neurological adaptation, both of which are affected by the use of volatile anesthetics, and to allow for early maternal-infant skin-to-skin bonding, honoring the mothers wish. Also important, the patient desired a regional anesthetic.\n\nOther management considerations to guide appropriate planning and reduce the risk of significant perioperative bleeding include: early pre-anesthetic consultation, pre-operative hematology consultation, expeditious use of intraoperative uterotonics, and maintenance of normothermia to prevent hypothermic coagulopathy. Co-loading with crystalloid or colloid prior to initiation of neuraxial anesthesia can attenuate hypotension—commonly observed with spinal-related sympathectomy—which may be exacerbated in the setting of acute hemorrhage. The use of non-steroidal anti-inflammatory drugs was avoided.\n\nSince laboratory results, including TEG, did not reflect the patient’s hemostasis defect, close intraoperative monitoring of bleeding status was performed by continually evaluating vital signs, examining blood loss in the operative field and suction canisters, and closely communicating with the obstetricians about uterine tone. The above-average blood loss was attributed to uterine atony due to observation by the surgeons, though incomplete hemostasis could not be ruled out completely. Postoperatively, close neurological exams for 24 hours to monitor for clinical signs of neuraxial hematoma were performed.\n\nIn summary, this case report highlights the importance of a multi-disciplinary approach. Close communication and coordination between the hematology, obstetric, and obstetric anesthesia services were crucial in achieving an uneventful CD in this patient with an otherwise extremely high bleeding risk.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\n\nConsent\n\nWritten informed consent for publication of their clinical details and/or clinical images was obtained from the patient.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nRuppen W, Derry S, McQuay H, et al.: Incidence of epidural hematoma, infection, and neurologic injury in obstetric patients with epidural analgesia/anesthesia. Anesthesiology. 2006; 105(2): 394–9. PubMed Abstract \| Publisher Full Text\n\nMoen V, Dahlgren N, Irestedt L: Severe neurological complications after central neuraxial blockades in Sweden 1990-1999. Anesthesiology. 2004; 101(4): 950–9. PubMed Abstract \| Publisher Full Text\n\nGulur P, Tsui B, Pathak R, et al.: Retrospective analysis of the incidence of epidural haematoma in patients with epidural catheters and abnormal coagulation parameters. Br J Anaesth. 2015; 114(5): 808–11. PubMed Abstract \| Publisher Full Text\n\nObaji S, Alikhan R, Rayment R, et al.: Unclassified bleeding disorders: outcome of haemostatic challenges following tranexamic acid and/or desmopressin. Haemophilia. 2016; 22(2): 285–291. PubMed Abstract \| Publisher Full Text\n\nChi C, Lee CA, England A, et al.: Obstetric analgesia and anaesthesia in women with inherited bleeding disorders. Thromb Haemost. 2009; 101(6): 1104–11. PubMed Abstract \| Publisher Full Text\n\nKatz D, Beilin Y: Disorders of coagulation in pregnancy. Br J Anaesth. 2015; 115 Suppl 2: ii75–88. PubMed Abstract \| Publisher Full Text\n\nGonzalez-Fiol A, Eisenberger A: Anesthesia implications of coagulation and anticoagulation during pregnancy. Semin Perinatol. 2014; 38(6): 370–7. PubMed Abstract \| Publisher Full Text\n\nDemers C, Derzko C, David M, et al.: No. 163-Gynaecological and Obstetric Management of Women With Inherited Bleeding Disorders. J Obstet Gynaecol Can. 2018; 40(2): e91–e103. PubMed Abstract \| Publisher Full Text\n\nPumberger M, Memtsoudis SG, Stundner O, et al.: An analysis of the safety of epidural and spinal neuraxial anesthesia in more than 100,000 consecutive major lower extremity joint replacements. Reg Anesth Pain Med. 2013; 38(6): 515–9. PubMed Abstract \| Publisher Full Text\n\nVandermeulen EP, Van Aken H, Vermylen J: Anticoagulants and spinal-epidural anesthesia. Anesth Analg. 1994; 79(6): 1165–77. PubMed Abstract\n\nHawkins JL, Chang J, Palmer SK, et al.: Anesthesia-related maternal mortality in the United States: 1979-2002. Obstet Gynecol. 2011; 117(1): 69–74. PubMed Abstract \| Publisher Full Text\n\nYoo KY, Lee JC, Yoon MH, et al.: The effects of volatile anesthetics on spontaneous contractility of isolated human pregnant uterine muscle: a comparison among sevoflurane, desflurane, isoflurane, and halothane. Anesth Analg. 2006; 103(2): 443–7, table of contents. PubMed Abstract \| Publisher Full Text" }	[ { "id": "38948", "date": "04 Oct 2018", "name": "Allison J. Lee", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a description of management of a challenging clinical scenario given the non-specific diagnosis of the bleeding disorder in this patient with a desire for neuraxial anesthesia.\nI was curious about the airway exam, considering that the team would have to prepare the possibility of inducing general anesthesia. I wonder if consideration was given to administering DDAVP or other treatment for von Willebrand's disease. Was an arterial line placed or considered, and what was the intravenous access in place?\nI am curious about the components of the intrathecal medications for the single shot spinal anesthetic. With the possibility for prolonged surgery, were any adjuvants, such as clonidine added to prolong the anesthetic?\nI would suggest editing the language in a few places. Instead of \"epidurals\", suggest instead, \"labor epidural analgesia\". Explain what is\"routine oxytocin\" in your institution. For what period postoperatively was the patient \"closely monitored\" and how? Were there nursing orders for frequent checks of neurologic status? Finally, I would suggest discussing the reasons for the low incidence of epidural hematoma in the obstetric population and including this reference: Lee L, et al1.\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Partly\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Partly\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes", "responses": [ { "c_id": "4179", "date": "06 Nov 2018", "name": "Laurent Bollag", "role": "Author Response", "response": "Dear Dr LeeThank you for taking the time to review our case report to improve its quality!I addressed your comments point by point below:I was curious about the airway exam, considering that the team would have to prepare the possibility of inducing general anesthesia. I wonder if consideration was given to administering DDAVP or other treatment for von Willebrand's disease. Was an arterial line placed or considered, and what was the intravenous access in place?Her airway exam was as follows: Mallampati Score 2, free head and neck movement (>6cm), normal mouth opening and no prominent incisors, in short pretty normal.Willebrand factor/DDAVP administration was not recommended by the hematologist based on the normal von Willebrand factor (VWF) parameters preoperatively.She had a well working 18g IV. We decided not to place an Aline a priori, because we felt that it could be sited easily if need be. In cases where the placement is deemed technically complicated we do place A lines when a PPH is deemed likely, the same applies to a second large bore IV.I am curious about the components of the intrathecal medications for the single shot spinal anesthetic. With the possibility for prolonged surgery, were any adjuvants, such as clonidine added to prolong the anesthetic?We used out routine SPA mixture 12,5mg hyperbaric Bupivacaine, 10mcg Fentanyl and 100mcg PF Morphine. We did not add spinal clonidine, to avoid worsening of hypotension in a case where a PPH was deemed likely. Clonidine could also be aded later IV, to extend spinal anesthesia duration, but again potentially worsen hypotension- for hours.Obviously, we opted not to place a CSE in this potentially coagulopathic patient. Explain what is\"routine oxytocin\" in your institution. For what period postoperatively was the patient \"closely monitored\" and how? Were there nursing orders for frequent checks of neurologic status? Finally, I would suggest discussing the reasons for the low incidence of epidural hematoma in the obstetric population and including this reference: Lee L, et al1Physiological hypercoagulability of pregnancy as well as the large epidural space in young parturients may explain why epidural hematomas in parturients are rare; additionally older guidelines are more conservative regarding platelet count cut-offs for neuraxial anesthesia.The risk of performing neuraxial anesthesia in patients with platelet counts of less than 70,000 however remains unknown.1We have addressed your other comments in our response to reviewer 1, who raised the same questions.References1. Lee LO, Bateman BT, Kheterpal S, Klumpner TT, Housey M, Aziz MF, Hand KW, MacEachern M, Goodier CG, Bernstein J, Bauer ME, Multicenter Perioperative Outcomes Group Investigators: Risk of Epidural Hematoma after Neuraxial Techniques in Thrombocytopenic Parturients: A Report from the Multicenter Perioperative Outcomes Group.Anesthesiology. 126 (6): 1053-1063 PubMed Abstract \| Publisher Full Text" } ] }, { "id": "38949", "date": "17 Oct 2018", "name": "Benjamin G. Cobb", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors describe the use of regional anesthesia for cesarean section in a parturient with an unspecified inherited bleeding disorder. In this case, the clinical bleeding history in conjunction with the hematologic lab workup leading to the diagnosis of an ‘unspecified’ bleeding disorder is well described. The authors highlight both the importance of early multidisciplinary collaboration in the parturient with a bleeding disorder in an effort to minimize hemorrhagic complications. In addition, the role of informed consent for a patient with an unspecified bleeding disorder and unquantifiable risk of spinal/epidural hematoma should not be understated. Additional attention to the potential limitations of factor/platelet/fibrinogen replacement in the setting of qualitative coagulation defects to prevent hemorrhagic complications of neuraxial anesthesia may be warranted in the discussion. I commend the authors for describing the often challenging anesthetic risk-benefit assessment of a patient with an unspecified bleeding disorders.\n\nIn addition to the cryoprecipitate transfusion, was von Willebrand factor/DDAVP administration considered given the unknown defect? Please clarify the institutional oxytocin protocol post-delivery. Following the postpartum hemorrhage, what guided the decision to transfuse two additional units of platelets postoperatively and not other factors/fibrinogen? Is there any published literature on the prevalence of ‘unspecified bleeding disorders’ and/or standardization of these clinical diagnoses in the setting of normal coagulation workups by hematologists? Please clarify post-operative neurologic monitoring plan, including nursing neurological check frequency and patient surveillance counseling (concerning signs/symptoms).\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Partly\n\nIs the case presented with sufficient detail to be useful for other practitioners? Partly", "responses": [ { "c_id": "4178", "date": "06 Nov 2018", "name": "Laurent Bollag", "role": "Author Response", "response": "Dear Dr. CobbThank you for your comments to improve the quality of our manuscript.I will reply point by point to your comments: In addition to the cryoprecipitate transfusion, was von Willebrand factor/DDAVP administration considered given the unknown defect? Willebrand factor/DDAVP administration was not given based on the norma; von Willebrand factor (VWF) parameters preoperatively. Please clarify the institutional oxytocin protocol post-delivery. Our protocol is as follows: 15 IU per hour given on a drip. Once adequate uterine tone is achieved, the dose is reduced to 7.5 IU per hour, and again, once the tone is reported to be adequate, reduced ca 50% (or 70ml /hr). This drip is continued postpartum. Following the postpartum hemorrhage, what guided the decision to transfuse two additional units of platelets postoperatively and not other factors/fibrinogen? While uterine atony was the likely reason for the PPH, incomplete hemostasis we assumed and platelet substitution deemed the most appropriate therapeutic first step. Is there any published literature on the prevalence of ‘unspecified bleeding disorders’ and/or standardization of these clinical diagnoses in the setting of normal coagulation workups by hematologists? The lack of such literature prompted this report. Please clarify post-operative neurologic monitoring plan, including nursing neurological check frequency and patient surveillance counseling (concerning signs/symptoms). Local guidelines include vigilant neurological checks, including motor function of both legs every 4 hours for 24 hours. Patients are informed about the common symptoms of spinal nerve compression and the need to urgently report these to the nursing staff, namely: rapid onset of pain new onset radiculopathy sensory and motor function deficits An emergent lumbar MRI will then confirm the diagnosis." } ] } ]	1	https://f1000research.com/articles/7-1482
https://f1000research.com/articles/7-1859/v1	28 Nov 18	{ "type": "Research Article", "title": "In vitro inhibitory and biofilm disruptive activities of ginger oil against Enterococcus faecalis", "authors": [ "Shahida Mohd-Said", "Wee Wee Kweh", "Chong Yi Than", "Zamirah Zainal-Abidin", "Siti Noor Adnalizawati Adnan", "Safura Anita Baharin", "Eason Soo", "Wee Wee Kweh", "Chong Yi Than", "Zamirah Zainal-Abidin", "Siti Noor Adnalizawati Adnan", "Safura Anita Baharin", "Eason Soo" ], "abstract": "Background: This study investigated the antibacterial effect of ginger (Zingiber officinale) oil against a common resistant root canal pathogen known as Enterococcus faecalis. The aim of the study was to determine the inhibition of E. faecalis growth in culture suspension and ability to inhibit growth of bacteria through disruption of pre-formed monospecies biofilm. Methods: Ginger rhizome oil was prepared in two-fold concentration series from 0.04 to 5.00 mg/mL and mixed with brain heart infusion broth inoculated with E. faecalis in anaerobic condition. Among the antibacterial tests performed were the minimum inhibitory concentrations (MIC) and minimum bactericidal concentrations using microdilution assays, and anti-biofilm assay on 3-day old pre-form monospecies biofilm on a 94 well-plate. Ampicillin was used as a positive control. Results: The result showed an in vitro dose-dependent bacteriostatic activity towards E. faecalis in suspension broth (MIC 0.04mg/mL) but no bactericidal activity within the tested concentration range. It was also found that the ginger oil inhibitory activity against E. faecalis was comparably less in anti-biofilm activity than against bacteria cultured in suspension solution. Conclusion: The study suggests that at determined concentrations, ginger oil has the potential to be used as an antibacterial agent in the management of root canal infections particularly where newly formed E. faecalis is involved.", "keywords": [ "Enterococcus", "biofilm", "endodontics", "Zingiber officinale" ], "content": "Introduction\n\nEnterococcus faecalis is an opportunistic facultative anaerobe that is well recognised as an oral pathogen associated with persistent apical periodontitis and is highly prevalent in failed root filled teeth1. The ability of E. faecalis to enter and survive in root canals, as well as its resistance to root canal medicaments, makes it one of the toughest pathogens to eradicate in endodontics2–7. It has been reported that this pathogen is associated with a high percentage of endodontic failures, which is about one third of the canals of root-filled teeth with persistent periapical lesions8. Inevitably, repeated use of calcium hydroxide and sodium hypochlorite as the two most commonly used root canal medicaments and irrigation solution, respectively, has been said to allow E. faecalis to adapt to the sub-lethal environment9.\n\nWithin the past decade, herbal medicine has begun to offer some beneficial antibacterial activities against oral pathogens10,11. More recently, studies also found the potential antibacterial properties of ginger on oral pathogens, including E. faecalis12–16. In most instances, the effects of ginger extracts were studied on bacteria cultured in vitro and in cell suspensions, using calcium hydroxide or sodium hypochlorite as a control. Limited data is available on the activity of ginger oil in disrupting established E. faecalis biofilm and its comparable effect to antibiotics. Hence, the aim of our study was to explore the in vitro potential of ginger oil as an antibacterial agent against E. faecalis cultured in suspension and biofilm in comparison to ampicillin, as a common antibiotic used for oral infection.\n\n\nMethods\n\nGinger oil was prepared from Zingiber officinale Roscoe rhizomes after fresh young ginger were finely sliced, dried and boiled in distilled water for 8h17. The oil stock solution (500 mg/mL) was prepared by mixing 100 µL ginger oil with 200 µL dimethyl sulfoxide (DMSO; Sigma-Aldrich, St Louis, MO USA) in an Eppendorf tube. Next, 100 µL of the oil stock solution was pipetted and mixed with 100 µL DMSO in a second tube to produce two-fold dilution at 250 mg/mL. This procedure was repeated six times to produce an eight concentration series of ginger oil solution at 500, 250, 125, 62.5, 31.3, 15.6, 7.8, 3.9 mg/mL respectively. The solutions were vortexed each time after mixing to ensure a thorough mixture were produced. Following this, 20 µL of each oil series were transferred into a 980 µL brain heart infusion (BHI; Oxoid Ltd., Cheshire UK) broth in 2.5 mL universal bottles to produce eight working concentration series of ginger oil mixtures in broth microdilution test at 10, 5, 2.5, 1.25, 0.63, 0.31, 0.16 and 0.08 mg/mL respectively. The positive control ampicillin solution was prepared by mixing 500 mg ampicillin powder (Sigma-Aldrich) with 1 mL distilled water, and then eight series of two-fold solutions were prepared similarly to the technique mentioned above.\n\nEnterococcus faecalis ATCC29212 (American Type Culture Collection, Virginia USA) used in this study was cultured in BHI (Oxoid Ltd., Cheshire UK) agar plates and regularly maintained in a 37°C incubator with 5% CO2. Identification and purification of E. faecalis were done through Gram stain test as well as morphology and colony identification. The quantity and viable bacterial number of E. faecalis were determined by colony forming unit (CFU) and standardized with McFarland 0.5 turbidity (1.5 x 108 cfu/mL) prior to the antibacterial assays. Cultures and broths were constantly checked for sterility and contamination. DMSO was used as solvent for ginger oil and ampicillin as the positive control.\n\nResearch and ethics approval was obtained from the Universiti Kebangsaan Malaysia Research and Ethics Committee (UKM 1.5.3.5/244/DD/2014/017) for methodology and dissemination of findings.\n\nAn antibacterial assay using the broth microdilution technique was done based on modified NCCLS guideline18. A 96-well test plate was divided into two sections where half was inoculated with 50 µL E. faecalis suspension in each well and another section was not inoculated (50 µL broth/well only), as shown in Figure 1, plate rows from A to H. Sample solution series prepared earlier were mixed with the bacterial suspensions in wells and produce a final concentration series of oil mixtures at final concentrations of 5, 2.5, 1.25, 0.63, 0.31, 0.16, 0.08 and 0.04 mg/mL respectively19. One hundred microliters sample series (ginger oil mixture or ampicillin solution) were dispensed into the prepared 96-well plate. Triplicates were done for each well and test plate. Plates were then incubated at 37°C for 24h. After this, the turbidity of wells on tested plates was measured at 595nm optical density using a microtiter plate reader (Thermo Scientific VARIOSKAN Flash, UK). Bacterial growth inhibition percentage was calculated using the following formula:\n\n\n\nThe reading generated by the device were analysed and the Minimum Inhibitory Concentration (MIC), defined as the lowest concentration of sample that inhibits visible growth of a microorganism after overnight incubation, was determined. Further to this, 20 µL of the mixture from each well that showed bacterial growth inhibition were cultured on BHI agar plates for 24h at 37°C in order to determine whether the inhibition is bacteriostatic or bactericidal in nature. The minimum bactericidal concentration (MBC), defined as the lowest concentration of samples that prevent the total growth of E. faecalis on test plates20, was determined.\n\nA monospecies biofilm was initially produced by inoculating 200 µL E. faecalis suspension culture at 1.5 x 108 cfu/mL cell density with BHI broth in 96-well plates (modified from Azizan et al19). The culture was checked daily for contamination for 4 days and cell density was measured at 595 nm optical density using a microtiter plate reader, and repeated in three separate tests. The findings from this monospecies biofilm development test was charted (Figure 2) and used to determine the age for a stable pre-formed biofilm for the anti-biofilm assay.\n\nIn the anti-biofilm assay, E. faecalis biofilm was developed and maintained for 3 days. On the third day, culture broth in the plate was pipetted out in slanting position to prevent disruption of biofilm and replaced with 100 µL of fresh broth. Then 100 µL ginger oil or ampicillin solutions at eight respective concentrations were dispensed into the wells and plates were incubated at 37°C for 24h. The turbidity of the bacterial cultures was measured using a microtiter plate reader at 595 nm optical density.\n\nEach assay was conducted in triplicate and two independent experiments were done. The data collected were analysed with SPSS 21.0. Non-parametric Mann-Whitney test were used with significant level set as 0.05. Results were accepted as statistically significant if the p value was <0.05.\n\n\nResults\n\nThe broth microdilution antibacterial assay showed a dose-dependent inhibition of E. faecalis growth when exposed to all tested concentration series of ginger oil solution (Figure 3), but no bactericidal effect of the oil was found within the range tested (0.04 – 5.0 mg/mL). While ampicillin showed almost 80% inhibition for all concentrations tested, ginger oil showed about 76% inhibition at 5.0 mg/mL and 50% and less at 2.50 – 0.04 mg/mL. The difference in means of inhibition between all ginger oil concentrations tested and between ginger oil and ampicillin was found to be statistically significant (p=0.002 and p=0.007, respectively).\n\n* significant difference (p<0.05) between ginger oil and ampicillin at 5.00mg/mL.\n\nIn the anti-biofilm test, ginger oil also inhibited the growth of pre-formed biofilm (Figure 4). However, it was found that both ginger oil and ampicillin showed lesser inhibitory effect on pre-formed biofilm than on suspended cells in broth (Figure 3). In this test, the anti-biofilm activity of ginger oil was observed as an increase between 1.25 to 5 mg/mL, yet the activity failed to achieve an acceptable level of inhibition of more than 50%. At 1.25 mg/mL, ginger oil was seen to have somewhat equal biofilm disruptive property as ampicillin, but the means of anti-biofilm activity between the two agents showed no statistically significant difference (p=0.052).\n\n\nDiscussion\n\nIt has been advocated that ginger, as the stem of Z. officinale plant, has strong antimicrobial properties including antibacterial and antifungal against pathogens, including bacteria from the oral cavity21–27. In the present study, it was further discovered that ginger rhizome oil exhibited bacteriostatic activity on E. faecalis cultured in plates as monospecies suspension and causes disruption of the pre-formed biofilm. The dose-dependent effect of the oil was accepted as a positive outcome of the study and we predict that with higher concentration (more than 5 mg/mL), ginger oil would have bactericidal effect on E. faecalis.\n\nThe ability of ginger oil to disrupt 3-day old pre-formed monospecies E. faecalis biofilm provided us with a new insight on the antibacterial property of this herbal oil. Although the activity was not as potent as the effect on suspension culture, ginger oil still produced acceptable inhibitory activities against the pathogen in vitro. This study provides clearer appreciation of the E. faecalis resistance to antibacterial agents when they are in biofilm form and conforms to other evidence of biofilm resistance28–30.\n\nThe use of systemic antibiotics such as ampicillin in acute endodontic infections has raised many concerns over the increase in the prevalence of bacterial resistance in dentistry31–33. On the other hand, the current use of antibacterial irrigation solutions such as sodium hypochlorite, chlorhexidine and iodine as well as calcium hydroxide as canal medicament reduces the needs for systemic antibiotics. However, some degree of toxicity towards vital tissues and allergy reactions to some patients have been reported34–36. Through our study, we found that ginger oil has lower but comparable antibacterial efficacy against E. faecalis compared to ampicillin. Further studies to investigate the effects of combined oil-antibiotics may perhaps provide some useful information on the range of its antibacterial activities. Studies have shown that combination of herbal oils with antibiotics does produce adequate synergistic and additive effects against microorganisms37.\n\nLastly, the results of this study demonstrated that at determined concentrations, ginger oil has the potential to be used as an antibacterial agent in the management of failed root canal therapy. Our recent study on the effect of ginger oil on tooth dentine microhardness also provide promising usage of this herbal oil in future endodontics as it offers comparable changes to dentine structure when used as irrigation solution in contrast to sodium hypochlorite and ethylenediaminetetraacetic acid38. These two findings may help to access infection deep within the root canal system. Nevertheless, the current clinical scenario does not offer immediate and accurate chairside information on the type of bacterial species involved in specific endodontic infection; hence this finding may not be readily useful for purpose as yet. More studies will be required to evaluate the clinical efficacy of the delivery of ginger oil as an antibacterial agent for endodontic infections.\n\n\nConclusion\n\nWithin the concentration tested (0.04 – 5.0 mg/mL), ginger oil showed an acceptable dose-dependent in vitro inhibitory activity against E. faecalis (MIC 0.04mg/mL), while its antibacterial activity towards bacteria in suspension broth was comparatively better than its anti-biofilm activity.\n\n\nData availability\n\nF1000Research: Dataset 1. File containing raw data for biofilm development test, anti-biofilm assay and suspension assay. , https://doi.org/10.5256/f1000research.16851.d22512239", "appendix": "Grant information\n\nThis study was funded by the Universiti Kebangsaan Malaysia Young Researchers’ Incentive Grant (GGPM-2013-096).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nAuthors wished to thank Miss Nurulnuha Kamal Baharin of Faculty of Dentistry UKM Microbiology Lab, for her assistance in this study.\n\n\nReferences\n\nPortenier I, Waltimo TMT, Haapasalo M: Enterococcus faecalis- the root canal survivor and ‘star’ in post-treatment disease. Endod Topics. 2003; 6(1): 135–159. Publisher Full Text\n\nChivatxaranukul P, Dashper SG, Messer HH: Dentinal tubule invasion and adherence by Enterococcus faecalis. Int Endod J. 2008; 41(10): 873–882. PubMed Abstract \| Publisher Full Text\n\nDuggan JM, Sedgley CM: Biofilm formation of oral and endodontic Enterococcus faecalis. J Endod. 2007; 33(7): 815–818. PubMed Abstract \| Publisher Full Text\n\nGajan EB, Aghazadeh M, Abashov R, et al.: Microbial Flora of Root Canals of Pulpally-infected Teeth: Enterococcus faecalis a Prevalent Species. J Dent Res Dent Clin Dent Prospects. 2009; 3(1): 24–27. PubMed Abstract \| Free Full Text\n\nLiu H, Wei X, Ling J, et al.: Biofilm formation capability of Enterococcus faecalis cells in starvation phase and its susceptibility to sodium hypochlorite. J Endod. 2010; 36(4): 630–635. PubMed Abstract \| Publisher Full Text\n\nMohammadi Z, Palazzi F, Giardino L, et al.: Microbial biofilms in endodontic infections: an update review. Biomed J. 2013; 36(2): 59–70. PubMed Abstract \| Publisher Full Text\n\nSvensater G, Bergenholtz G: Biofilms in endodontic infections. Endod Topics. 2004; 9(1): 27–36. Publisher Full Text\n\nRoth KA, Friedman S, Lévesque CM, et al.: Microbial biofilm proliferation within sealer-root dentin interfaces is affected by sealer type and aging period. J Endod. 2012; 38(9): 1253–1256. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEvans M, Davies JK, Sundqvis G, et al.: Mechanisms involved in the resistance of Enterococcus faecalis to calcium hydroxide. Int Endod J. 2002; 35(3): 221–228. PubMed Abstract \| Publisher Full Text\n\nMaekawa LE, Valera MC, Oliveira LD, et al.: Effect of Zingiber officinale and propolis on microorganisms and endotoxins in root canals. J Appl Oral Sci. 2013; 21(1): 25–31. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPanpatil VV, Tattari S, Kota N, et al.: In vitro evaluation on antioxidant and antimicrobial activity of spice extracts of ginger , turmeric and garlic. J Pharmacogn Phytochem. 2013; 2(3): 143–148. Reference Source\n\nAlrazhi BA, Diab AH, Essa SA, et al.: Antibacterial Activity of the Ethanolic Extracts of Allium Sativum L. Bulbs and Zingiber officinale Roscoe Rhizomes as Irrigating Solutions. World J Pharm Pharm Sci. 2014; 3(6): 324–337. Reference Source\n\nAzizi A, Aghayan S, Zaker S, et al.: In Vitro Effect of Zingiber officinale Extract on Growth of Streptococcus mutans and Streptococcus sanguinis. Int J Dent. 2015; 2015: 489842. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGiriraju A, Yunus G: Assessment of antimicrobial potential of 10% ginger extract against Streptococcus mutans, Candida albicans, and Enterococcus faecalis: an in vitro study. Indian J Dent Res. 2013; 24(4): 397–400. PubMed Abstract \| Publisher Full Text\n\nShahsiah S, Moghimipour M, Shamsizadeh N, et al.: Evaluation of antibacterial effect of Eucalyptus and ginger on dentinal tubules Enterococcus faecalis in the presence of smear layer. Bioscience Biotechnology Research Communications. 2017; 10(2): 225–232. Publisher Full Text\n\nValera MC, Oliveira SA, Maekawa LE, et al.: Action of Chlorhexidine, Zingiber officinale, and Calcium Hydroxide on Candida albicans, Enterococcus faecalis, Escherichia coli, and Endotoxin in the Root Canals. J Contemp Dent Pract. 2016; 17(2): 114–118. PubMed Abstract \| Publisher Full Text\n\nZainal-Abidin Z, Abdul-Wahab NA, Ghazi-Ahmad MK, et al.: In vitro Antibacterial Activity of Zingiber officinale and Orthosiphon stamineus on Enterococcus faecalis. J Agric Sci. 2017; 9(13): 112–121. Publisher Full Text\n\nClinical and Laboratory Standards Institute: M11-A8 Methods for Antimicrobial Susceptibility Testing of Anaerobic Bacteria; Approved Standard - Eight Edition. 2012. Reference Source\n\nAzizan N, Mohd Said S, Zainal Abidin Z, et al.: Composition and Antibacterial Activity of the Essential Oils of Orthosiphon stamineus Benth and Ficus deltoidea Jack against Pathogenic Oral Bacteria. Molecules. 2017; 22(12): pii: E2135. PubMed Abstract \| Publisher Full Text\n\nHelal MM, Osman MY, Ghobashy MOI, et al.: Study of some biological activities of aqueous extract of ginger (Zingiber officinale). Egypt Pharmaceut J. 2014; 13(2): 144–150. Publisher Full Text\n\nBhardwaj A, Velmurugan N, Sumitha, et al.: Efficacy of passive ultrasonic irrigation with natural irrigants (Morinda citrifolia juice, Aloe Vera and Propolis) in comparison with 1% sodium hypochlorite for removal of E. faecalis biofilm: An in vitro study. Indian J Dent Res. 2013; 24(1): 35–41. PubMed Abstract \| Publisher Full Text\n\nJantan IB, Yassin MS, Chin CB, et al.: Antifungal Activity of the Essential Oils of Nine Zingiberaceae Species. Pharm Biol. 2003; 41(5): 392–397. Publisher Full Text\n\nMueller M, Hobiger S, Jungbauer A: Anti-inflammatory activity of extracts from fruits, herbs and spices. Food Chem. 2010; 122(4): 987–996. Publisher Full Text\n\nPatel RV, Thaker VT, Patel VK: Antimicrobial activity of ginger and honey on isolates of extracted carious teeth during orthodontic treatment. Asian Pac J Trop Biomed. 2011; 1(1): S58–S61. Publisher Full Text\n\nSasidharan I, Nirmala Menon A: Comparative Chemical Composition and Antimicrobial Activity Fresh & Dry Ginger Oils (Zingiber Officinale Roscoe). Int J Curr Pharm Res. 2010; 2(4): 4–7. Reference Source\n\nSebiomo A, Awofodu AD, Awosanya AO, et al.: Comparative studies of antibacterial effect of some antibiotics and ginger (Zingiber officinale) on two pathogenic bacteria. J Microbiol Antimicrob. 2011; 3(1): 18–22. Reference Source\n\nSivasothy Y, Chong WK, Hamid A, et al.: Essential Oils of Zingiber officinale var. rubrum Theilade and their Antibacterial Activities. Food Chem. 2011; 124(2): 514–517. Publisher Full Text\n\nDavies D: Understanding biofilm resistance to antibacterial agents. Nat Rev Drug Discov. 2003; 2(2): 114–122. PubMed Abstract \| Publisher Full Text\n\nGupta A, Duhan J, Tewari S, et al.: Comparative evaluation of antimicrobial efficacy of Syzygium aromaticum, Ocimum sanctum and Cinnamomum zeylanicum plant extracts against Enterococcus faecalis: A preliminary study. Int Endod J. 2013; 46(8): 775–783. PubMed Abstract \| Publisher Full Text\n\nNeelakantan P, Subbarao C, Sharma S, et al.: Effectiveness of curcumin against enterococcus faecalis biofilm. Acta Odontol Scand. 2013; 71(6): 1453–1457. PubMed Abstract \| Publisher Full Text\n\nGuivarc’h M, Ordioni U, Ahmed HM, et al.: Sodium Hypochlorite Accident: A Systematic Review. J Endod. 2017; 43(1): 16–24. PubMed Abstract \| Publisher Full Text\n\nLevy SB, Marshall B: Antibacterial resistance worldwide: causes, challenges and responses. Nat Med. 2004; 10(12 Suppl): S122–S129. PubMed Abstract \| Publisher Full Text\n\nMatsuura M, Nakazawa H, Hashimoto T, et al.: Combined antibacterial activity of amoxicillin with clavulanic acid against ampicillin-resistant strains. Antimicrob Agents Chemother. 1980; 17(6): 908–911. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWilliamson EM: Synergy and other interactions in phytomedicines. Phytomedicine. 2001; 8(5): 401–409. PubMed Abstract \| Publisher Full Text\n\nYost RA, Bergeron BE, Kirkpatrick TC, et al.: Evaluation of 4 Different Irrigating Systems for Apical Extrusion of Sodium Hypochlorite. J Endod. 2015; 41(9): 1530–1534. PubMed Abstract \| Publisher Full Text\n\nZandi H, Rodrigues RC, Kristoffersen AK, et al.: Antibacterial Effectiveness of 2 Root Canal Irrigants in Root-filled Teeth with Infection: A Randomized Clinical Trial. J Endod. 2016; 42(9): 1307–1313. PubMed Abstract \| Publisher Full Text\n\nZainol SN, Mohd Said S, Zainal Abidin Z, et al.: Synergistic benefit of eugenia caryophyllata L. and cinnamomum zeylanicum blume essential oils against oral pathogenic bacteria. Chem Eng Trans. 2017; 56: 1429–1434. Publisher Full Text\n\nMohd Zaki SK, Baharin SA, Qamaruz Zaman J, et al.: The Effect of Zingiber officinale Roscoe (Ginger) on Dentin Microhardness: An in vitro Study. J Agric Sci. 2017; 9(13): 102–111. Publisher Full Text\n\nMohd-Said S, Kweh WW, Than CY, et al.: Dataset 1 in: In vitro inhibitory and biofilm disruptive activities of ginger oil against Enterococcus faecalis. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16851.d225122" }	[ { "id": "42492", "date": "16 Jan 2019", "name": "Lara Thieme", "expertise": [ "Reviewer Expertise infectious diseases", "clinical microbiology", "antibiotic resistant bacteria", "biofilms", "enterococci" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this study, the authors assessed the inhibitory activities of ginger oil in comparison to ampicillin against Enterococcus faecalis, both against planktonic and biofilm cultures, with the clinical background of analyzing novel treatment ideas for root canal infections caused by E. faecalis. Ginger oil has been shown to exhibit antibacterial effects against E. faecalis in some in vitro studies, but its anti-biofilm activity has not been investigated yet.\n\nAlthough the authors came up with an in principle adequate experimental set-up, the main criticism concerns the fact that only one laboratory strain has been analyzed (ATCC 29212), but no clinical isolates recovered from root canal infections. Clinical isolates usually show a different behavior than laboratory strains concerning antimicrobial susceptibility patterns, making it absolutely necessary to include an adequate number of clinical isolates when drawing conclusion for therapeutic treatments of infections. Further, it remains questionable whether the analyzed concentrations of ginger oil apply as therapeutic dosages. The lowest concentration (range 40 mg/L to 5000 mg/L) analyzed for both ginger oil and ampicillin, the reference substance in this study, is still 40 times higher than the MIC of ampicillin of 1 mg/L for E. faecalis ATCC 29212 1.\n\nUnfortunately, the conclusion drawn from the presented results is in our opinion over-interpreted (mainly due to n=1) and not conclusive from the data presented. As the authors state in the result part, the biofilm inhibitory activity of ginger oil was not strong enough to make it a significant result (quote “activity failed to achieve an acceptable level of inhibition of more than 50%”), contradicting very much the conclusion of the discussion part (quote “Although the activity was not as potent as the effect on suspension culture, ginger oil still produced acceptable inhibitory activities against the pathogen in vitro”). Ampicillin, a common antibiotic for oral infections and the reference substance, showed stronger effects than ginger oil at all concentrations tested in both of the planktonic and biofilm experiments!\n\nRegarding the language, only the sentences in which the meaning was not clear were marked and suggestions for improvement were made. Typos, wrong choices of words and sentence constructions were neglected.\nIntroduction: approved Methods: approved with reservations/not approved due to n=1 Results: approved Discussion: not approved\n\nMajor comments:\nMethods: Please include at least three root canal E. faecalis isolates in your study. Further, please state some data/references about therapeutic doses of ginger oil to better understand the suitability of the tested concentrations for therapeutic treatment. Method anti-biofilm assay: We are struggling with your method of choice (OD measurement) for the establishment of the biofilm growth. Firstly, a measurement of the OD value rather assesses the planktonic but not the biofilm state of the bacteria and further does not discriminate between viable and dead cells. We recommend performing a CFU determination of the detached biofilms at day 1, 2, 3 and 4 of biofilm growth 2 to determine the viable cell number. This will tell you how the OD measurement correlates with cell numbers. Also, you might want to consider changing the medium every 24hr to enhance biofilm formation. Secondly, we do not understand why you have chosen a 3-day old biofilm for your subsequent anti-biofilm assay as the highest OD value is observed on day 1 (see Figure 2). Please explain your choice. For the anti-biofilm assay, please also verify for at least one strain that lower OD values (compared to the control) correlate with lower cell numbers and therefore stronger inhibitory effects of ginger oil. Results Bactericidal effects (MBC): If you plated the samples on agar plates to determine the MBC as described in the method section, it would be interesting to see the CFU numbers to see the correlation between OD values and viable bacteria (see above)! Discussion: Disruption is the wrong word because you were analyzing biofilm growth inhibitory, but not biofilm eradicative effects! Also, “we found that ginger oil has lower but comparable antibacterial efficacy compared to ampicillin” is a very contradicting statement and in terms of your results wrong. Ampicillin showed stronger effects than ginger oil at all concentrations tested in both of the planktonic and biofilm experiments! Conclusion: Why do you state 0.04 mg/mL as the MIC of ginger oil for ATCC 29212? At this concentration, only approximately 5% growth inhibition was observed (Fig. 3).\n\nMinor comments:\nAbstract Methods: You state that E. faecalis was grown in anaerobic conditions but this is not described in the method part of the main paper. For MIC testing according to CLSI guidelines, enterococci are grown aerobically (so at 37°C and 5 % CO2 as you stated correctly in the method section). Abstract methods: Please consider revision of the sentence: “Among the antibacterial tests performed were the minimum inhibitory concentrations (MIC) and minimum bactericidal concentrations using microdilution assays, and anti-biofilm assay on 3-day old pre-form monospecies biofilm on a 94 well-plate.” E.g. “Among the antibacterial tests performed were the minimum inhibitory concentration (MIC) and minimum bactericidal concentration (MBC) microdilution assay as well as an anti-biofilm assay on 3-day old pre-formed biofilms on 96 well-plates.”\n\nFurther, it would be helpful for the reader to clarify which parameter was assessed in your anti-biofilm assay. If I understood your experiments correctly, you assessed the so-called minimum biofilm inhibitory concentration (MBIC 3) in contrast to biofilm eradicative effects (aka MBEC). Abstract results: Please consider revision of the sentence. “It was also found that the ginger oil inhibitory activity against E. faecalis was comparably less in anti-biofilm activity than against bacteria cultured in suspension solution.” E.g. “It was also found that the inhibitory activity of ginger oil was lower in biofilm cultures compared to planktonic cultures of E. faecalis.” Introduction: Your introduction is precise and on point. Maybe you could add 1-2 sentences about the definition/meaning and clinical problem of biofilm formation. Introduction: First paragraph, last sentence: “Inevitably, repeated use of calcium hydroxide and sodium hypochlorite as the two most commonly used root canal medicaments and irrigation solution, respectively, has been said to allow E. faecalis to adapt to the sub-lethal environment\". Please re-write this sentence as I do not get the meaning. Do you mean E. faecalis is tolerant to sodium hypochloride and calcium hydroxide treatment? Also, please change “Within the past decade, research has begun to explore some beneficial antibacterial activities of herbal medicine against oral pathogens.” Methods: Please state the source of the ginger root as the quality and therefore the reproducibility of the experiments might vary depending on the growth conditions. Method, part culture and maintenance of bacteria: The last sentence (DMSO was used as solvent for ginger oil…) is redundant and does not fit in this paragraph. Method, part broth microdilution assay: This sentence is rather confusing because the two sections you describe are not visible in Figure 1. Please clarify your pipetting scheme. Results: Please make the captions below the figures more informative by including more details about how the results presented in the figure were created (e.g. Figure 2: which strain is shown, how were the biofilms grown,…).\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [] }, { "id": "44138", "date": "13 Jun 2019", "name": "Fathilah Binti Abdul Razak", "expertise": [ "Reviewer Expertise Oral environment", "Oral biofilm", "Antimicrobial agent" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe study aimed to assess the antibacterial and antibiofilm properties of ginger oil on E. faecalis which is a pathogen often isolated from tooth root canals. E. faecalis used were in the planktonic (cell suspension) and monospecies biofilm forms. Tests performed include MIC, MBC and an antibiofilm assay.\n\nComments:\n1. Introduction: Adequate\n2. Methods: (i) OD readings from ELISA plate reader were used as a measurement of cell density in both the planktonic and biofilm state. This procedure should be revised as in the former, the cells are free in the suspension while in the later, the cells are sedimented/attached to the bottom base, which blocks vertical light source from the bottom and may contribute to higher OD readings. A different set of readings would be obtained if using a spectrophotometer with horizontal light source.\n\n(ii) An inoculation of 20ul of the broth mixture on agar plates would not be able to indicate a bacteriostatic effect of a test agent, it can only confirm the bactericidal activity of a test agent i.e. when there is no colony growth.\n\n(iii) How was a stable 3-day preformed biofilm defined, and how can Fig. 2 explain this? (iv) The 3-day biofilm could have resulted from sedimentation since the biofilm was left static for 3 days. It is thus suggested that the biofilm is gently washed first before adding the test agent. The use of a shaking incubator is also suggested.\n3. Results: (i) What were the MIC and MBC for ginger and ampicillin towards E. faecalis? (ii) Figure 3 does not really show a dose-dependent antibacterial response of E. faecalis to ginger oil, or to ampicillin. This would be obvious if error bars were shown in the figures.\n\n(iii) It is not accurate to claim that the anti-biofilm activity of ginger oil was observed to increase between 1.25 to 5 mg/ml. referring to Fig. 4 the increase were insignificant, and again this would be obvious if the error bars were shown.\n\n4. Discussion: Cross referencing to more relevant journals/works to defend and justify the study outcome would help improve the discussion.\n\nOverall comment: The manuscript was poorly prepared and certain details such as error bars in the figures were missing. Many steps were over simplified. Analysis and presentation of results can be further improved to give meaningful outcome. I agree with Reviewer 1 that the conclusion was rather over-interpreted, and very weak considering only one species of bacterium was used. More experiments are required to come to such conclusion.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [] } ]	1	https://f1000research.com/articles/7-1859
https://f1000research.com/articles/7-1242/v1	10 Aug 18	{ "type": "Research Note", "title": "Polyglutamine toxicity assays highlight the advantages of mScarlet for imaging in Saccharomyces cerevisiae", "authors": [ "Maram B. Albakri", "Yuwei Jiang", "Patrick Lajoie", "Maram B. Albakri", "Yuwei Jiang" ], "abstract": "Development of fluorescent proteins (FPs) enabled researchers to visualize protein localization and trafficking in living cells and organisms. The extended palette of available FPs allows simultaneous detection of multiples fluorescent fusion proteins. Importantly, FPs are originally derived from different organisms from jelly fish to corals and each FP display its own biophysical properties. Among these properties, the tendency of FPs to oligomerize inherently affects the behavior of its fusion partner. Here we employed the budding yeast Saccharomyces cerevisiae to determine the impact of the latest generation of red FPs on their binding partner. We used a yeast assay based on the aggregation and toxicity of misfolded polyQ expansion proteins linked to Huntington’s disease. Since polyQ aggregation and toxicity are highly dependent on the sequences flanking the polyQ region, polyQ expansions provide an ideal tool to assess the impact of FPs on their fusion partners. We found that unlike yemRFP and yFusionRed, the synthetically engineered ymScarlet displayed severe polyQ toxicity and aggregation similar to what is observed for green FP variants. Our data indicate that ymScarlet might have significant advantages over the previous generation of red FPs for use in fluorescent fusions in yeast.", "keywords": [ "fluorescent proteins", "mScarlet", "yeast", "polyglutamine toxicity", "aggregation", "Huntington’s disease" ], "content": "Introduction\n\nFollowing the development of the green fluorescent protein (GFP) from the jellyfish Aqueaora victoria (Chalfie et al., 1994), several other FPs with various spectral properties have been characterized (Thorn, 2017), allowing simultaneous detection of multiple fluorescent reporters. Among the most popular alternatives to GFP are the red fluorescent proteins (RFPs) isolated from Anthozoa coral and anemone species. One of the drawbacks of RFPs is that Anthozoa derived FPs are obligate tetramers (Baird et al., 2000; Verkhusha & Lukyanov, 2004). While development of RFPs into monomeric versions has been successful, it is often associated with reduced brightness of the fluorescent signal (Campbell et al., 2002) and therefore reduced overall performance of the resulting monomeric FPs. Moreover, RFPs such as TagRFP and mRuby2 reported as monomeric by passing purified proteins through sizing columns still display high tendency to oligomerize in living mammalian cells (Costantini et al., 2015; Costantini et al., 2012). Thus, under specific circumstances, FPs reported as monomeric can still be prone to oligomerization. Unwanted formation of oligomers could potentially significantly alter the function/localization of the protein of interest fused to the FP and render reporters unreliable (Costantini et al., 2015; Snapp et al., 2003; Zacharias et al., 2002). Indeed, various RFPs (mCherry, mKate2, mRuby, mKO2, mApple, TagRFP-T) have been shown to have differential effects on localization of cdc12 in yeast (Lee et al., 2013). Thus, being able to assess the behavior of fluorescent reporter in a given organism and/or cellular compartment is critical to help optimize fluorescent reporter design (Snapp, 2009).\n\nWe recently established a method to rapidly compare the behavior of FPs against a monomeric variant of superfolder GFP (msfGFP) in yeast (Jiang et al., 2017). The assays exploit the ability of polyglutamine expansions associated with Huntington’s disease (HD) to form toxic aggregates in yeast cells. The cause of HD can be traced back to abnormal expansion of a polyQ stretch within the first exon of the gene encoding the Huntingtin protein (Httex1) resulting in chorea and cognitive defects in patients (Gusella & MacDonald, 1995; Huntington, 1872; Penney et al., 1997). Expansion over 36 repeats is known to cause the Htt protein to misfold and aberrantly accumulate into detergent-insoluble amyloid-like aggregates in the cytoplasm of striatal neurons (Penney et al., 1997). Expression of expanded Httex1 in yeast results in severe polyQ aggregation and growth defect (Duennwald, 2013; Krobitsch & Lindquist, 2000; Mason & Giorgini, 2011; Meriin et al., 2002). Interestingly, the nature of the sequences flanking the polyQ regions (in this case fluorescent or epitope tags) greatly affects the propensity of the polyQ expansions to aggregate and to display significant growth defects in yeast (Duennwald et al., 2006). Using polyQ toxicity assays in yeast, we previously showed that a yeast-optimized version of mCherry (termed yemRFP (Keppler-Ross et al., 2008)) displays only a mild growth defects compared to yeast-optimized msfGFP (ymsfGFP) (Jiang et al., 2017). These results lead us to exploit the polyQ toxicity and aggregation assays to explore to effects of two of the most recently available RFPs. Here, we focused on FusionRed, a red monomeric fluorescent variant of mKate2 known for its low cytotoxicity in cells (Shemiakina et al., 2012) that displays low propensity to oligomerize in mammalian cells (Costantini et al., 2015). We also included mScarlet, a monomeric synthetic RFP that was recently shown to outperform other RFPs in terms of brightness of the fluorescent signal (Bindels et al., 2017). Both have yet to be characterized for expression in yeast.\n\n\nMethods\n\nAll strains are derived from W303A (Thomas & Rothstein, 1989). All experiments were conducted in synthetic complete media (SC) at 30°C.\n\nyemRFP (Keppler-Ross et al., 2008) was previously described. yFusionRed and ymScarlet were codon optimized for expression in yeast and synthetized by Genscript Inc. based on previously published sequences (Bindels et al., 2017; Shemiakina et al., 2012). RFPs were cloned into the SpeI/SalI site of p415 GPD. Alternatively, RFPs were cloned into the SpeI/SalI sites of p415 GAL1 25Q/68Q Httex1 lacking the proline rich domain, as previously described (Jiang et al., 2017). See Table 1 for a list of plasmids used in this study.\n\nYeast growth was measured by spotting assay on agar plates. Briefly, cells were cultured overnight to saturation in appropriate selection media. The next day, cells densities were equalized to OD600 0.2 and 5x serial dilutions were spotted on agar plates.\n\nAfter induction in galactose media overnight, cells were lysed using glass beads in lysis buffer (100 mM Tris pH 7.5; 200 mM NaCl; 1 mM EDTA; 5% glycerol, 1 mM Dithiothreitol (DTT) 4 mM phenylmethylsulfonyl fluoride (PSMF) and protease inhibitor cocktail). Equal amount of proteins were spotted on a nitrocellulose membrane. Membranes were blocked for 30 min in PBS-0.05%Tween at room temperature were then processed for immunoblot. Membranes were probed with anti-FLAG primary antibody (Sigma F3040) overnight at 4°C and subsequently with a secondary anti-mouse fluorescent antibody (Thermo Alexa 555 #A21424) for 1h at room temperature and imaged using a Bio-Rad ChemiDoc MP imaging system. Membranes were then stripped using the Gene Bio-Application stripping buffer and reprobed with an anti-Pgk1 primary antibody (Thermo 22C5D8). In Figure 2, for each individual antibody, both membranes were imaged simultaneously to allow direct comparison of fluorescent signal.\n\nUnder the different experimental conditions, cells were diluted 10x in growth media and plated in Lab-tek (Thermo Inc.) imaging chambers and processed for fluorescent microscopy. Images in Figure 1 were acquired using a Zeiss AxioVert A1 wide field fluorescent microscope equipped with a 63X NA 1.4 Plan Apopchromat objective, a 560 to 600nM excitation/630 to 705 nm emission bandpass filter and Zeiss Axiocam 506 mono camera. Images presented in Figure 2 were collected using a Zeiss 800 confocal microscope equipped with 488 nm and 561 nm diode lasers and a 63x PlanApochromat NA 1.4 objective.\n\n(A) yemRFP, yFusionRed and ymScarlet were introduced into centromeric vectors under the control of a constitutive GPD promoter. (B) Representative images from 3 fields of yeast cells expressing different RFPs. Imaging conditions were kept constant between samples to allow direct comparison of fluorescent intensities. Inverted black and white images are shown for clarity. Bar: 5µm (C) Yeast cells expressing the different RFPs were analyzed by flow cytometry and compared to cells carrying an empty vector. (D) Median fluorescent intensities of the various RFPs were calculated from fluorescent data acquired using flow cytometry. *p<0.05.\n\n(A) Fluorescent proteins (FPs) were cloned in frame with FLAG-Httex1 into a centromeric vector carrying a GAL1 inducible promoter. (B) Yeast cells carrying an empty vector or 25/68Q Httex1 fused to either ymsfGFP, yemRFP, yFusionRed or ymScarlet were grown to saturation overnight in glucose (control) or galactose (polyQ induced) containing media. The next days, cell concentrations were equalized to OD600 0.2 and 5 fold serial dilutions of the cell suspension spotted on synthetic complete agar media plates containing either glucose or galactose. (C) Yeast cells carrying an empty vector or 25/68Q Httex1 fused to ymsfGFP, yemRFP, yFusionRed or ymScarlet or carrying an empty vector were induced overnight in galactose containing media and protein levels analyzed by dot blot using either an anti-FLAG (detection of fluorescent fusions) or anti-Pgk1 antibody (loading control). (D) Representative fluorescent images from 3 fields of yeast cells expressing 25/68Q Httex1 fused to ymsfGFP, yemRFP, yFusionRed or ymScarlet after overnight induction in galactose-containing media.\n\nCell were cultured with appropriate media and processed for flow cytometry using a BD Bioscience FACS Celesta flow cytometer equipped with a 561 Yellow laser for imaging of RFPs. Data were analyzed using the BD FACS Diva software. All conditions were performed in triplicates, 20,000 cells were analyzed and median fluorescence intensities were calculated. No gates were applied.\n\nA two-tailed student t-test was used to determine statistical significance between the different experimental conditions in Figure 1D using GraphPad Prism v6.0h.\n\n\nResults and discussion\n\nTo analyze the performance of the three different RFPs in yeast, we first generated codon optimized versions of both FusionRed and mScarlet (termed yFusionRed and ymScarlet, respectively) (Table 2). Centromeric plasmids encoding yFusionRed, yemRFP and ymScarlet under the control of the constitutive GPD promoter were transformed in yeast (Figure 1A). Fluorescent intensities were compared using wide-field fluorescence microscopy (Figure 1B). Median fluorescent intensity (MFI) was then quantified using flow cytometry. Quantification revealed that yFusionRed was significantly dimmer than yemRFP (Figure 1C and D). This result was surprising given that previously published data reported a slightly increased brightness for FusionRed when compared to mCherry (Shemiakina et al., 2012). However, it is known that fluorescent brightness of FPs expressed in yeast can differed from the ones registered for pure purified proteins (Lee et al., 2013). As opposed to yFusionRed, ymScarlet displayed the strongest fluorescent signal (Figure 1C and D). These results are in agreement with previous studies reporting increased brightness of mScarlet compared to other RFs variants (Bindels et al., 2017). Based on the intensity of the fluorescent signal, ymScarlet appears to be the optimal RFP for imaging in yeast.\n\nNext, we sought to determine how the three different FPs affect their fusion partners in living yeast. To this end, we employed the polyQ toxicity assays. Each RFP was cloned in frame with a galactose inducible version of Httex1 carrying either 25Q (non-pathological length) or 68Q (HD-associated) (Figure 2A). 25Q constructs show no growth differences across the different FPs in both uninduced (glucose media) and polyQ-induced (galactose media) conditions indicating that expression of the different constructs results in similar growth phenotypes. When fused to 68Q Httex1, yFusionRed displayed no significant toxicity when compared to the non-toxic 25Q fusion (Figure 2B). Interestingly, ymScarlet displayed severe toxicity, showing a slow growth phenotype comparable to what was observed for ymsfGFP (Figure 2B). In addition, assessment of protein abundance for each constructs using dot blot revealed that both 25 and 68Q yFusionRed fusions were present at lower levels compared to other fluorescent counterparts Figure 2C). The cause of this phenotype is unclear and could result from increased degradation of the fusions. Based on these observations, we then investigated the effects of the different RFPs on polyQ aggregation using fluorescent microscopy. We found that yemRFP displayed robust 68Q aggregation similar to ymsfGFP as we previously described (Jiang et al., 2017). It is important to note that while prone to aggregation, yemRFP polyQ proteins were shown to form aggregates with different biophysical properties (increased detergent solubility) that can account for their moderately toxic nature (Jiang et al., 2017). In accordance with the absence of toxicity noted in the growth assay, 68Q-FusionRed did not form visible aggregates, while ymScarlet displayed strong aggregation propensity (Figure 2D). The absence of aggregation for the 68Q-yFusionRed is consistent with our previous observation showing that yomTagBFP2, a blue fluorescent proteins similarly does not form toxic aggregates (Jiang et al., 2017). In fact, both FusionRed and mTagBFP2 (Subach et al., 2011) are evolved versions of the wild-type RFP from sea anemone Entacmaea quadricolor (Merzlyak et al., 2007). In the case of mScarlet, the protein was evolved from a synthetic template design for generating a monomeric protein. Therefore, based on our data, mScarlet appears to be an attractive alternative to mCherry, which minimizes the effect of the FP on its fusion partner.\n\n\nData availability\n\nDataset 1: Raw data behind Figure 1 and Figure 2. DOI, 10.5256/f1000research.15829.d213385 (Albakri et al., 2018).\n\nBoth p415 GPD-yFusionRed and p415 GPD-ymScarlet are available from addgene (#111916/11917). All plasmids are available upon request from the corresponding author.", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis study was supported by an operating grant (MOP-325538) from the Canadian Institutes for Health Research (CIHR). YJ is supported by a M.Sc. to Ph.D. transfer scholarship from the Dean of the Schulich School of Medicine and Dentistry at The University of Western Ontario.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nThe authors would like to thanks Dr Martin Duennwald and members of his laboratory as well as Sarah Chadwick for their help during the realization of this project. The authors also thank Neta Dean (Stony Brook University, New York) for the yemRFP plasmid.\n\n\nReferences\n\nAlbakri MB, Jiang Y, Lajoie P: Dataset 1 in: Polyglutamine toxicity assays highlight the advantages of mScarlet for imaging in Saccharomyces cerevisiae. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15829.d213385\n\nBaird GS, Zacharias DA, Tsien RY: Biochemistry, mutagenesis, and oligomerization of DsRed, a red fluorescent protein from coral. Proc Natl Acad Sci U S A. 2000; 97(22): 11984–11989. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBindels DS, Haarbosch L, van Weeren L, et al.: mScarlet: a bright monomeric red fluorescent protein for cellular imaging. Nat Methods. 2017; 14(1): 53–56. PubMed Abstract \| Publisher Full Text\n\nCampbell RE, Tour O, Palmer AE, et al.: A monomeric red fluorescent protein. Proc Natl Acad Sci U S A. 2002; 99(12): 7877–7882. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChalfie M, Tu Y, Euskirchen G, et al.: Green fluorescent protein as a marker for gene expression. Science. 1994; 263(5148): 802–805. PubMed Abstract \| Publisher Full Text\n\nCostantini LM, Baloban M, Markwardt ML, et al.: A palette of fluorescent proteins optimized for diverse cellular environments. Nat Commun. 2015; 6: 7670. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCostantini LM, Fossati M, Francolini M, et al.: Assessing the tendency of fluorescent proteins to oligomerize under physiologic conditions. Traffic. 2012; 13(5): 643–649. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDuennwald ML: Yeast as a platform to explore polyglutamine toxicity and aggregation. Methods Mol Biol. 2013; 1017: 153–161. PubMed Abstract \| Publisher Full Text\n\nDuennwald ML, Jagadish S, Muchowski PJ, et al.: Flanking sequences profoundly alter polyglutamine toxicity in yeast. Proc Natl Acad Sci U S A. 2006; 103(29): 11045–11050. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGusella JF, MacDonald ME: Huntington’s disease: CAG genetics expands neurobiology. Curr Opin Neurobiol. 1995; 5(5): 656–662. PubMed Abstract \| Publisher Full Text\n\nHuntington G: On Chorea. Medical and Surgical Reporter of Philadelphia. 1872; 26(15): 317–321. Reference Source\n\nJiang Y, Di Gregorio SE, Duennwald ML, et al.: Polyglutamine toxicity in yeast uncovers phenotypic variations between different fluorescent protein fusions. Traffic. 2017; 18(1): 58–70. PubMed Abstract \| Publisher Full Text\n\nKeppler-Ross S, Noffz C, Dean N: A new purple fluorescent color marker for genetic studies in Saccharomyces cerevisiae and Candida albicans. Genetics. 2008; 179(1): 705–710. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKrobitsch S, Lindquist S: Aggregation of huntingtin in yeast varies with the length of the polyglutamine expansion and the expression of chaperone proteins. Proc Natl Acad Sci U S A. 2000; 97(4): 1589–1594. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLee S, Lim WA, Thorn KS: Improved blue, green, and red fluorescent protein tagging vectors for S. cerevisiae. PLos One. 2013; 8(7): e67902. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMason RP, Giorgini F: Modeling Huntington disease in yeast: perspectives and future directions. Prion. 2011; 5(4): 269–276. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMeriin AB, Zhang X, He X, et al.: Huntington toxicity in yeast model depends on polyglutamine aggregation mediated by a prion-like protein Rnq1. J Cell Biol. 2002; 157(6): 997–1004. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMerzlyak EM, Goedhart J, Shcherbo D, et al.: Bright monomeric red fluorescent protein with an extended fluorescence lifetime. Nat Methods. 2007; 4(7): 555–557. PubMed Abstract \| Publisher Full Text\n\nMumberg D, Müller R, Funk M: Yeast vectors for the controlled expression of heterologous proteins in different genetic backgrounds. Gene. 1995; 156(1): 119–122. PubMed Abstract \| Publisher Full Text\n\nPenney JB Jr, Vonsattel JP, MacDonald ME, et al.: CAG repeat number governs the development rate of pathology in Huntington’s disease. Ann Neurol. 1997; 41(5): 689–692. PubMed Abstract \| Publisher Full Text\n\nShemiakina II, Ermakova GV, Cranfill PJ, et al.: A monomeric red fluorescent protein with low cytotoxicity. Nat Commun. 2012; 3: 1204. PubMed Abstract \| Publisher Full Text\n\nSnapp EL: Fluorescent proteins: a cell biologist’s user guide. Trends Cell Biol. 2009; 19(11): 649–655. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSnapp EL, Hegde RS, Francolini M, et al.: Formation of stacked ER cisternae by low affinity protein interactions. J Cell Biol. 2003; 163(2): 257–269. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSubach OM, Cranfill PJ, Davidson MW, et al.: An enhanced monomeric blue fluorescent protein with the high chemical stability of the chromophore. PLos One. 2011; 6(12): e28674. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nThomas BJ, Rothstein R: Elevated recombination rates in transcriptionally active DNA. Cell. 1989; 56(4): 619–630. PubMed Abstract \| Publisher Full Text\n\nThorn K: Genetically encoded fluorescent tags. Mol Biol Cell. 2017; 28(7): 848–857. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVerkhusha VV, Lukyanov KA: The molecular properties and applications of Anthozoa fluorescent proteins and chromoproteins. Nat Biotechnol. 2004; 22(3): 289–296. PubMed Abstract \| Publisher Full Text\n\nZacharias DA, Violin JD, Newton AC, et al.: Partitioning of lipid-modified monomeric GFPs into membrane microdomains of live cells. Science. 2002; 296(5569): 913–916. PubMed Abstract \| Publisher Full Text" }	[ { "id": "37079", "date": "17 Aug 2018", "name": "Danny M. Hatters", "expertise": [ "Reviewer Expertise Experts in mammalian cell biology and Httex1 mechanisms." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis study reports the impact of three different red fluorescent protein (RFP) tags on the expression and toxicity patterns of Httex1 in yeast by benchmarking them against the green fluorescent protein tagged Httex1 as the established model system in the literature. The authors reveal some differences in their effects on Httex1 toxicity and these results are likely to be of interest to those in the Httex1-polyQ field. Overall, the manuscript is well-written and figures are clearly presented. It should be noted that the phenomena reported here is a direct extension of a broader study of other fluorescent protein fusions previously reported by the same group (and has been been referenced).\nMajor queries:\nThe authors state that the median fluorescent intensities if Fig 1 “reveals yFusionRed was significantly dimmer than yemRFP … ymScarlet displayed the strongest fluorescent signal”. As written, this implies the median fluorescence intensity provides a measure of the relative brightness of individual fluorophores. However, without quantification of the amount of protein expressed this difference in median fluorescence could simply reflect differences in total fluorophore abundance. We are not familiar with yeast transfection methods - but certainly in mammalian transfection experiments cells will invariably have a massive (many fold) variation in expression levels within populations and between different constructs including a notable population of cells that have no expression. If this is the case in yeast, then the interpretation of difference in brightness is not sufficiently supported by the data. This point needs clarification. Following from the above point, it is difficult to ascertain whether the differences in expression, toxicity and aggregate extents of the various fluorescent protein tagged Htt constructs (Fig 2) arises from the same issues. To us this would be a more likely (and simpler) explanation than what the authors explain as the protein constructs having differences in inherent toxicity (molecule per molecule). Further clarification of this issue is required.\n\nThe level of reproducibility should be indicated by including at least 3 independent replicates for the dot blots and growth assays (in Fig 2). Also, the dot blots should be quantified with the appropriate statistical methods. In Fig 1D, the authors report a significant difference in the median fluorescent intensities using a Student’s t-test to compare only two of the RFPs examined. It is suggested that an ANOVA would be more appropriate because there are more than two groups of interest. Based on the data provided, our ANOVA reports a higher degree of statistical significance for the comparison currently reported (yemRFP v ymScarlet, ) and also supports a significant difference between ymFusionRed and ymScarlet ().\n\nMinor suggestions:\nThis sentence in the abstract has an unclear meaning as it implies ymScarlet is inherently affected by polyQ: “We found that unlike yemRFP and yFusionRed, the synthetically engineered ymScarlet displayed severe polyQ toxicity and aggregation similar to what is observed for green FP variants.” Thus the sentence should be revised. Additional details for the Dot blot method (secondary antibody for Pkt-1, dilutions) would enhance reproducibility Include dimension for scale bar in Figure 2D\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? No\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [ { "c_id": "4208", "date": "26 Nov 2018", "name": "Patrick Lajoie", "role": "Author Response", "response": "We would like to thank the reviewers for their helpful comments. Below is our point-by-point response addressing the concerns raised during the review process. Major queries: The authors state that the median fluorescent intensities if Fig 1 “reveals yFusionRed was significantly dimmer than yemRFP … ymScarlet displayed the strongest fluorescent signal”. As written, this implies the median fluorescence intensity provides a measure of the relative brightness of individual fluorophores. However, without quantification of the amount of protein expressed this difference in median fluorescence could simply reflect differences in total fluorophore abundance. We are not familiar with yeast transfection methods - but certainly in mammalian transfection experiments cells will invariably have a massive (many fold) variation in expression levels within populations and between different constructs including a notable population of cells that have no expression. If this is the case in yeast, then the interpretation of difference in brightness is not sufficiently supported by the data. This point needs clarification. Yeast transformations are similar to E. coli transformations that the reviewers must be familiar with. All cells in the population are expressing the contructs since the cells are kept under constant selection. Since all constructs are expressed under the same promoter in a low copy centromeric vector (1-2 copies per cell), there is no reason to believe that some reason yFusionRed cells do not express fusion at the mRNA level. Quantification of the relative protein levels is problematic since anti-RFP antibodies could recognize mScarlet and yFusionred with different affinities. Addition of an additional epitope tag could affect behavior of the FPs. That said, we have used the polyQ fusions to show that yFusionRed protein levels are decreased using dot blot (see #3) and this appears to mask its ability to aggregate and consequently generate a toxic phenotype (Figure 4). Following from the above point, it is difficult to ascertain whether the differences in expression, toxicity and aggregate extents of the various fluorescent protein tagged Htt constructs (Fig 2) arises from the same issues. To us this would be a more likely (and simpler) explanation than what the authors explain as the protein constructs having differences in inherent toxicity (molecule per molecule). Further clarification of this issue is required. We agree with the reviewer on this point. We have now performed experiments where polyQ fusions are expressed using multicopy plasmids (2µ) were cells can express up to 50 copies of the constructs. Interestingly, under these conditions, 68Q-yFusionred can indeed aggregate and induce toxicity, albeit to lower extant than ymsfGFP-tagged polyQ counterparts. Therefore it is reasonable to hypothesize that lower expression of yFusionRed (maybe due to increased protein turnover or mRNA level) is responsible for decreased toxicity. Further experiments beyond the scope of this research note will be required to explain the relative lower expression of FusionRed in yeast. The level of reproducibility should be indicated by including at least 3 independent replicates for the dot blots and growth assays (in Fig 2). Also, the dot blots should be quantified with the appropriate statistical methods. We have now added quantification of the dot blots and perfrom quantitative growth assays in liquid media to complement the spot assays on agar plates. In Fig 1D, the authors report a significant difference in the median fluorescent intensities using a Student’s t-test to compare only two of the RFPs examined. It is suggested that an ANOVA would be more appropriate because there are more than two groups of interest. Based on the data provided, our ANOVA reports a higher degree of statistical significance for the comparison currently reported (yemRFP v ymScarlet, ) and also supports a significant difference between ymFusionRed and ymScarlet (**). We thank the reviewer and have now added appropriate statistical analysis. Minor suggestions: This sentence in the abstract has an unclear meaning as it implies ymScarlet is inherently affected by polyQ: “We found that unlike yemRFP and yFusionRed, the synthetically engineered ymScarlet displayed severe polyQ toxicity and aggregation similar to what is observed for green FP variants.” Thus the sentence should be revised. We have modified accordingly. Additional details for the Dot blot method (secondary antibody for Pkt-1, dilutions) would enhance reproducibility We have added additional information. Include dimension for scale bar in Figure 2D. We have added the information." } ] }, { "id": "37082", "date": "21 Aug 2018", "name": "Fedor V. Subach", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this manuscript Albakri Maram et al. demonstrated superior performance of the mScarlet red fluorescent protein over the FusionRed protein in the Huntington’s disease related yeast assay. First, authors showed the higher brightness of non-fused mScarlet protein vs FusionRed during constitutive expression in yeast. Albakri Maram et al. then made fusions of the mScarlet and FusionRed proteins with first exon of the Huntingtin protein (Httex1) carrying FLAG-tag and either non-pathological 25 repeats of glutamines or Huntington’s disease-associated 68 repeats of glutamines. Authors expressed these fusions in yeast and found that in opposite to the FusionRed protein, mScarlet displayed right cytotoxicity and polyQ aggregation. Overall, mScarlet red fluorescent protein can be used as a right tag in polyQ yeast assay for the Huntington’s disease studies. The strategy suggested in this paper may be used researchers for characterization of the properties of other newly developed fluorescent markers.\n\nMajor points: 1) Results and discussion section, page 6, right column. Analyzing protein expression level on dot blot authors revealed lower levels of yFusionRed protein as compared to other RFPs and based on this fact they suggested its increased degradation in yeast cells (actually, one cannot exclude that yFusionRed lower expression level is not related with its high degradation on RNA level not protein level). These findings suggest that lower expression level of yFusionRed protein in yeast my prevent FusionRed protein from toxicity and formation of aggregates, which could be observed at higher FusionRed protein concentrations only. So we cannot say unambiguously about the real reasons of the absence of toxicity and aggregates formation in the case of FLAG-68Q-yFusionRed fusion. Lower expression level of yFusionRed also may masks its tendency to oligomerize inherently and prevent from correct comparison in this respect with other fluorescent proteins.\n\nMinor points: 1) Introduction section, page 3, left column, last sentence, please, replace misprint “…and aggregation assays to explore to effects of…” with “…and aggregation assays to explore the effects of…”. 2) Methods section, page 3, right column, Yeast strains and culture conditions section, authors use term W303A. However according to provided reference it should be named as W303-1A. Please, correct. 3) Methods section, page 3, right column, DNA constructs section, authors mention that they do not use proline rich domain. Please, add explanation of the reason of this. 4) Methods section, page 3, right column, Dot blot section, please, replace misprint “…PSMF…” with “…PMSF…”. 5) Methods section, page 3, right column, Dot blot section, please, replace misprint “…at room temperature were then processed…” with “…at room temperature and then processed …”. 6) Methods section, page 3, right column, Fluorescent microscopy section and elsewhere in the text, please, replace misprint “…Fluorescent microscopy…” with “…Fluorescence microscopy…”. 7) Methods section, page 4, Table 1. I would recommend to change the abbreviations of the constructs in order to mention Httex1 and avoid 25Q abbreviation which can be mistakenly interpreted as a substitution of the residue in position 25 with Q, i.e. for example, replace “p415 Gal1-FLAG-25Q-ymsfGFP” with “p415 Gal1-FLAG-Httex1-(Q)25-ymsfGFP”. 8) Methods section, page 5, Figure 2, panel A. I would recommend replace “Httex1” with “Httex1(Q)25/(Q)68”. 9) Methods section, page 5, Figure 2, panel A and elsewhere in the text. In accordance to the 6th point I would recommend replace “25Q” and “Q68” with “(Q)25” and “(Q)68”. 10) Methods section, page 5, Legend to the Figure 1. Please, add explanation that panel (A) represents a “Scheme of fluorescent proteins expression vectors …”. Also, please, be consistent in the same respect for the panel (C). 11) Methods section, page 5, legend to the Figure 2. Please, add explanation that panel (A) represents a “Scheme of fluorescent proteins expression system …”, panel (B) – “Image of yeast colonies …” and panel (C) – “Dot blots of …”. 12) Results and discussion section, page 6, left column and elsewhere in the text and figure legends, please, replace misprint “Fluorescent intensities…” with “…Fluorescence intensities…”. 13) Results and discussion section, page 6, left column, authors mention that “…yFusionRed was significantly dimmer than yemRFP”. Below they write about the strongest fluorescent signal for ymScarlet. Please, provide these comparisons in more quantitative and scientific manner, i.e. mention in how many fold yFusionRed and ymScarlet RFPs were dimmer and brighter than yemRFP, respectively. 14) Results and discussion section, page 6, left column, please, replace misprint “… in yeast can differed from …” with “… in yeast can be different from …”. 15) Results and discussion section, page 6, left column, please, replace misprint “… of mScarlet compared to other RFs…” with “…of mScarlet compared to other RFs …”. 16) Results and discussion section, page 6, right column. Authors use 68 repeats of Q however in their previous paper of the year 2017 they utilized 72 Q repeats. Please, include in the text explanation why did you choose 68 not 72 repeats. It is important for the comparison of the data. 17) Results and discussion section, page 6, right column. Authors write that “…yFusionRed displayed no significant toxicity…”. The term “significant” suggests statistical analysis, could you, please, provide which statistic criterion was used in this case, e.g. a two-tailed student t-test or Mann-Whitney or someone else? Now Statistical analysis section describes Figure1D data only.\n\nOverall, if the points raised above will be adequately addressed, indexing in F1000Research is appropriate.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4207", "date": "26 Nov 2018", "name": "Patrick Lajoie", "role": "Author Response", "response": "We thank the reviewer for helpful comments. We have addressed concerns in the new version of the manuscript. See below for the authors’ response to specific points raised during the review process. Major points: 1) Results and discussion section, page 6, right column. Analyzing protein expression level on dot blot authors revealed lower levels of yFusionRed protein as compared to other RFPs and based on this fact they suggested its increased degradation in yeast cells (actually, one cannot exclude that yFusionRed lower expression level is not related with its high degradation on RNA level not protein level). These findings suggest that lower expression level of yFusionRed protein in yeast my prevent FusionRed protein from toxicity and formation of aggregates, which could be observed at higher FusionRed protein concentrations only. So we cannot say unambiguously about the real reasons of the absence of toxicity and aggregates formation in the case of FLAG-68Q-yFusionRed fusion. Lower expression level of yFusionRed also may masks its tendency to oligomerize inherently and prevent from correct comparison in this respect with other fluorescent proteins. We agree with the reviewer on this point. We have now performed additional experiments where polyQ fusions are expressed using multicopy plasmids (2µ) were cells can express up to 50 copies of the constructs. Interestingly, under these conditions, 72Q-yFusionred can indeed aggregate and induce toxicity, albeit to lower extant than ymsfGFP-tagged polyQ counterparts. Therefore it is reasonable to hypothesize that lower expression of yFusionRed (maybe due to increased protein turnover) is responsible for decreased toxicity. Further experiments beyond the scope of this research note will be required to explain the relative lower expression of FusionRed in yeast. Minor points: 1) Introduction section, page 3, left column, last sentence, please, replace misprint “…and aggregation assays to explore to effects of…” with “…and aggregation assays to explore the effects of…”. We have modified accordingly. 2) Methods section, page 3, right column, Yeast strains and culture conditions section, authors use term W303A. However according to provided reference it should be named as W303-1A. Please, correct. We have modified accordingly. 3) Methods section, page 3, right column, DNA constructs section, authors mention that they do not use proline rich domain. Please, add explanation of the reason of this. We added the rationale for the use of the Δ proline constructs which is described in Duennwald et al., 2006 PNAS. 4) Methods section, page 3, right column, Dot blot section, please, replace misprint “…PSMF…” with “…PMSF…”. We have modified accordingly. 5) Methods section, page 3, right column, Dot blot section, please, replace misprint “…at room temperature were then processed…” with “…at room temperature and then processed …”. We have modified accordingly. 6) Methods section, page 3, right column, Fluorescent microscopy section and elsewhere in the text, please, replace misprint “…Fluorescent microscopy…” with “…Fluorescence microscopy…”. We have modified accordingly. 7) Methods section, page 4, Table 1. I would recommend to change the abbreviations of the constructs in order to mention Httex1 and avoid 25Q abbreviation which can be mistakenly interpreted as a substitution of the residue in position 25 with Q, i.e. for example, replace “p415 Gal1-FLAG-25Q-ymsfGFP” with “p415 Gal1-FLAG-Httex1-(Q)25-ymsfGFP”. 8) Methods section, page 5, Figure 2, panel A. I would recommend replace “Httex1” with “Httex1(Q)25/(Q)68”. 9) Methods section, page 5, Figure 2, panel A and elsewhere in the text. In accordance to the 6th point I would recommend replace “25Q” and “Q68” with “(Q)25” and “(Q)68”. We thank the reviewer for the suggestion but the authors believe that to be consistent with the literature in the field, the current nomenclature should be used. 10) Methods section, page 5, Legend to the Figure 1. Please, add explanation that panel (A) represents a “Scheme of fluorescent proteins expression vectors …”. Also, please, be consistent in the same respect for the panel (C). We have modified accordingly. 11) Methods section, page 5, legend to the Figure 2. Please, add explanation that panel (A) represents a “Scheme of fluorescent proteins expression system …”, panel (B) – “Image of yeast colonies …” and panel (C) – “Dot blots of …”. 12) Results and discussion section, page 6, left column and elsewhere in the text and figure legends, please, replace misprint “Fluorescent intensities…” with “…Fluorescence intensities…”. We have modified accordingly. 13) Results and discussion section, page 6, left column, authors mention that “…yFusionRed was significantly dimmer than yemRFP”. Below they write about the strongest fluorescent signal for ymScarlet. Please, provide these comparisons in more quantitative and scientific manner, i.e. mention in how many fold yFusionRed and ymScarlet RFPs were dimmer and brighter than yemRFP, respectively. We added values in the text. 14) Results and discussion section, page 6, left column, please, replace misprint “… in yeast can differed from …” with “… in yeast can be different from …”. We have modified accordingly. 15) Results and discussion section, page 6, left column, please, replace misprint “… of mScarlet compared to other RFs…” with “…of mScarlet compared to other RFs …”. We have modified accordingly. 16) Results and discussion section, page 6, right column. Authors use 68 repeats of Q however in their previous paper of the year 2017 they utilized 72 Q repeats. Please, include in the text explanation why did you choose 68 not 72 repeats. It is important for the comparison of the data. The polyQ repeats are known to be unstable. Sequencing of the DNA revealed that our 72Q original plasmid has mutated into 68Q. This change does not impact the toxicity of the fusion. We added a note in the material and methods section. 17) Results and discussion section, page 6, right column. Authors write that “…yFusionRed displayed no significant toxicity…”. The term “significant” suggests statistical analysis, could you, please, provide which statistic criterion was used in this case, e.g. a two-tailed student t-test or Mann-Whitney or someone else? Now Statistical analysis section describes Figure1D data only. Statistical analysis was performed on the growth curve data and described in the updated figure legends." } ] } ]	1	https://f1000research.com/articles/7-1242
https://f1000research.com/articles/5-1791/v1	22 Jul 16	{ "type": "Research Article", "title": "NG-Tax, a highly accurate and validated pipeline for analysis of 16S rRNA amplicons from complex biomes", "authors": [ "Javier Ramiro-Garcia", "Gerben D. A. Hermes", "Christos Giatsis", "Detmer Sipkema", "Erwin G. Zoetendal", "Peter J. Schaap", "Hauke Smidt", "Gerben D. A. Hermes", "Christos Giatsis", "Detmer Sipkema", "Erwin G. Zoetendal", "Peter J. Schaap", "Hauke Smidt" ], "abstract": "Background Massive high-throughput sequencing of short, hypervariable segments of the 16S ribosomal RNA (rRNA) gene has transformed the methodological landscape describing microbial diversity within and across complex biomes. However, several studies have shown that the methodology rather than the biological variation is responsible for the observed sample composition and distribution. This compromises true meta-analyses, although this fact is often disregarded. Results To facilitate true meta-analysis of microbiome studies, we developed NG-Tax, a pipeline for 16S rRNA gene amplicon sequence analysis that was validated with different mock communities and benchmarked against QIIME as the currently most frequently used pipeline. The microbial composition of 49 independently amplified mock samples was characterized by sequencing two variable 16S rRNA gene regions, V4 and V5-V6, in three separate sequencing runs on Illumina’s HiSeq2000 platform. This allowed evaluating important factors of technical bias in taxonomic classification: 1) run-to-run sequencing variation, 2) PCR–error, and 3) region/primer specific amplification bias. Despite the short read length (~140 nt) and all technical biases, the average specificity of the taxonomic assignment for the phylotypes included in the mock communities was 96%. On average 99.94% and 92.02% of the reads could be assigned to at least family or genus level, respectively, while assignment to ‘spurious genera’ represented on average only 0.02% of the reads per sample. Analysis of α- and β-diversity confirmed conclusions guided by biology rather than the aforementioned methodological aspects, which was not the case when samples were analysed using QIIME. Conclusions Different biological outcomes are commonly observed due to 16S rRNA region-specific performance. NG-Tax demonstrated high robustness against choice of region and other technical biases associated with 16S rRNA gene amplicon sequencing studies, diminishing their impact and providing accurate qualitative and quantitative representation of the true sample composition. This will improve comparability between studies and facilitate efforts towards standardization.", "keywords": [ "16S rRNA amplicon analysis", "microbial community analysis", "microbial ecology", "next-generation sequencing", "bioinformatic pipeline" ], "content": "Background\n\nRecent advances in massive high-throughput, short-amplicon sequencing are revolutionizing efforts to describe microbial diversity within and across complex biomes1. Cultivation-independent whole metagenome sequencing has received increasing attention in the functional characterization of individual communities. These efforts, however, remain relatively expensive on a per sample basis, and the richer but much more unstructured information content requires complex data modelling and analysis procedures2. Therefore targeted surveys for specific taxonomic marker genes, such as the 16S ribosomal RNA (rRNA) gene3,4, remain essential in many microbial ecological studies. These surveys rely on sequencing of short, PCR amplified, hypervariable subregions rather than of the full-length gene, mostly for reasons of throughput, sequence depth and cost-efficiency.\n\nThere have been great efforts to address the accuracy and reproducibility of findings from 16S rRNA gene amplicon sequencing studies through increased levels of standardization, and software pipelines provide comprehensive protocols to analyze microbial ecology datasets. However, these efforts have arguably enhanced replicability rather than reproducibility, by providing widely adopted defaults5. To this end, Drummond6 suggested that exact replication of an experiment (i.e., replicability) is less informative (although a necessary pre-requisite for any scientific endeavour) than the corroboration of findings by reproduction in different independent setups (i.e., reproducibility)7, because biological findings that are robust to independent methodologies are arguably more dependable than any single-track analysis5. This distinction is highly relevant for the field of microbial ecology, where replicability is often confused with reproducibility, which is apparent from many often non-interchangeable methodologies.\n\nAccuracy can typically be evaluated by the addition of positive controls. Generally these are synthetic or mock communities (MCs) consisting of phylotypes that, ideally, are representative of the ecosystem of interest. MCs allow researchers to answer two essential questions concerning accuracy. 1) Do I retrieve the number of species I put in, and if so are they correctly assigned? 2) How well does the sequencing and data analysis procedure reproduce species relative abundances? Reproducibility can be evaluated by comparing separate sequencing runs and different primer pairs that cover distinct 16S rRNA gene regions. Although replicability is often achieved, accuracy has been shown to be challenging especially at higher taxonomic resolution such as at genus level8,9.\n\nCentral to all 16S rRNA gene amplicon studies are Operational Taxonomic Units (OTUs). These are often regarded as the synthetic proxy for microbial species and are typically clustered at 97% sequence similarity. However, the prokaryotic species definition remains a hotly debated topic without any satisfying solution so far10–12. Moreover, the 97% sequence similarity threshold is based on the complete 16S rRNA gene (~1500 nt), and although sequence variability is not evenly distributed it is routinely applied to short reads of 100–500 nt. Different regions would therefore require their own species level cut-off. This combination of an ambiguous prokaryotic species definition and its application to short reads, is the foundation for many complications regarding ‘correct’ OTU clustering. Hence there is little consensus on key experimental choices such as primers, targeted variable regions and OTU picking/clustering algorithms. Each of these technical aspects generate biases, and different methods produce clearly distinct results, leading to a situation where results of current studies cannot be easily compared or extrapolated to other study designs. Therefore there is a strong need for standardization.\n\nHistorically, 16S rRNA gene sequences generated in a project were initially clustered de novo into OTUs at >97% sequence similarity using various clustering algorithms, mostly because available 16S rRNA gene reference databases were thought to provide insufficient coverage13–16. Although new clustering algorithms that reduce the influence of clustering parameters, such as a hard cutoff for cluster similarity, have been specifically developed for amplicons17, cluster generation is context-dependent, i.e. different datasets generate different clusters, and different algorithms may produce different end-results5,18. Therefore, even though the same analysis framework is used, independent studies remain incomparable at OTU level. Consequently, reference-based OTU clustering has received increasing attention, due to the need for standardization, and because de-novo OTU clustering for very large datasets, such as those generated by Hiseq and Miseq sequencers has become computationally very intensive, unless greedy heuristics are employed which suffer from the problems described above. With reference-based OTU clustering, sequences are mapped to pre-clustered reference sets of curated 16S rRNA gene sequences, provided by dedicated databases such as the Ribosomal Database Project (RDP), Greengenes and SILVA19–21. The consequence of this approach is that the ‘quality’ of the clustering of the reference set propagates to reference-picked OTUs. Clustering has limited robustness5,18,22, and unbalances in databases due to over- or under-representation of certain species as well as error hotspots that are not necessarily matched to the variable regions23, can potentially lead to a biased cluster formation, driven by non-biological factors. These effects have been previously ignored or underestimated in reference OTU picking protocols5.\n\nAnother essential experimental choice concerns the selection of targeted variable region, because it should represent the sequence variability encountered with the full-length gene. Despite several studies comparing the performance of diverse regions, sequence lengths, sequencing platforms and taxon assignment methodologies, both within and across laboratories23–29, there still is no standard or consensus of best choices for variable regions. There are several factors that can lead to the commonly observed highly region-specific behaviour across datasets: 1) PCR bias of varying degrees23,27,30, 2) different regions are associated with different error profiles and different rates of chimera formation23,31, and 3) the actual variation contained in the sequence is dissimilar (e.g. some regions are not variable enough to differentiate between genera, while others are), which in turn can affect clustering5.\n\nApart from the use of a diverse range of primers and OTU picking protocols that can cause differences in results between studies and/or laboratories, sequencing error is a third important factor that defines data quality. Massive high throughput, short read length sequencing platforms have not been developed for amplicon sequencing but rather for whole genome sequencing, where sequence errors in individual reads is less important. However, in 16S rRNA gene amplicon sequencing every sequencing error could potentially lead to the false discovery of a new species. To avoid overestimation of microbial diversity, stringent quality filtering is therefore considered essential9.\n\nMethodology rather than biology has often been shown to be the largest driver of variation in microbiome studies5,18,23,27,29,32–34, and this aspect of amplicon sequencing is increasingly addressed in literature. Nevertheless, a satisfactory solution has not been found. To address the aforementioned challenges we have applied several recommendations from literature to validate high throughput, high-resolution microbiota profiling, using Illumina Hiseq2000 101nt paired end sequencing data as a test case. We implemented redundancy by sequencing two tandem variable 16S rRNA gene regions in parallel (V4 and V5-V6). To find optimal filtering settings and to empirically determine the noise floor, multiple standardized mock communities specifically designed to tackle issues associated with filtering parameter optimization were added to each sequencing run. To our knowledge a similar setup with multiple MCs and different variable regions has not been applied to datasets generated by the Illumina platform. This set-up has enabled us to develop NG-Tax, a pipeline that better accounts for error associated with a range of technical aspects of 16S rRNA gene amplicon sequencing. NG-Tax will improve comparability by removing technical bias and facilitate efforts towards standardization, by focusing on reproducibility as well as accuracy. To assess the performance we benchmarked the results obtained with NG-Tax with results obtained with QIIME35, a common pipeline used for the analysis of this type of data.\n\n\nResults and discussion\n\nNG-Tax consists of three core elements, namely barcode-primer filtering, OTU-picking and taxonomic assignment (Figure 1).\n\nThe steps that are performed per sample are depicted in grey boxes. Optional steps are indicated with dashed lines.\n\nBarcode-Primer filtering. In a first step, paired end libraries are combined, and only read pairs with perfectly matching primers and barcodes are retained. To this end, both primers are barcoded to facilitate identification of chimeras produced during library generation after pooling of individual PCR products.\n\nOTU picking. For each sample an OTU table is created with the most abundant sequences, using a minimum user defined relative abundance threshold. In this particular study we employed a threshold of 0.1% minimum relative abundance. Lowering the threshold will lead to the acceptance of low abundant OTUs, with an increased probability of these OTUs being artifacts due to sequencing and PCR errors. Abundance thresholds are commonly used to remove spurious OTUs generated by sequencing and PCR errors8,36, but previous studies applied a fraction threshold defined by the complete dataset under study, thereby ignoring sample size heterogeneity which may lead to under-representation of asymmetrically distributed OTUs.\n\nCommonly employed quality filtering parameters based on Phred score, such as minimum average Phred score, maximum number of ambiguous positions, maximum bad run length, trimming and minimum read length after quality trimming, are not utilized in NG-Tax because quality scores from the Illumina base caller have been shown to be of limited use for the identification of actual sequence errors for 16S rRNA gene amplicon studies9,37. Additionally, these quality scores only check for errors that occurred during sequencing, but do not account for other sources of error, such as PCR amplification, whereas quality filtering by abundance is sensitive to any source of error. Moreover, the application of global parameters (e.g. average Phred score) ignores that error is sequence-specific, and hence some sequences could be affected more than others. If a species specific amplicon is more prone to PCR or sequencing errors, the relative abundance of that particular species will be underestimated. To compensate for this potential bias, discarded reads are clustered to the OTUs with one mismatch.\n\nFinally, all OTUs are subjected to non-reference based chimera checking according to the following principle: given three OTUs named A, B and C, C will be considered a chimera when the following conditions are satisfied: C and A 5’ reads are identical, C and B 3’ reads are identical and both OTUs, A and B, are at least twice as abundant as OTU C. A complete overview of the number of sequences retained in both pipelines, i.e. NG-Tax and QIIME, as well as the final number of OTUs, is provided in Dataset 1.\n\nTaxonomic assignment. In the current version of NG-Tax, taxonomy is assigned to OTUs utilizing the uclust algorithm16 and the Silva_111_SSU Ref database, containing 731,863 unique full length 16S rRNA gene sequences. To ensure maximum resolution and avoid the risk of errors due to clustering-associated flaws (e.g. reference sequence error hotspots, overrepresentation of certain species and lack of robustness in cluster formation by clustering algorithms), we use the non-clustered database. To speed up the procedure by several orders of magnitude, 16S rRNA gene sequences from the reference database are trimmed to contain only the region amplified by the primers. For each OTU, a taxonomic assignment is retrieved at six different identity thresholds levels (100%, 98%, 97%, 95%, 92% and 90%) and at two taxonomic levels (genus and family). The final taxonomic label is determined by the assignments that show concordance at the highest taxonomic resolution.\n\n\nValidation\n\nOur main objective was to develop a pipeline that accurately reproduces the synthetic MCs and also reduces the impact of experimental choices on the results. To achieve this goal, four synthetic communities of varying complexity were created, consisting of 16S rRNA gene amplicons of phylotypes (PTs) associated with the human GI-tract (Table 1). This specific setup limited the likelihood of overfitting to a particular OTU composition or distribution and allowed us to assess (1) the quantification potential, (2) noise floor and (3) the effect of richness and diversity on quality filtering parameters, thus ensuring a higher fidelity with biological samples than by using a single MC. As a reference, to assess the quality of the taxonomic classifications, full length sequences for all PTs were obtained through Sanger sequencing. Expected MCs were created by trimming the full length sequences to the sequenced region. MC1 and MC2 consisted of equimolar amounts of 17 and 55 PTs, respectively. MC3 contained 55 PTs in staggered concentrations typical for the human GI-tract, and MC4 included 50 PTs with relative abundances ranging between 0.001 and 2.49%. To account for pipetting errors, each of the four MCs was produced in triplicate. To design a pipeline that puts more focus on biology, these 12 MC templates were used to sequence the MCs with different conditions that cover most of the technical bias associated with 16S rRNA gene amplicon studies reported in literature. To this end, we 1) targeted either region V4 or region V5-V6, 2) used four PCR protocols differing in the number of PCR cycles and reaction volumes 3) PCR products were analysed in three different sequencing runs and in seven different libraries, and 4) two different library preparation protocols (with and without an extra amplification of 10 cycles) were applied (Table 2). In addition the sequencing depth ranged from 2363 to 335822 reads per sample (Dataset 1).\n\nTaxonomy of 55 reference sequences used in this study was based on classification of the full length reference sequences according to the RDP classifier.\n\nTo evaluate the accuracy and reproducibility of taxonomic classification using a low information content of ~140 nt compared to a maximal information content of ~1500 nt, we compared the NG-Tax classification of all 55 reference sequences used in this study trimmed to V4 and V5-V6, with a classification of the corresponding full length reference sequences using the RDP classifier (RDPc)24 (Figure 2). At family level, all three classifications (i.e. full length, V4 and V5-V6) were in complete concordance for all phylotypes. Correspondingly, the consistency at genus level was very high. Only a few phylotypes (nine and five for V4 and V5-V6 amplicons, respectively), that belong to poorly classified families such as Peptostreptococcaceae, Ruminococceae and Enterobacteriaceae, attained higher resolution using the full length sequences and RDPc. In turn, for Pseudobutyrivibrio (PT35), a higher resolution was attained with short reads due to the high specificity of the hypervariable regions, which can be overshadowed when using the full length sequence. Lastly, only two (PT51, PT52, V4) and one (PT51, V5-V6) assignments at genus level (both members of the Enterobacteriaceae) were incongruent between classification of the short and full length sequences. Overall, the V5-V6 amplicons outperformed the V4 amplicons because this region allowed for differentiation between members of the Enterobacteriaceae. The average taxonomic specificity (percentage of hits with an identical taxonomic label) for all reference phylotypes was 96% for both regions with an average of 1485 and 635 hits for regions V4 and V5-V6, respectively. The high specificity and high number of hits at very high identity thresholds, combined with the fact that the vast majority of V4 and V5-V6 based assignments matched, testifies for the reliability and quality of the assignments.\n\nThe x-axis shows taxonomic assignments by three methods: RDP full length/NG-Tax V5-V6 trimmed/NG-Tax V4 trimmed. If all three assignments are in agreement, that classification is shown. The y-axis depicts assignment specificity (the fraction of hits with an identical label) and the total number of hits supporting this taxonomic label.\n\nTo assess the ability to reproduce the expected composition of the MCs we benchmarked NG-Tax with QIIME, a commonly employed 16S rRNA gene amplicon analysis pipeline. The reproduction of MC compositional profiles generated by amplicon sequencing on Illumina platforms commonly suffers from a high fraction of poorly classified and spurious OTUs9. Using QIIME, up to 20% of OTUs per sample could not be assigned beyond the class level (Figure 3). In contrast, with NG-Tax we observed excellent reproduction of the expected profiles (Figure 4). An average of 92.02% of the reads could be assigned to genus level and 99.94% to at least family level. Spurious genera (Robinsoniella, Subdoligranulum, Cupriavidus, Ralstonia, Kluyvera and Pantoea) represented on average only 0.02% of the reads per sample compared to an average of 23% misclassified reads using QIIME38. One template, PT17 (Parabacteroides), attracted so much sequencing error in the V4 region that it was rendered undetectable although it was amplified by the primers (Supplementary Figure 1). Therefore, to test both pipelines without this sequencing anomaly, it was removed from the analysis.\n\nRichness and diversity measures are important for understanding community complexity and dynamics. Among these measures, α-diversity is defined as the diversity within a sample, which is often estimated based on the abundance distribution (evenness) and number (richness) of species, whereas β-diversity is defined as the partitioning of diversity among communities. The ability of researchers to quantify richness and diversity hinges on an accurate assessment of the composition of these communities39. For microbial communities, this has been particularly challenging, as none of the existing molecular microbial ecology methods normally capture more than a small proportion of the estimated total richness in most microbial communities40. For deep sequencing based approaches, filtering strategies that remove low-abundance reads make it impossible to apply richness estimation metrics such as the Chao1 index and the ACE coverage estimator, because low-abundance read counts are included in their calculations. Conversely, richness estimates based on unfiltered datasets are unlikely to be accurate, if many of the reads actually represent PCR and/or sequencing errors9. In contrast to purely OTU-based methods, divergence-based methods account for the fact that not all species within a sample are equally related to each other, considering two communities to be similar if they harbour the same phylogenetic lineages, even if the species representing those lineages in each of the communities are different. Consequently, these methods are more powerful than purely OTU-based methods, because similarity in 16S rRNA gene sequence often correlates with phenotypic similarity in key features such as metabolic capabilities. An added benefit is that small errors that are likely due to unfiltered sequencing errors, are punished less severely because OTUs that are only a few nt distant from each other due to error are still closely related using divergence based indices. Therefore, these indices probably provide a better estimate of the true diversity for data generated by high throughput next generation technology sequencers.\n\nBecause the focus of NG-Tax is to retain as much biological signal as possible while minimizing the impact of any technical choice, divergence-based α-diversity (Phylogenetic Diversity (PD)41) and β-diversity (Unifrac39) metrics were used to visualize the diversity within and between MCs (Figure 5). The results obtained with QIIME suffered from all of the previously described technological artifacts. The MCs clustered by primer pair instead of MC, and within each cluster the structure, i.e. the position of MCs relative to each other, was different. More importantly, the true biological variation depicted by the expected composition was reproduced by neither primer pair (Figure 5C). Based on these results not only the Principle Coordinates Analysis (PCoA) based conclusions would have been different for both primer pairs, but also the differences in taxonomic classification could lead to significant changes in identified biomarkers, in line with what has previously been observed by He and co-workers28. Here we show that replicability within a variable region was attained. The more important reproducibility, however, i.e. the corroboration of findings by reproduction in different independent setups that use e.g. different primers, was not. This is an important observation because biological findings should be insensitive to independent methodologies5. In line with the above, also the observed α-diversity (PD) was found highly inflated and the biological order was not reproduced (Figure 5D). In contrast, NG-Tax provided a clear separation of samples by MC type and their representative expected samples regardless of variable region, PCR protocol, sequencing run, library and sequencing depth. These results are remarkable, given the biases associated with each of these categories and the difference in resolution between the two regions (Figure 5A). Moreover, MC2, MC3 and MC4 were very similar, sharing the same genera and largely the same phylotypes, only differing in relative distribution (Table 1). Correspondingly, rarefaction curves for α-diversity (Figure 5B) showed excellent reproduction of the true diversity. A perfect overlap cannot be achieved since the expected MCs do not account for sequencing or PCR errors, and these factors cannot be completely removed from real sequencing data.\n\nA/C. PCoA using Weighted Unifrac of all sequenced and expected MCs as obtained after processing of data using NG-Tax (A) and QIIME (C). Darker colored triangles represent the expected composition while lighter colored circles represent sequenced samples. B/D. Rarefaction curves of PD for all MCs and their expected counterparts for NG-Tax (B) and QIIME (D). Dashed lines represent the expected composition while solid lines represent sequenced samples.\n\n\nConclusions\n\nAn increasing number of studies have shown that the chosen methodology rather than the natural variance is responsible for the greatest variance in microbiome studies5,18,23,27,32–34. Some authors raised their concern with comparing data generated using different strategies29, which basically suggests that true reproducibility (i.e. using different approaches and drawing the same biological conclusions) cannot be attained. This is an alarming observation since studies are often used to identity biomarker organisms, associated with certain host phenotypes (often comparing a diseased state to a healthy state), yet the use of different primers might show different biomarkers8,22,23,27,28,30. So far, neither currently available pipelines nor taxonomic classifiers have been able to efficiently reduce the noise in this type of data. Nevertheless, in properly de-noised datasets, taxonomical profiles, richness and diversity should be close to the expected values and the abundance of unassigned and poorly assigned reads should be low except when dealing with largely unexplored environments that are not sufficiently covered yet by the reference databases. At lower noise levels different variable regions should yield similar conclusions with small variations due to region specific resolution, and minor changes in the experiment should still deliver the same biological conclusions. Here we presented NG-Tax, an improved pipeline for 16S rRNA gene amplicon sequencing data, which continues to be a backbone in the analysis of microbial ecosystems. Several novel steps ensure much needed improved robustness against errors associated with technical aspects of these studies, such as PCR protocols, choice of 16S rRNA gene variable region and variable rates of sequencing error27,29,32. The commonly reported problems such as many un- or poorly classified OTUs, inflated richness and diversity, taxonomic profiles that do not match the expected ones, region dependent taxonomic classification and results being highly dependent on minor changes in the experimental setup have been tackled with NG-Tax. Despite the short read length (~140 nt) and all technical biases, the average taxonomic assignment specificity for the phylotypes included in the MCs was 96%. In addition, 92.02% of the reads could be assigned to at genus level and 99.94% to at least family. Spurious genera represented only 0.02% of the reads per sample. More importantly, rarefaction curves and PCoA plots confirmed improved performance of NG-Tax with respect to clustering based on biology rather than technical aspects, such as sequencing run, library or choice of 16S rRNA gene region. Therefore NG-Tax represents a method for 16S rRNA gene amplicon analysis with improved qualitative and quantitative representation of the true sample composition. Additionally, the high robustness against technical bias associated with 16S rRNA gene amplicon studies will improve comparability between studies and facilitate efforts towards standardization.\n\n\nMethods\n\nPrimer pairs 515F (5’-GTGCCAGCMGCCGCGGTAA) - 806R (5’-GGACTACHVGGGTWTCTAAT) and BSF784F (5’-RGGATTAGATACCC) - 1064R (5’-CGACRRCCATGCANCACCT) have been previously reported for amplification of the V48 and V5- V627 regions of the bacterial 16S rRNA gene, respectively. They were selected based on 1) experimental validation, 2) taxonomic coverage of the relevant ecosystem (Supplementary Figure 2) and 4) adherence to specific rules associated with the sequencing platform, such as a maximum amplicon size of <500 nt. Unless noted otherwise all primers were ordered at Biolegio (Nijmegen, Netherlands).\n\nAt the time of sequencing Illumina’s Hiseq2000 allowed for multiplexing of up to 12 samples per lane using an index or barcode read provided by Illumina. To achieve optimal sample throughput and phylogenetic depth, 70 primers containing a custom designed 8nt barcode were developed to combine with the Illumina barcodes to reach a maximum throughput of 12×70 samples per lane. Each set of 70 barcoded samples are referred to as “library”. Low diversity samples, such as 16S rRNA gene amplicons, can lead to problems with base calling due to overexposure of fluorescent labels. Therefore, the set of 70 barcodes was specifically designed to possess an equal base distribution over their complete length. Additionally, to avoid differential amplification, a two-base “linker” sequence that is not complementary to any 16S rRNA sequence at the corresponding position, from a database that contains 1132 phylotypes associated with the Human GI tract42, was inserted between the primer and barcode. The resulting set of 70 barcoded primers was checked for avoidance of secondary structure formation within or between primers (i.e., primer-dimers) or between barcodes and primers, using PrimerProspector43.\n\nAll MCs were mixed in triplicate to account for pipetting error. These MCs ranged from 17–55 species in both equimolar and staggered compositions. One MC contained members at very low abundances of 0.1, 0.01 and 0.001% (Table 2). Amplicons were generated either from cloned 16S rRNA gene amplicons, isolates available in the local culture collection of the Laboratory of Microbiology, Wageningen University, or strains ordered from DSMZ and cultured according to DSMZ recommendations, after which genomic DNA was isolated using the Genejet genomic DNA isolation kit (Thermo fisher scientific AG, Reinach, Zwitserland). A 16S rRNA gene specific PCR was performed using the universal primers 27F (5’-GTTTGATCCTGGCTCAG) - 1492R (5’-GGTTACCTTGTTACGACTT) to obtain full length amplicons of which size and concentration were checked on a 1% agarose gel and which were column purified and quantified with the Qubit 2.0 fluorometer, and dsDNA BR assay kit (Invitrogen, Eugene, USA). Amplicons were mixed in the MCs to obtain the specified relative abundances. High quality full length reference sequences of all MC members were obtained by Sanger sequencing at GATC Biotech AG (Constance, Germany) with sequencing primers 27F - 1492R in order to confirm their identity. The MCs were diluted 103-fold and subsequently used as templates in PCRs for the generation of barcoded PCR products.\n\n\nBarcoded PCR\n\nUnless noted otherwise, each sample was amplified in triplicate using Phusion hot start II high fidelity polymerase (Thermo fisher scientific AG), checked for correct size and concentration on a 1% agarose gel and subsequently combined and column-purified with the High pure PCR cleanup micro kit (Roche diagnostics, Mannheim, Germany). Forty μl PCR reactions contained 28.4 μL nucleotide free water (Promega, Madison, USA), 0.4 μL of 2 U/μl polymerase, 8 μL of 5× HF buffer, 0.8 μl of 10 μM stock solutions of each of the forward (515F) and reverse (806R) primers, 0.8 μL 10mM dNTPs (Promega) and 0.8 μL template DNA (103 × diluted 200 ng/μl stock). Reactions were held at 98°C for 30 s and amplification proceeding for 25 cycles at 98°C for 10 s, 50°C for 10 s, 72°C for 10 s and a final extension of 7 min at 72°C. Purified amplicons were quantified using Qubit. For primer pair BSF784F-1064R the thermal cycling conditions were identical to those detailed above except that the annealing temperature was 42°C. To quantify noise generated by the PCR protocol, several reactions were performed with 30 or 35 cycles and 1× 100μl reaction instead of pooling 40μl in triplicate (Table 2).\n\nA composite sample for sequencing was created by combining equimolar amounts of amplicons from the individual samples, followed by gel purification and ethanol precipitation to remove any remaining contaminants. The resulting libraries were sent to GATC Biotech AG for sequencing on an Illumina Hiseq2000 instrument.\n\n\nSequence analysis with QIIME\n\nWe have used QIIME to benchmark NG-Tax. Illumina fastq files were de-multiplexed, quality filtered and analyzed using QIIME (v. 1.9)35 with closed reference OTU picking, using default settings and quality parameters as previously reported9.\n\n\nNG-tax pipeline and user manual\n\nThe NG-tax pipeline, user manual and files and code to reproduce the presented results, are available for download at the ftp server http://systemsbiology.nl/NG-Tax/.\n\n\nAbbreviations\n\nrRNA: ribosomal RNA; MC: Mock Community; OTU: Operational Taxonomic Unit; PT: Phylotype; RDP: Ribosomal Database Project; RDPc: RDP classifier; PD: Phylogenetic Diversity; PCoA: Principle Coordinates Analysis\n\n\nData availability\n\nF1000Research: Dataset 1. Raw data of NG-Tax pipeline for analysis of 16S rRNA amplicons from complex biome, 10.5256/f1000research.9227.d13012044\n\nSequence data have been deposited in the European Nucleotide Archive45, accession number [ENA:PRJEB11702]) http://www.ebi.ac.uk/ena/data/view/PRJEB11702 (amplicon sequencing data for all 49 samples) and [ENA:LN907729-LN907783]) (full length 16S rRNA gene sequences for all 55 PTs).", "appendix": "Author contributions\n\n\n\nGDAH and JRG wrote the manuscript. JRG conceived NG-Tax and performed the statistical analysis. GDAH, PS, EGZ and HS designed the experiment, GDAH constructed the MCs and prepared libraries 1-2 for sequencing. DS and CG provided the data for libraries 3-7. HS, DS, EGZ and PS helped to draft the manuscript, of which the final version was read and approved by all the authors.\n\n\nCompeting interests\n\n\n\nThe authors declare that they have no competing interests.\n\n\nGrant information\n\nThis work was funded by Top Institute Food and Nutrition (TIFN, Wageningen, The Netherlands), a public - private partnership on precompetitive research in food and nutrition. We are grateful for additional support from the European Community’s Seventh Framework Program (FP7/2007–2013) under grant agreement no. 227197 Promicrobe.\n\n\nAcknowledgements\n\nWe thank Gianina Bacanu for generating libraries 3–7 and Jesse van Dam for revising the scripts.\n\n\nSupplementary Figures\n\nA) Nucleotide distribution of PT17 (Parabacteroides) for each of the four primers. Positions under the black segment are fixed and specific for PT17 preventing the inclusion of sequences belonging to a different PT. B) Percentage of 10 most abundant sequences for PT17 obtained with each of the primers. PT17 (Parabacteroides) presented a sequencing anomaly in the reverse V4 region (primer R806) (Supplementary Figure 1A). From positions 50 to 67 this region had higher error rate than the other three regions. The noise generated from this anomaly masked the biological signal rendering PT17 undetectable. In fact the most abundant sequence represented less than 0.45% of the total reads, while for the other three regions the most abundant sequence represented more than 80% (Supplementary Figure 1B). We decided to remove the sequences belonging to PT17 from V5-V6 samples to avoid region clustering due to the presence of PT17. Our intention in this study was to test region performance under conditions in which sequencing anomalies like the one showed in Supplementary Figure 1 are not present.\n\nForward (left bars) and reverse (right bars) primer coverage of the major bacterial phyla associated with the human GI tract using RDP’s probematch program with one mismatch allowed.\n\n\nReferences\n\nHuman Microbiome Project Consortium: Structure, function and diversity of the healthy human microbiome. Nature. 2012; 486(7402): 207–14. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nQin J, Li R, Raes J, et al.: A human gut microbial gene catalogue established by metagenomic sequencing. Nature. 2010; 464(7285): 59–65. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOlsen GJ, Lane DJ, Giovannoni SJ, et al.: Microbial ecology and evolution: a ribosomal RNA approach. Annu Rev Microbiol. 1986; 40: 337–65. PubMed Abstract \| Publisher Full Text\n\nLane DJ, Pace B, Olsen GJ, et al.: Rapid determination of 16S ribosomal RNA sequences for phylogenetic analyses. Proc Natl Acad Sci U S A. 1985; 82(20): 6955–9. PubMed Abstract \| Free Full Text\n\nSchmidt TS, Matias Rodrigues JF, von Mering C: Limits to robustness and reproducibility in the demarcation of operational taxonomic units. Environ Microbiol. 2015; 17(5): 1689–706. PubMed Abstract \| Publisher Full Text\n\nDrummond C: Replicability is not reproducibility: nor is it good science. Proc Eval Methods Mach Learn. Workshop 26th ICML, Montreal, Quebec, Canada., 2009. Reference Source\n\nCasadevall A, Fang FC: Reproducible science. Infect Immun. 2010; 78(12): 4972–5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCaporaso JG, Lauber CL, Walters WA, et al.: Global patterns of 16S rRNA diversity at a depth of millions of sequences per sample. Proc Natl Acad Sci U S A. 2011; 108(Suppl 1): 4516–22. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBokulich NA, Subramanian S, Faith JJ, et al.: Quality-filtering vastly improves diversity estimates from Illumina amplicon sequencing. Nat Methods. 2013; 10(1): 57–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStackebrandt E, Goebel BM: A place for DNA–DNA reassociation and 16S ribosomal-RNA sequence analysis in the present species definition in bacteriology. Int J Syst Bacteriol. 1994; 44: 846–849. Publisher Full Text\n\nStackebrandt E, Frederiksen W, Garrity GM, et al.: Report of the ad hoc committee for the re-evaluation of the species definition in bacteriology. Int J Syst Evol Microbiol. 2002; 52(Pt 3): 1043–7. PubMed Abstract \| Publisher Full Text\n\nStackenbrandt E, Ebers J: Taxonomic parameters revisited: tarnished gold standards. Microbiol Today. 2006; 33(4): 152–155. Reference Source\n\nSchloss PD, Westcott SL, Ryabin T, et al.: Introducing mothur: open-source, platform-independent, community-supported software for describing and comparing microbial communities. Appl Environ Microbiol. 2009; 75(23): 7537–41. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCai Y, Sun Y: ESPRIT-Tree: hierarchical clustering analysis of millions of 16S rRNA pyrosequences in quasilinear computational time. Nucleic Acids Res. 2011; 39(14): e95. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLi W, Godzik A: Cd-hit: a fast program for clustering and comparing large sets of protein or nucleotide sequences. Bioinformatics. 2006; 22(13): 1658–9. PubMed Abstract \| Publisher Full Text\n\nEdgar RC: Search and clustering orders of magnitude faster than BLAST. Bioinformatics. 2010; 26(19): 2460–1. PubMed Abstract \| Publisher Full Text\n\nMahé F, Rognes T, Quince C, et al.: Swarm: robust and fast clustering method for amplicon-based studies. PeerJ. 2014; 2: e593. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSun Y, Cai Y, Huse SM, et al.: A large-scale benchmark study of existing algorithms for taxonomy-independent microbial community analysis. Brief Bioinform. 2011; 13(1): 107–21. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCole JR, Wang Q, Fish JA, et al.: Ribosomal Database Project: data and tools for high throughput rRNA analysis. Nucleic Acids Res. 2014; 42(Database issue): D633–42. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDeSantis TZ, Hugenholtz P, Larzen N, et al.: Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB. Appl Environ Microbiol. 2006; 72(7): 5069–72. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYilmaz P, Parfrey LW, Yarza P, et al.: The SILVA and “All-species Living Tree Project (LTP)” taxonomic frameworks. Nucleic Acids Res. 2014; 42(Database issue): D643–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHe Y, Caporaso JG, Jiang XT, et al.: Stability of operational taxonomic units: an important but neglected property for analyzing microbial diversity. Microbiome. 2015; 3: 20. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJumpstart Consortium Human Microbiome Project Data Generation Working Group. Evaluation of 16S rDNA-based community profiling for human microbiome research. PLoS One. 2012; 7(6): e39315. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWang Q, Garrity GM, Tiedje JM, et al.: Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microbiol. 2007; 73(16): 5261–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiu Z, Lozupone C, Hamady M, et al.: Short pyrosequencing reads suffice for accurate microbial community analysis. Nucleic Acids Res. 2007; 35(18): e120. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWang Y, Qian PY: Conservative fragments in bacterial 16S rRNA genes and primer design for 16S ribosomal DNA amplicons in metagenomic studies. PLoS One. 2009; 4(10): e7401. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nClaesson MJ, Wang Q, O'Sullivan O, et al.: Comparison of two next-generation sequencing technologies for resolving highly complex microbiota composition using tandem variable 16S rRNA gene regions. Nucleic Acids Res. 2010; 38(22): e200. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHe Y, Zhou BJ, Deng GH, et al.: Comparison of microbial diversity determined with the same variable tag sequence extracted from two different PCR amplicons. BMC Microbiol. 2013; 13: 208. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nClooney AG, Fouhy F, Sleator RD, et al.: Comparing Apples and Oranges?: Next Generation Sequencing and Its Impact on Microbiome Analysis. PLoS One. 2016; 11(2): e0148028. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEngelbrektson A, Kunin V, Wrighton KC, et al.: Experimental factors affecting PCR-based estimates of microbial species richness and evenness. ISME J. 2010; 4(5): 642–7. PubMed Abstract \| Publisher Full Text\n\nMay A, Abeln S, Crielaard W, et al.: Unraveling the outcome of 16S rDNA-based taxonomy analysis through mock data and simulations. Bioinformatics. 2014; 30(11): 1530–8. PubMed Abstract \| Publisher Full Text\n\nTremblay J, Singh K, Fern A, et al.: Primer and platform effects on 16S rRNA tag sequencing. Front Microbiol. 2015; 6: 771. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBarb JJ, Oler AJ, Kim HS, et al.: Development of an Analysis Pipeline Characterizing Multiple Hypervariable Regions of 16S rRNA Using Mock Samples. PLoS One. 2016; 11(2): e0148047. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKoskinen K, Auvinen P, Björkroth KJ, et al.: Inconsistent Denoising and Clustering Algorithms for Amplicon Sequence Data. J Comput Biol. 2015; 22(8): 743–51. PubMed Abstract \| Publisher Full Text\n\nCaporaso JG, Kuczynski J, Stombaugh J, et al.: QIIME allows analysis of high-throughput community sequencing data. Nat Methods. 2010; 7(5): 335–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDegnan PH, Ochman H: Illumina-based analysis of microbial community diversity. ISME J. 2012; 6(1): 183–94. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchirmer M, Ijaz UZ, D'Amore R, et al.: Insight into biases and sequencing errors for amplicon sequencing with the Illumina MiSeq platform. Nucleic Acids Res. 2015; 43(6): e37. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFaith JJ, Guruge JL, Charbonneau M, et al.: The long-term stability of the human gut microbiota. Science. 2013; 341(6141): 1237439. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLozupone C, Knight R: UniFrac: a new phylogenetic method for comparing microbial communities. Appl Environ Microbiol. 2005; 71(12): 8228–35. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHong SH, Bunge J, Jeon SO, et al.: Predicting microbial species richness. Proc Natl Acad Sci U S A. 2006; 103(1): 117–22. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFaith DP: The role of the phylogenetic diversity measure, PD, in bio-informatics: getting the definition right. Evol Bioinform Online. 2007; 2: 277–83. PubMed Abstract \| Free Full Text\n\nRajilic-Stojanovic M, Heilig HG, Molenaar D, et al.: Development and application of the human intestinal tract chip, a phylogenetic microarray: analysis of universally conserved phylotypes in the abundant microbiota of young and elderly adults. Environ Microbiol. 2009; 11(7): 1736–51. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWalters WA, Caporaso JG, Lauber CL, et al.: PrimerProspector: de novo design and taxonomic analysis of barcoded polymerase chain reaction primers. Bioinformatics. 2011; 27(8): 1159–61. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRamiro Garcia J, DA Hermes G, Giatsis C, et al.: Dataset 1 in: NG-Tax, a highly accurate and validated pipeline for analysis of 16S rRNA amplicons from complex biomes. F1000Research. 2016. Data Source\n\nEuropean Nucleotide Archive. Reference Source" }	[ { "id": "15177", "date": "02 Aug 2016", "name": "Thomas S. B. Schmidt", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn their manuscript, the authors introduce NG-Tax, an open-source software for the (meta-)analysis of 16S rRNA-based microbiome datasets. Their tool focuses on an important and so-far arguably understudied aspect of microbial ecology research: the integration of results across studies, in view of both technical and biological variation.\nThe approach is interesting and addresses important points. In particular, several sequencing datasets of different mock communities were generated, even using different primer sets: this is great data to benchmark on, and many (most) other papers introducing tools do not provide benchmarks on such an array of real (mock) data. In general, I feel that this is very interesting work and that NG-Tax can be a promising alternative to existing tools in the field.\nHowever, there are several points that I feel would need to be addressed in order for the manuscript to stand tall, and for the reader to get a good understanding of how NG-Tax can be useful in practice.\n\nMajor comments:\nEven after reading the manuscript and online user manual repeatedly, I have to admit that it is not completely clear to me how NG-Tax works in detail, and in which points exactly it differs from existing approaches. Based on the introduction, I gather that NG-Tax relies on closed-reference OTU picking, but this is not mentioned explicitly anywhere in the text. Also, does reference-based OTU picking in NG-Tax rely on uclust? If yes, which version and parameters were used, and how do they differ from QIIME’s defaults? Also, the Background and Discussion sections do not elaborate on the various disadvantages of closed-reference approaches; most importantly, closed-ref only takes into account sequences matching the database and removes everything else. When integrating sequence data from different primer sets, this is arguably the most straightforward approach; however, the limitations should be discussed.\n\nI gather from the text that NG-Tax’s main innovations are the use of primer-tailored reference databases and a different (more conservative) read abundance filtering scheme. It is perfectly valid to benchmark these against QIIME’s default settings; however, it would be great to see how QIIME performs with similarly conservative settings, to better understand where NG-Tax’s edge in performance comes from.\n\nRegarding taxonomy assignments, it is valid to compare NG-Tax’s uclust-based approach to QIIME’s uclust-based approach. However, I believe that the gold standard continues to be the RDP Classifier, and it would be interesting to see a performance comparison to this tool (on the short-read data, not only on full-length reads). Also, how does taxonomic classification by NG-Tax differ conceptually from RTAX (http://www.uio.no/english/services/it/research/hpc/abel/help/software/rtax.html)? I do believe that they are not equivalent, but the approaches appear somewhat related.\n\nIn general, the results on taxonomic classification are not discussed quantitatively. From Figures 3&4, the visual impression is that NG-Tax indeed better approximates expected taxonomic profiles than QIIME, but it is hard to quantify this from stacked bar charts. I would suggest to compute e.g. Euclidean or more sophisticated distances of classified taxonomic profiles to the expected distribution. Also, it would be interesting to see quantitative sensitivities and specificities (or F1-scores?) on the taxonomic assignments; particularly also when running on the exact same (more conservatively filtered) dataset for QIIME. Some numbers on specificity are provided in the Abstract and Conclusion sections – but I am not sure if specificity may be gained at the expense of sensitivity based on the more rigid read filtering upstream.\n\nAs a suggestion, but certainly not as a request, I would recommend to maybe include additional, independent datasets to benchmark on. For example, Tremblay et al. (2015) have published data on mock communities sequenced with different primer sets and on different platforms. Such data could contribute to a yet more general assessment of NG-Tax performance.\n\nMinor comments (chronologically, not in order of importance):\n\nBackground, “The consequence of this approach is that the ‘quality’ of the clustering of the reference set propagates to reference-picked OTUs.” I believe that as such, this statement is not fully valid or supported. In fact, the negative complement is arguably true: reference-based OTU picking against a “bad” reference can never provide “good” OTUs (a garbage-in, garbage-out problem, so to say). However, even with a good reference, a bad mapping algorithm can generate non-informative reference-based OTU sets. Schloss & Westcott have recently published a study which discusses this point, among others (Westcott & Schloss, 2015).\n\nBackground, “However, in 16S rRNA gene amplicon sequencing every sequencing error could potentially lead to the false discovery of a new species.” I have two comments on this statement. First, I believe that the term “species” in this context can be misleading and I feel that the neutral term OTU or diversity unit would be more appropriate. Second, there is a large body of literature on how sequencing errors affect 16S-based diversity studies beyond the cited Bokulich et al paper (starting from Kunin et al., 2010), and it would be worth to at least mention these, although an in-depth discussion would probably lead away from this study’s focus. Also, it may be worth mentioning recent algorithmic approaches to tackling this issue, such as DADA2 (Callahan et al., 2016).\n\nResults & Discussion, chimera filtering. The implemented method for chimera filtering appears a little ad hoc and heuristic, although the proposed approach certainly makes sense intuitively. However, given the long history of “chimera-slaying” algorithms and the quite sobering benchmark studies on them, some context would be helpful for the reader here – maybe even as a short supplement or as a reference to the user manual. For example, how is the proposed approach conceptually different from existing tools like UCHIME etc? And why was it implemented as is? What was the (empirical?) motivation to do it like this, not otherwise? Personally, I am not very convinced of the performance of chimera-filtering algorithms overall and several recent pipelines side-step the issue more or less elegantly. In the case of NG-Tax (or other reference-based OTU callers), one could even argue that if the reference database is perfectly chimera-free, a closed-reference approach would not need a chimera filtering approach at all, or only one which is based on differential mapping of a sequence to two (highly unrelated) OTUs.\n\nTable 1 is very large and (on the PDF) unfortunately rotated by 90 degrees. I suggest to convert it into a supplemental Excel sheet which would be more reader-friendly.\n\nFigure 2 has rotated horizontal axis labels, a 90deg rotated legend – maybe that’s just due to formatting of the PDF. It is also difficult to read taxonomic names on the vertical axis in all-caps.\n\n“Consequently, these methods are more powerful than purely OTU-based methods, […].” While I agree with this sentence to a certain extent, I believe that the statement should be supported by referring to previous work on the topic. It is not necessarily consensus that 16S “sequence often correlates with phenotypic similarity in key features”, but it is even less clear to what extent phylogenetic diversity estimators capture this signal in a useful way. Arguably, a PD-estimator of UniFrac can only be as good as their underlying tree, which in turn is based on the (representative) sequences of OTUs and thus depends on many factors in the background.\n\nIn particular, the weighted UniFrac measure used in this study seems to be more sensitive to quite a number of factors (including sequencing errors and inflation of small clusters, not irrelevant for the points made in this study) than its unweighted sister in my personal experience, and according to a number of researchers I have talked to on this point. However, since “personal experience” and “people I’ve talked to” are certainly not a dependable scientific source, and because performance on mock communities should not be severely impacted, I would formulate this as a suggestion and certainly not as a reviewer’s request: were the weighted UF-based results double-checked using unweighted UF and/or a non-phylogenetic method, such as Bray-Curtis?\n\nIn the PCoA (Figure 5, A&C), it is quite hard to decide which method looks “better” purely based on visual impression, not least because the % variance explained on the axes is not equivalent. It would be good to see a more quantitative statement on which approach better recovers expected clusters from the mock communities. The most straightforward approach would be to perform MANOVA analyses, structured by the different factors to test for and then use the effect sizes to quantify the goodness of separation (or non-separation). I would suggest to run e.g. Anderson’s PERMANOVA (http://www.entsoc.org/PDF/MUVE/6_NewMethod_MANOVA1_2.pdf; implementation available through the function “adonis” in the R package vegan) or ANOSIM to this end. Alternatively, samples could be clustered based on beta div and the resulting clusterings (or dendrograms) quantitatively compared to expectations based on different factors.\n\nThank you for providing Supplementary Figures 1&2; they are informative in the interpretation of the presented data.\n\nSimilarly, thank you for providing code and data as supplements!", "responses": [ { "c_id": "4231", "date": "02 Jan 2019", "name": "Javier Ramiro-Garcia", "role": "Author Response", "response": "In their manuscript, the authors introduce NG-Tax, an open-source software for the (meta-)analysis of 16S rRNA-based microbiome datasets. Their tool focuses on an important and so-far arguably understudied aspect of microbial ecology research: the integration of results across studies, in view of both technical and biological variation.The approach is interesting and addresses important points. In particular, several sequencing datasets of different mock communities were generated, even using different primer sets: this is great data to benchmark on, and many (most) other papers introducing tools do not provide benchmarks on such an array of real (mock) data. In general, I feel that this is very interesting work and that NG-Tax can be a promising alternative to existing tools in the field.We thank the reviewer for his nice comments and also his suggestions about the manuscript.However, there are several points that I feel would need to be addressed in order for the manuscript to stand tall, and for the reader to get a good understanding of how NG-Tax can be useful in practice.Major comments: Even after reading the manuscript and online user manual repeatedly, I have to admit that it is not completely clear to me how NG-Tax works in detail, and in which points exactly it differs from existing approaches. Based on the introduction, I gather that NG-Tax relies on closed-reference OTU picking, but this is not mentioned explicitly anywhere in the text. Also, does reference-based OTU picking in NG-Tax rely on uclust? If yes, which version and parameters were used, and how do they differ from QIIME’s defaults? Also, the Background and Discussion sections do not elaborate on the various disadvantages of closed-reference approaches; most importantly, closed-ref only takes into account sequences matching the database and removes everything else. When integrating sequence data from different primer sets, this is arguably the most straightforward approach; however, the limitations should be discussed. Thanks for the suggestion. In the new version we included a more detailed Figure 1 including those unique aspects of NG-Tax.We agree with the reviewer that close reference OTU picking has the disadvantage of only taking sequences into account that have a match in the database, and this is incompatible with having stable OTUs since databases change over time. For this reason, NG-Tax employs an open reference approach to remain independent of reference databases.Different clustering algorithms also lead to different OTUs, and hence no clustering process is applied and the generation of OTUs is independent for each sample. Existing open reference approaches generate OTUs for the whole study by clustering the reads from all the samples together. Then, if new samples are included to a previous study, the OTUs need to be regenerated with the reads from the previous study and new samples together, which will lead to discrepancies in the former and new composition of the samples because some of the previous OTUs may not be present in the new analysis anymore.Instead, in NG-Tax OTUs are generated sample by sample using the following strategy:For each sample reads are ranked by read abundance and OTUs are added to an OTU table starting from the most abundant sequence until the read abundance is lower than a percentage defined by the user (recommended is at is 0.1%). Subsequently, the discarded reads are clustered to the OTU table allowing one mismatch.In practical terms it is guided clustering where seeds are determined by abundance. The differences with an normal clustering approach is that there is no clustering to define the seeds, which allows seeds that differ as little as one nucleotide. The clustering is applied only afterwards to compensate for potential bias due to PCR and sequencing errors. Error is sequence-specific, and hence some sequences could be affected more than others. If a species specific amplicon is more prone to PCR or sequencing errors, the relative abundance of that particular OTU will be underestimated. But after clustering, OTUs more prone to error receive a higher percentage of discarded reads than others, this differential recovery helps to reestablish the true composition that was lost due to sequence specific error rates.We substituted uclust by usearch in the scripts of the new version. I gather from the text that NG-Tax’s main innovations are the use of primer-tailored reference databases and a different (more conservative) read abundance filtering scheme. It is perfectly valid to benchmark these against QIIME’s default settings; however, it would be great to see how QIIME performs with similarly conservative settings, to better understand where NG-Tax’s edge in performance comes from. We think that the main innovation of NG-Tax is the way OTUs are generated. This may seem counter-intuitive because it does not follow the standard approach but it is the discerning step compared with other existing pipelines. This innovative OTU generation algorithm is the reason of the NG-Tax’s edge in performance.With QIIME those conservative thresholds cannot be used because the filtering percentage threshold is defined using the whole library and within a library there are samples that contain 20 times more reads than others. A conservative threshold like 0.1% is conservative for an average sample, not conservative for a big sample at all and extreme for small samples. Hence, OTUs present in only small samples can be discarded even if they represent 1% of that sample but less than 0.1% of the whole dataset. On the other hand, NG-Tax applies thresholds defined by sample accounting for sample heterogeneity.In the manuscript we used the setting recommended by QIIME and described in Bokulich et al 2013 1. For NG-Tax analysis we also employed recommended default settings so we thought that even if this is not optimal and has its limitations, this could be a fair approach. Nevertheless, we benchmarked with QIIME not to compare performances but rather to show that this dataset is not an exceptional case and the commonly reported problems such as many un- or poorly classified OTUs, inflated richness and diversity, taxonomic profiles that do not match the expected ones, region dependent taxonomic classification and results being highly dependent on minor changes in the experimental setup are also found in this dataset when standard approaches are used.In Bokulich et al 2013, supplementary material 2; pages 8, 9 and 10, the authors compare expected composition against real sample composition using different parameters, one of them being OTU abundance, and the plot shows that the obtained profiles do not change much using different abundance thresholds.Nonetheless, as suggested by the reviewer we reproduced the QIIME analysis with a 0.1% abundance threshold and this is included in the supplementary data. The results using 0.1% or 0.005% are consistent. Regarding taxonomy assignments, it is valid to compare NG-Tax’s uclust-based approach to QIIME’s uclust-based approach. However, I believe that the gold standard continues to be the RDP Classifier, and it would be interesting to see a performance comparison to this tool (on the short-read data, not only on full-length reads). Also, how does taxonomic classification by NG-Tax differ conceptually from RTAX (http://www.uio.no/english/services/it/research/hpc/abel/help/software/rtax.html)? I do believe that they are not equivalent, but the approaches appear somewhat related. In the manuscript we wanted to show that taxonomy assignment using short reads should be comparable with the assignment using the complete 16S rRNA gene (Figure 2). This is why we employed full length sequences. We could have included also RDP short read based taxonomy but the reads were too short for RDP, and hence genus and many times even family assignment could not be achieved with a minimum threshold value of 50%. In the supplementary data we supplied the theoretical compositions for all mock communities. The files for MC2 V4 and MC2 V5V6 contain all phylotypes and can be uploaded to the RDP classifier to verify the poor performance. In the new manuscript we substituted RDP for SILVA Incremental Aligner (SINA) to classify the full length sequences and we also updated the database in NG-Tax to SILVA 128, improving in both cases the classification.I read the manuscript suggested by the reviewer and I can say that NG-Tax taxonomic classification is very similar to rtax.The NG-Tax classifier works as follows:For each OTU, usearch is used to retrieve hits for the forward and reverse reads against their respective trimmed reference database. Hits that are common between both reads are divided in 6 identity thresholds 100, 98, 97, 95, 92, 90. A hit belongs to a certain level, for example 97, when both reads have at least a 97 percentage identity with that hit. Using the highest available identity threshold, NG-Tax assigns the consensus taxonomy to the OTU if this taxonomy is supported for at least half of the hits. Genus, Family or Order remains unassigned if the maximum identity percentage level is lower or equal to 97%, 95% and 92% respectively. These are the main differences:rtax clusters the reference database at 99%, while NG-Tax does not.rtax averages the percentage identity for both reads and then considers the hits that have an averaged percentage identity 0.5% lower than the maximum averaged percentage identity as valid. NG-Tax does not average the percentage identities and uses fixed values 100, 98, 97, 95, 92 and 90 as thresholds.For the rest they are indeed very similar approaches.Therefore we have added rtax to the references and acknowledge in the manuscript that similar dynamic identity thresholds have been already employed to assign taxonomy. Furthermore, all the details about NG-Tax taxonomic assignment have been added to the user manual. In general, the results on taxonomic classification are not discussed quantitatively. From Figures 3&4, the visual impression is that NG-Tax indeed better approximates expected taxonomic profiles than QIIME, but it is hard to quantify this from stacked bar charts. I would suggest to compute e.g. Euclidean or more sophisticated distances of classified taxonomic profiles to the expected distribution. Also, it would be interesting to see quantitative sensitivities and specificities (or F1-scores?) on the taxonomic assignments; particularly also when running on the exact same (more conservatively filtered) dataset for QIIME. Some numbers on specificity are provided in the Abstract and Conclusion sections – but I am not sure if specificity may be gained at the expense of sensitivity based on the more rigid read filtering upstream. As suggested by the reviewer, distances between compositional profiles and expected profiles are now shown in Figure 5. Distances between taxonomic profiles were calculated as the sum of the weighted differences. Given two taxonomical profiles x and y, for each taxa i, we defined the difference in abundance as difi(x,y)=( xi –yi) and a weighing factor wi as wi(x,y)=( xi –yi )/avg(xi + yi). Weighted difference was the result of multiplying the difference in abundance by its weighting factor. This weighting factor is useful to take into account the relative change and not only the absolute change, because a 1% absolute change becomes a 200% or 20% relative change depending on whether the expected abundance is 0.5% or 5% respectively. We performed t tests to compare the performance of NG-Tax versus QIIME from a quantitative point of view.We have also included an Excel spreadsheet with compositional profiles in the supplementary data.Figure 2 shows specificity of the taxonomical assignments and has been has been modified to improve readability.The QIIME analysis at 0.1% abundance threshold can be found in the supplementary material. As a suggestion, but certainly not as a request, I would recommend to maybe include additional, independent datasets to benchmark on. For example, Tremblay et al. (2015) have published data on mock communities sequenced with different primer sets and on different platforms. Such data could contribute to a yet more general assessment of NG-Tax performance. We thank the reviewer for the suggestion, but including new datasets will imply a rewrite of a big part of the manuscript. We consider that 49 samples can give an idea of NG-Tax performance. Additionally, we would like to mention that NG-Tax has been the reference method for 16S rRNA gene amplicon analysis in our lab for more than two years and has been used in more than 30 manuscripts that have been submitted or are in preparation. One of these manuscripts 2 was published before this manuscript. Since then another fifteen studies using NG-Tax have been published 3-17. These studies contain biological samples that belong to very different and specific environments and were sequenced both on MiSeq and HiSeq instruments. These will contribute to the assessment of NG-Tax performance, however these were not included in the current manuscript since they are accessible on the aforementioned publications.Minor comments (chronologically, not in order of importance): Background, “The consequence of this approach is that the ‘quality’ of the clustering of the reference set propagates to reference-picked OTUs.” I believe that as such, this statement is not fully valid or supported. In fact, the negative complement is arguably true: reference-based OTU picking against a “bad” reference can never provide “good” OTUs (a garbage-in, garbage-out problem, so to say). However, even with a good reference, a bad mapping algorithm can generate non-informative reference-based OTU sets. Schloss & Westcott have recently published a study which discusses this point, among others (Westcott & Schloss, 2015). With this sentence we did not imply that only ‘good quality’ is transferred from the clustered databases to the OTUs, we meant both, pros and cons are transferred. In fact, we agree that references have their limitations and clustered databases also contain bias due to clustering. For this reason NG-Tax employs a de novo OTU picking with no references or clustering involved. Background, “However, in 16S rRNA gene amplicon sequencing every sequencing error could potentially lead to the false discovery of a new species.” I have two comments on this statement. First, I believe that the term “species” in this context can be misleading and I feel that the neutral term OTU or diversity unit would be more appropriate. Second, there is a large body of literature on how sequencing errors affect 16S-based diversity studies beyond the cited Bokulich et al paper (starting from Kunin et al., 2010), and it would be worth to at least mention these, although an in-depth discussion would probably lead away from this study’s focus. Also, it may be worth mentioning recent algorithmic approaches to tackling this issue, such as DADA2 (Callahan et al., 2016). As suggested by the reviewer we rephrased the sequence to avoid the use of “species”. Now we stated: “However, in 16S rRNA gene amplicon sequencing every sequencing error could potentially lead to an incorrect OTU classification which may ultimately lead to the false discovery of a new phylotype”We added Kunin et al 2010 and Callahan et al. 2016 to the references but we just wanted to point out that sequencing error is an important factor in 16S analysis rather than make an in-depth discussion about it. Results & Discussion, chimera filtering. The implemented method for chimera filtering appears a little ad hoc and heuristic, although the proposed approach certainly makes sense intuitively. However, given the long history of “chimera-slaying” algorithms and the quite sobering benchmark studies on them, some context would be helpful for the reader here – maybe even as a short supplement or as a reference to the user manual. For example, how is the proposed approach conceptually different from existing tools like UCHIME etc? And why was it implemented as is? What was the (empirical?) motivation to do it like this, not otherwise? Personally, I am not very convinced of the performance of chimera-filtering algorithms overall and several recent pipelines side-step the issue more or less elegantly. In the case of NG-Tax (or other reference-based OTU callers), one could even argue that if the reference database is perfectly chimera-free, a closed-reference approach would not need a chimera filtering approach at all, or only one which is based on differential mapping of a sequence to two (highly unrelated) OTUs. First, we would like to recall that NG-Tax is not reference-based.We fully agree with the reviewer opinion about ‘chimera-slaying’ algorithms. Chimera detectors are often validated using in-silico datasets generated by determining an initial set of valid sequences and a chimera formation pattern. This pattern or “rule” for chimera formation is commonly defined by considering that any two sequences in the initial dataset are equally probable to lead to a chimera and any nucleotide is equally probable to be the point in which these two sequences merge to form the chimera. It is conceivable that maybe the initial set is not representative of the sequences present in a specific real biological sample, not every pair of sequences has the same probability to form chimeras and not all the nucleotides may have the same odd to be the merging point of two sequences. Many different sequence sets can be selected as initial valid sequences and also many different chimera formation patterns can be chosen, but it is very difficult to really determine whether our choices mimic the way in which chimeras are formed in real sequencing data and therefore it is hard to verify if those in-silico created chimeras represent the chimeras that can be found in real sequencing samples. We consider that the proper validation should be using the real sequencing samples. If the chimera detection algorithm works, we would expect a very small number of non-assigned reads (since most chimeras should be aberrant). In case we have positive controls like MC, sequencing profiles and diversity should resemble the expected ones, and this is exactly what we observe with the results of NG-Tax.We think that there are no perfect chimera-free databases, and a valid OTU can be found in the reference database and at the same time be a perfect combination of 2 other OTUs, especially for regions with lower variability (V4). If all those 3 OTUs are present in the same sample, how can we know whether it is a chimera or real?In our opinion chimera detection is the weakest step in 16S pipelines because there is no satisfactory solution to the problem mentioned above. So the prevention against chimeras should come from the experimental design by reducing the PCR cycles and selecting regions of high variability. Chimera removal has many limitations and human supervision is recommended. For this reason, we decided to simplify the chimera detection as much as possible so the researcher can quickly identify why an OTU has been discarded. And also apply stringent parameters (100% identity) to avoid false positives. False negatives should be easier to detect afterwards since most of the chimeras should be aberrant.In the manuscript of UCHIME they stated that “UCHIME searches for a chimeric alignment between a query sequence (Q) and two candidate parents (A and B)” and “candidate parents are required to have abundance at least λ times that of the query sequence, on the assumption that a chimera has undergone fewer rounds of amplification and will therefore be less abundant than its parents. The parameter λ is called the abundance skew, and by default λ=2 “, so NG-Tax approach is very similar to de novo UCHIME approach, but NG-Tax treats forward and reverse reads separately. Table 1 is very large and (on the PDF) unfortunately rotated by 90 degrees. I suggest to convert it into a supplemental Excel sheet which would be more reader-friendly. As suggested by the reviewer Table 1 is now supplied as an excel spreadsheet in the supplementary material Figure 2 has rotated horizontal axis labels, a 90deg rotated legend – maybe that’s just due to formatting of the PDF. It is also difficult to read taxonomic names on the vertical axis in all-caps. As suggested by the reviewer we modified Figure 2 to increase readability. “Consequently, these methods are more powerful than purely OTU-based methods, […].” While I agree with this sentence to a certain extent, I believe that the statement should be supported by referring to previous work on the topic. It is not necessarily consensus that 16S “sequence often correlates with phenotypic similarity in key features”, but it is even less clear to what extent phylogenetic diversity estimators capture this signal in a useful way. Arguably, a PD-estimator of UniFrac can only be as good as their underlying tree, which in turn is based on the (representative) sequences of OTUs and thus depends on many factors in the background. Taxonomic assignment of the OTUs suffers from the same problems raised by the reviewer. Not always does 16S rRNA gene sequence similarity correlate with phenotypic similarity and the taxonomical assignment is as good as the reference database and the classifier employed. But having a composition barplot with OTUs named by number rather than by taxonomy would mean that all the information provided by the nucleotide sequence is discarded. This information may not be perfect but we cannot neglect that this information transformed into taxonomical assignment is useful at least to some extent.The same criterion was applied to evaluate diversity. We used phylogenetic methods, which retain the information of the nucleotide sequence. We acknowledge the limitations but we argue that a sample containing 5 OTUs with a 99% pairwise sequence identity should not be given the same (potential) diversity that a sample containing 5 OTUs with less than 85% pairwise sequence identity. We consider that phylogenetic methods are more powerful because they use all information available, however, we should not over extrapolate the results. 16S rRNA gene amplicon sequencing should be taken as exploratory approach, whereas metagenomic and metatranscriptomic sequencing provides a more suitable and precise approach if we really want to focus on microbial functionality. In particular, the weighted UniFrac measure used in this study seems to be more sensitive to quite a number of factors (including sequencing errors and inflation of small clusters, not irrelevant for the points made in this study) than its unweighted sister in my personal experience, and according to a number of researchers I have talked to on this point. However, since “personal experience” and “people I’ve talked to” are certainly not a dependable scientific source, and because performance on mock communities should not be severely impacted, I would formulate this as a suggestion and certainly not as a reviewer’s request: were the weighted UF-based results double-checked using unweighted UF and/or a non-phylogenetic method, such as Bray-Curtis? As suggested by the reviewer we have included the unweighted UniFrac and Bray-Curtis analysis in the supplementary material. The results obtained by all three methods are in concordance. In the PCoA (Figure 5, A&C), it is quite hard to decide which method looks “better” purely based on visual impression, not least because the % variance explained on the axes is not equivalent. It would be good to see a more quantitative statement on which approach better recovers expected clusters from the mock communities. The most straightforward approach would be to perform MANOVA analyses, structured by the different factors to test for and then use the effect sizes to quantify the goodness of separation (or non-separation). I would suggest to run e.g. Anderson’s PERMANOVA (http://www.entsoc.org/PDF/MUVE/6_NewMethod_MANOVA1_2.pdf; implementation available through the function “adonis” in the R package vegan) or ANOSIM to this end. Alternatively, samples could be clustered based on beta div and the resulting clusterings (or dendrograms) quantitatively compared to expectations based on different factors. Thanks for the suggestion. In the new version of the manuscript we performed a more quantitative analysis of the sequencing data and the expected MC. We performed a permanova analysis under MC type factor and it was significant for both pipelines meaning that some of the variance is explained by the Mock type. But to really evaluate accuracy and reproducibility and compare pipelines performances we used pairwise distances and t tests (Figure 7 and Dataset 1). Thank you for providing Supplementary Figures 1&2; they are informative in the interpretation of the presented data. Thank you. Similarly, thank you for providing code and data as supplements! Thank you.References:1. Bokulich NA, Subramanian S, Faith JJ, et al. Quality-filtering vastly improves diversity estimates from Illumina amplicon sequencing. Nat Methods 2013;10:57-9.2. Timmers PH, Widjaja-Greefkes HC, Ramiro-Garcia J, et al. Growth and activity of ANME clades with different sulfate and sulfide concentrations in the presence of methane. Front Microbiol 2015;6:988.3. Giatsis C, Sipkema D, Ramiro-Garcia J, et al. Probiotic legacy effects on gut microbial assembly in tilapia larvae. Sci Rep 2016;6:33965.4. Atashgahi S, Lu Y, Ramiro-Garcia J, et al. Geochemical Parameters and Reductive Dechlorination Determine Aerobic Cometabolic vs Aerobic Metabolic Vinyl Chloride Biodegradation at Oxic/Anoxic Interface of Hyporheic Zones. Environ Sci Technol 2017;51:1626-1634.5. Atashgahi S, Lu Y, Zheng Y, et al. Geochemical and microbial community determinants of reductive dechlorination at a site biostimulated with glycerol. Environ Microbiol 2017;19:968-981.6. Azman S, Khadem AF, Plugge CM, et al. Effect of humic acid on anaerobic digestion of cellulose and xylan in completely stirred tank reactors: inhibitory effect, mitigation of the inhibition and the dynamics of the microbial communities. Appl Microbiol Biotechnol 2017;101:889-901.7. Dieho K, van den Bogert B, Henderson G, et al. Changes in rumen microbiota composition and in situ degradation kinetics during the dry period and early lactation as affected by rate of increase of concentrate allowance. J Dairy Sci 2017;100:2695-2710.8. Lu Y, Ramiro-Garcia J, Vandermeeren P, et al. Dechlorination of three tetrachlorobenzene isomers by contaminated harbor sludge-derived enrichment cultures follows thermodynamically favorable reactions. Appl Microbiol Biotechnol 2017;101:2589-2601.9. van Lingen HJ, Edwards JE, Vaidya JD, et al. Diurnal Dynamics of Gaseous and Dissolved Metabolites and Microbiota Composition in the Bovine Rumen. Front Microbiol 2017;8:425.10. Paulo LM, Ramiro-Garcia J, van Mourik S, et al. Effect of Nickel and Cobalt on Methanogenic Enrichment Cultures and Role of Biogenic Sulfide in Metal Toxicity Attenuation. Front Microbiol 2017;8:1341.11. van Gastelen S, Visker M, Edwards JE, et al. Linseed oil and DGAT1 K232A polymorphism: Effects on methane emission, energy and nitrogen metabolism, lactation performance, ruminal fermentation, and rumen microbial composition of Holstein-Friesian cows. J Dairy Sci 2017;100:8939-8957.12. Gerritsen J, Hornung B, Renckens B, et al. Genomic and functional analysis of Romboutsia ilealis CRIBT reveals adaptation to the small intestine. PeerJ 2017;5:e3698.13. van der Waals MJ, Pijls C, Sinke AJC, et al. Anaerobic degradation of a mixture of MtBE, EtBE, TBA, and benzene under different redox conditions. Appl Microbiol Biotechnol 2018;102:3387-3397.14. Umanets A, de Winter I, F IJ, et al. Occupancy strongly influences faecal microbial composition of wild lemurs. FEMS Microbiol Ecol 2018;94.15. Steinert G, Gutleben J, Atikana A, et al. Coexistence of poribacterial phylotypes among geographically widespread and phylogenetically divergent sponge hosts. Environ Microbiol Rep 2018;10:80-91.16. Gu F, Borewicz K, Richter B, et al. In Vitro Fermentation Behavior of Isomalto/Malto-Polysaccharides Using Human Fecal Inoculum Indicates Prebiotic Potential. Mol Nutr Food Res 2018;62:e1800232.17. Dat TTH, Steinert G, Thi Kim Cuc N, et al. Archaeal and bacterial diversity and community composition from 18 phylogenetically divergent sponge species in Vietnam. PeerJ 2018;6:e4970." } ] }, { "id": "15175", "date": "02 Aug 2016", "name": "Julien Tremblay", "expertise": [], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper describes a pipeline for processing 16S rRNA amplicon data. They implemented an experimental design in which they used data coming from three different HiSeq2000 runs using two variable regions (V4 and V5-V6). It is however not clear if their data has been generated in-house or if their data was actually coming from public databases. This should be explicitly stated somewhere (unless I missed it). Using this data as input, the authors developed a pipeline labeled NG-Tax, which according to them: 1) better accounts (compared to what?) for errors associated with a range of technical aspects of 16S rRNA amplicon sequencing and 2) improves comparability be removing technical bias and facilitating efforts towards standardization. In my view, the problem is that why their pipeline does 1) and 2) is not addressed in depth. The description of the technical aspects of their pipeline in the first part of the result section only very summarily describes the general workflow of the pipeline, but nowhere do they describe how exactly OTU picking is done (see comment below). How exactly Chimera are detected? With an open-source package? In-house script? Taxonomic assignment methodology is unclear as well. The authors state that they are using uclust for taxonomic assignment, while uclust is a sequence clustering software (also see comments below).\nThen the authors compares their pipeline results with the ones generate by Qiime with default paramters. Qiime with its default parameters is already known to not perform optimally (See UPARSE paper, Edgar, 2013). I think that comparing with Qiime for validation is okay, but do not spend too much time dissecting the results. What the authors should focus on is, I think, on improving substantially on the technical description of their pipeline – describe each step in details. If open source packages are being used, say so, if not, describe your script/software/algorithm. Also please make the source code available under a code repository (Github or Bitbucket for instance).\nIn my view the paper is not acceptable in its current form.\n\nSpecific comments:\nAt the sentence \"mostly because available 16S rRNA gene reference databases were thought to provide insufficient coverage13–16.\" Can you please elaborate on that? What do exactly mean by that?\n\n\"there still is no standard or consensus of best choices for variable regions.\"\nI don't fully agree with this. Depending on your field of study, a certain consensus can usually be found. For instance, the Earth Microbiome project recommends two primer sets (V4 and the 'newer' V4-V5) - Many labs investigating soil or environmental samples in general will effectively favor these primers because they are being used by a large part of the community which readily enables inter-lab community/study comparisons.\n\nConcerning the OTU picking section: It is not clear how exactly you pick your OTUs. Basically, you are kind of dereplicating/clustering your raw reads data set at 100% ID and then create a one column OTU table for each sample? Please clarify.\n\n“Phred score, such as minimum average Phred score, maximum number of ambiguous positions, maximum bad run length, trimming and minimum read length after quality trimming, are not utilized in NG-Tax because quality scores from the Illumina base caller have been shown to be of limited use for the identification of actual sequence errors for 16S rRNA gene amplicon studies9,37.”\nYes Q scores have their limitation, but it is unwise to not filter for reads containing Ns and reads of very poor Q scores. Some basic filtering should be implemented to at least filter for very bad data. For instance if you have a read with 10 bases with Q score lower than 10, this read should obviously be removed.\n\n“To speed up the procedure by several orders of magnitude, 16S rRNA gene sequences from the reference database are trimmed to contain only the region amplified by the primers.”\nPlease specify how you generated your trimmed database of 16S rRNA genes ref. In silico PCR? A multiple alignment that was trimmed at specific coordinates?\n\n\"In the current version of NG-Tax, taxonomy is assigned to OTUs utilizing the uclust algorithm16 and the Silva_111_SSU Ref database, containing 731,863 unique full length 16S rRNA gene sequences. To ensure maximum resolution and avoid the risk of errors due to clustering-associated flaws (e.g. reference sequence error hotspots, overrepresentation of certain species and lack of robustness in cluster formation by clustering algorithms),we use the non-clustered database. To speed up the procedure by several orders of magnitude\",\nUclust is for clustering sequences/reads and not for taxonomic assignment…? Taxonomic assignment is done by other means (RDP classifier), but certainly not with uclust.\n\nFor each OTU, a taxonomic assignment is retrieved at six different identity thresholds levels (100%, 98%, 97%, 95%, 92% and 90%) and at two taxonomic levels (genus and family).\nHow exactly are OTUs classified? With an in-house method? The RDP classifier? Please elaborate.\n\nFigure 1. Please add more details. Are you using open-source packages in your pipeline? If so please indicate.\n\nTable 1: Table 1 is heavy and not really meaningful. Would fit in more appropriately in suppl. material.\n\nFigure 3 and 4: Please find another way of displaying data of figure 3. It is simply not feasible to associate a color to a given bar graph. Maybe consider using a heatmap with hierarchical clustering or a PCA/PCoA? Typically for taxonomiy stacked barplots you can’t really go above 20 different colors. After that it becomes indistinguishable.\n\n\"Because the focus of NG-Tax is to retain as much biological signal as possible while minimizing the impact of any technical choice,\"\nBut how exactly does NG-Tax retain more biological signal than other pipelines, what does that mean?\n\nDiscussion: The authors say that their pipeline outperforms Qiime, but nowhere is discussed how exactly does Qiime works. How exactly does Qiime generate OTUs, how are the reads QCed? How is the classification performed, what training sets are being used for classification? It is already known that Qiime does not perform well with default parameters (see R. Edgar’s UPARSE paper), so Qiime does not represent a gold standard, especially with default parameters.\n\nNG-Tax pipeline availability. Please include the pipeline on a Github or bitbucket repository.", "responses": [ { "c_id": "4230", "date": "02 Jan 2019", "name": "Javier Ramiro-Garcia", "role": "Author Response", "response": "This paper describes a pipeline for processing 16S rRNA amplicon data. They implemented an experimental design in which they used data coming from three different HiSeq2000 runs using two variable regions (V4 and V5-V6). It is however not clear if their data has been generated in-house or if their data was actually coming from public databases. This should be explicitly stated somewhere (unless I missed it). To further clarify this we added the section Datasets:Datasets:Four synthetic communities of varying complexity were created, consisting of 16S rRNA gene amplicons of phylotypes (PTs) associated with the human GI-tract (Dataset 1). This specific setup limited the likelihood of overfitting to a particular OTU composition or distribution and allowed us to assess (1) the quantification potential, (2) noise floor and (3) the effect of richness and diversity on quality filtering parameters, thus ensuring a higher fidelity with biological samples than by using a single MC. As a reference, to assess the quality of the taxonomic classifications, full length sequences for all PTs were obtained through Sanger sequencing. Expected MCs were created by trimming the full length sequences to the sequenced region. MC1 and MC2 consisted of equimolar amounts of 17 and 55 PTs, respectively. MC3 contained 55 PTs in staggered concentrations typical for the human GI-tract, and MC4 included 50 PTs with relative abundances ranging between 0.001 and 2.49%. To account for pipetting errors, each of the four MCs was produced in triplicate. To design a pipeline that puts more focus on biology, these 12 MC templates were used to sequence the MCs with different conditions that cover most of the technical bias associated with 16S rRNA gene amplicon studies reported in literature. To this end, we: Targeted either region V4 or region V5-V6, Used four PCR protocols differing in the number of PCR cycles and reaction volumes. PCR products were analysed in three different sequencing runs and in seven different libraries. Two different library preparation protocols (with and without an additional amplification of 8 cycles) were applied (Dataset 1). In addition the sequencing depth ranged from 2363 to 335822 reads per sample (Dataset 1). One phylotype, PT17 (Parabacteroides), attracted so much sequencing error in the V4 region that it was rendered undetectable although it was amplified by the primers (Supplementary Figure 1). Therefore, to test both pipelines without this sequencing anomaly, it was removed from the analysis.In this section we explain how we created and sequenced the MCs. The sequencing data was generated by a sequencing company (GATC, Constance, Germany; see section Materials and Methods). The sequencing data has been submitted to the ENA repository, and we added the following sequence data availability section:Sequence data availability:Sequence data have been deposited in the European Nucleotide Archive46, accession number [ENA:PRJEB11702]) http://www.ebi.ac.uk/ena/data/view/PRJEB11702 (amplicon sequencing data for all 49 samples) and [ENA:LN907729-LN907783]) (full length 16S rRNA gene sequences for all 55 Pts).”Using this data as input, the authors developed a pipeline labeled NG-Tax, which according to them: 1) better accounts (compared to what?) for errors associated with a range of technical aspects of 16S rRNA amplicon sequencing and 2) improves comparability be removing technical bias and facilitating efforts towards standardization. In my view, the problem is that why their pipeline does 1) and 2) is not addressed in depth.We agree with the reviewer that the highlighted elements were not sufficiently clear and lacked an explanation why we believe that NG-Tax performs better. Therefore we replaced this sentence with:”This allowed for the development of NG-Tax, a pipeline that accounts for biases associated with this range of technical aspects associated with 16S rRNA gene amplicon sequencing. Therefore NG-Tax will improve comparability by removing technical bias and facilitate efforts towards standardization, by focusing on reproducibility as well as accuracy. To assess the performance regarding key output parameters such as taxonomic classification, composition and richness, and α and β diversity measures, we benchmarked the results obtained with NG-Tax.”In order to account for errors and increase comparability by removing technical bias from 16S rRNA amplicon studies, NG-Tax should fulfill the following requirements: Taxonomy assignment using short reads should be comparable with the assignment using the complete 16S rRNA gene. Composition profiles based on sequencing data should resemble the real composition of the biological sample. α and β diversity should match the expected α and β diversity. Results should be reproducible and therefore robust against biological variation (different sample compositions) and technical (PCR and sequencing settings) biases. We consider that these requirements were met by NG-Tax, as supported by the following data.Figure 2 shows the high similarity of the taxonomic classification of the V4 and V5V6 amplicon results compared to full length sequences using SILVA Incremental Aligner (SINA). The specificity and the number of hits testify to the reliability of the assignments.Table 1 shows the low number and percentage of spurious reads.Figure 3 shows that NG-Tax derived compositional profiles based on sequencing data accurately resemble the expected profiles.Figure 5 quantifies the distances to the expected profiles.Figure 6 & 7: the PCoA plots show that MCs group by type, despite all technical bias associated with 16S rRNA gene amplicons sequencing, such as PCR settings, and primer or region selection. Figure 7 shows that all within-MC pairwise comparisons and the dispersion of all pairwise comparisons are significantly smaller in NG-Tax meaning that distances within and between MC types are robust. These results could not have been achieved without a proper reduction of the aforementioned biases. This will improve comparability by enabling direct comparison between studies even when using slightly different approaches.The description of the technical aspects of their pipeline in the first part of the result section only very summarily describes the general workflow of the pipeline, but nowhere do they describe how exactly OTU picking is done (see comment below). How exactly Chimera are detected? With an open-source package? In-house script? Taxonomic assignment methodology is unclear as well. The authors state that they are using uclust for taxonomic assignment, while uclust is a sequence clustering software (also see comments below).In the revised manuscript we substantially increased the amount and detail of information on the description of the general work flow. All the critical steps, including barcode & primer filtering, OTU picking, mapping rejected reads to accepted OTUs, de novo chimera filtering, taxonomic assignment and the generation of a phylogenic tree are now detailed in Figure 1 and explained in the user manual.Then the authors compares their pipeline results with the ones generate by Qiime with default paramters. Qiime with its default parameters is already known to not perform optimally (See UPARSE paper, Edgar, 2013). I think that comparing with Qiime for validation is okay, but do not spend too much time dissecting the results. We agree with the reviewer’s view on the default parameters of QIIME, however, the major improvements are gained by not clustering and processing the reads per sample. Therefore, the presented results cannot be achieved with QIIME independent of the parameters we choose. Besides that, testing QIIME under different settings has been already extensively covered elsewhere 5 and if we would change parameters, reviewers could argue that our chosen parameters are less than optimal and therefore we stayed with the default settings.Nonetheless, as suggested by the reviewer we reproduced the QIIME analysis with a 0.1% abundance threshold, and this is now included in the supplementary data. The results using 0.1% or 0.005% are consistent and show no performance gain (“Supplementary data. QIIME beta-div results all settings”). This is also in line with the result shown by Bokulich et al 2013 1, supplementary material 2; pages 8, 9 and 10. This text includes a comparison of the expected composition against real sample composition using different filtering parameters. One of these parameters is OTU abundance and the plot shows that the obtained profiles do not change much using different filtering abundance thresholds.Although we agree with the reviewer that we need not to put too much emphasis on the QIIME results, they do show the consequences when technical bias is not adequately taken care of, which makes it easier for the non-technical reader to place the results achieved by NG-Tax into context.What the authors should focus on is, I think, on improving substantially on the technical description of their pipeline – describe each step in details. If open source packages are being used, say so, if not, describe your script/software/algorithm. Also please make the source code available under a code repository (Github or Bitbucket for instance).In my view the paper is not acceptable in its current form.We thanks the reviewer for the constructive suggestions and hope that the changes introduced in the manuscript and supplementary data help to change his opinion.As suggested by the reviewer the source code is now available on Github (https://github.com/JavierRamiroGarcia/NG-Tax.git).Specific comments:At the sentence \"mostly because available 16S rRNA gene reference databases were thought to provide insufficient coverage 13–16.\" Can you please elaborate on that? What do exactly mean by that?In the past, when pyrosequencing was the standard sequencing method, open reference base OTU picking was the common strategy for two main reasons: 1) reference databases were still very small and thought to provide insufficient coverage and 2) the more computationally intense open reference methods were used because the amount of reads generated were lower than nowadays. Over time databases were more complete and the amount of data generated increased the needed computational time, so close reference OTU picking gained popularity. Currently, with new bioinformatics solutions, open reference OTU picking is gaining ground and NG-Tax is following that trend by implementing a new open reference OTU picking algorithm.\"there still is no standard or consensus of best choices for variable regions.\"I don't fully agree with this. Depending on your field of study, a certain consensus can usually be found. For instance, the Earth Microbiome project recommends two primer sets (V4 and the 'newer' V4-V5) - Many labs investigating soil or environmental samples in general will effectively favor these primers because they are being used by a large part of the community which readily enables inter-lab community/study comparisons.We agree with the reviewer that there is a certain consensus for some projects, but still there are many publications addressing the differences in results when different primers are used. Therefore, when choosing a primer pair, whether these primers are used by the community becomes an important factor if afterwards the researcher wants to compare the results with existing studies. The idea of NG-Tax is to decrease the importance of this factor by providing comparable results across different primer sets, giving more freedom to the researcher to explore new possibilities. But as suggested by the reviewer we softened our statement and rephrased as:“There still is no complete consensus regarding best choices for variable regions even if some initiatives like the Earth Microbiome Project are setting standards that are increasingly being adopted by the field.”Concerning the OTU picking section: It is not clear how exactly you pick your OTUs. Basically, you are kind of dereplicating/clustering your raw reads data set at 100% ID and then create a one column OTU table for each sample? Please clarify.We tried to make the paper as readable as possible by not adding too much technical information. Realizing that we have excessively reduced technical detail in the original manuscript, in the new version all technical details can be found in the user manual. An indication that this information can be found in the user manual is now included in the manuscript.In NG-Tax OTUs are generated per sample using the following strategy:For each sample reads are ranked by abundance and OTUs are added to an OTU table starting from the most abundant sequence until the read abundance is lower than a percentage defined by the user (recommended is at is 0.1%). Subsequently, the discarded reads are clustered to the OTU table allowing one mismatch.In practical terms this is in fact guided clustering where seeds are determined by abundance. The difference with an normal clustering approach is that there is no clustering to define the seeds. This allows seeds that differ as little as one nucleotide. The clustering is applied only afterwards to compensate for potential bias due to PCR and sequencing errors. Error is sequence-specific, and hence some sequences could be affected more than others. If a species specific amplicon is more prone to PCR or sequencing errors, the relative abundance of that particular OTU will be underestimated. But after clustering, OTUs more prone to error receive a higher percentage of discarded reads than others, this differential recovery helps to reestablish the true composition that was lost due to sequence specific error rates.“Phred score, such as minimum average Phred score, maximum number of ambiguous positions, maximum bad run length, trimming and minimum read length after quality trimming, are not utilized in NG-Tax because quality scores from the Illumina base caller have been shown to be of limited use for the identification of actual sequence errors for 16S rRNA gene amplicon studies9,37.”Yes Q scores have their limitation, but it is unwise to not filter for reads containing Ns and reads of very poor Q scores. Some basic filtering should be implemented to at least filter for very bad data. For instance if you have a read with 10 bases with Q score lower than 10, this read should obviously be removed.We fully agree with the reviewer. A filtering process is needed, and this is already implemented in NG-Tax. The point is that it is not based in quality score but based on abundance. Illumina have reported that 95%-97% of the reads have Q>30 (http://www.illumina.com/documents/products/technotes/technote_Q-Scores.pdf). This 3 to 5 percent of reads with lower quality will contain reads for all the different phylotypes, and within phylotypes there will be reads with errors in different positions and with different base substitutions. This decreases the probability of having exactly the same erroneous read. Therefore we expect that any specific erroneous read should be in low abundance. Subsequently, when samples are filtered by discarding low abundance sequences, those low quality reads will be removed without the need to check for quality scores.In addition quality scores do not account for PCR errors since the base caller will give them very high scores, because according to the sequencer they are real sequences. In contrast, filtering by abundance is insensitive to the error source, and hence if the reads with PCR errors are in low abundance (especially if high fidelity taq polymerase is used), they will be also removed.A good example of how stringent quality thresholds can bias the results can be found in Bokulich et al 2013, supplementary material 2; pages 8, 9 and 101. “To speed up the procedure by several orders of magnitude, 16S rRNA gene sequences from the reference database are trimmed to contain only the region amplified by the primers.”Please specify how you generated your trimmed database of 16S rRNA genes ref. In silico PCR? A multiple alignment that was trimmed at specific coordinates?We thank the reviewer for the suggestion. This information is important and now it has been added to the user manual. NG-Tax applies an in silico PCR using the primers and a reference database given by the user. Degenerated primer positions are allowed and alternative primers with mismatches can be supplied.\"In the current version of NG-Tax, taxonomy is assigned to OTUs utilizing the uclust algorithm16 and the Silva_111_SSU Ref database, containing 731,863 unique full length 16S rRNA gene sequences. To ensure maximum resolution and avoid the risk of errors due to clustering-associated flaws (e.g. reference sequence error hotspots, overrepresentation of certain species and lack of robustness in cluster formation by clustering algorithms),we use the non-clustered database. To speed up the procedure by several orders of magnitude\",Uclust is for clustering sequences/reads and not for taxonomic assignment…? Taxonomic assignment is done by other means (RDP classifier), but certainly not with uclust.For each OTU, a taxonomic assignment is retrieved at six different identity thresholds levels (100%, 98%, 97%, 95%, 92% and 90%) and at two taxonomic levels (genus and family).How exactly are OTUs classified? With an in-house method? The RDP classifier? Please elaborate.First we wanted to inform that uclust has been substituted by usearch in the scripts for the second version of the manuscript.Any or at least most methods for taxonomic assignment contain two main steps. First, the read to be classified is linked to sequences in a reference database by sequence similarity, and then the taxonomic information of linked sequences, termed hits, is transferred to the sequence to be classified. Different methods can be used to perform the linking step. In our case we used usearch (previously uclust). We used dynamic thresholds to get hits at 6 different identity levels, after which the taxonomic information is transferred to the read of unknown taxonomy by the NG-Tax classifier algorithm. Similar dynamic thresholds are used by rtax 2.A description of how the NG-Tax classifier works:For each OTU, usearch is used to retrieve hits for the forward and reverse reads against their respective trimmed reference database. Hits that are in common between both reads are divided in 6 identity thresholds 100, 98, 97, 95, 92, 90. A hit belongs to a certain level, for example 97, when both reads have at least a 97 percentage identity with that hit. Using the highest available identity threshold, NG-Tax assigns the consensus taxonomy to the OTU if this taxonomy is supported for at least half of the hits. Genus, Family or Order remains unassigned if the maximum identity percentage level is lower or equal to 97%, 95% and 92% respectively. The levels lower than 97% are only useful for unexplored environments; otherwise most of the OTUs are assigned at 100% identity.As suggested by the reviewer we have included a detailed explanation of how the algorithm works in the user manual provided in the supplementary files.Figure 1. Please add more details. Are you using open-source packages in your pipeline? If so please indicate.We thank the reviewer for the suggestion, now we added a figure with more details. As stated in the user manual we use USEARCH and QIIME.Table 1: Table 1 is heavy and not really meaningful. Would fit in more appropriately in suppl. material.As suggested by the reviewer the table 1 is now supplied as supplementary material.Figure 3 and 4: Please find another way of displaying data of Figure 3. It is simply not feasible to associate a color to a given bar graph. Maybe consider using a heatmap with hierarchical clustering or a PCA/PCoA? Typically for taxonomy stacked barplots you can’t really go above 20 different colors. After that it becomes indistinguishable.With figure 3 and 4 we just wanted to show in one glance, how close the sample compositions resembled the expected composition and how many different taxa are found in the data. In the new version of the manuscript we have added boxplots showing distances to the expected profiles to improve interpretation. An excel file with taxonomic profiles is also added to the supplementary material for further interpretation.PCoA plots showing distances between samples and expected for both pipelines are provided in figure 6. Figure 7 shows those distances as pairwise comparisons.\"Because the focus of NG-Tax is to retain as much biological signal as possible while minimizing the impact of any technical choice,\"But how exactly does NG-Tax retain more biological signal than other pipelines, what does that mean?We agree with the reviewer that the sentence is confusing. Therefore it has been rephrased.”Therefore, these indices probably provide a better estimate of the true diversity for data generated by high throughput next generation technology sequencers.Because the aim of NG-Tax is to enhance the biological signal as much as possible by minimizing the impact of any technical choice, divergence-based α-diversity (Phylogenetic Diversity (PD) [41]) and β-diversity (Unifrac [39]) metrics were used to visualize the diversity within and between MCs (Figure 6)”.Discussion: The authors say that their pipeline outperforms Qiime, but nowhere is discussed how exactly does Qiime works. How exactly does Qiime generate OTUs, how are the reads QCed? How is the classification performed, what training sets are being used for classification? It is already known that Qiime does not perform well with default parameters (see R. Edgar’s UPARSE paper), so Qiime does not represent a gold standard, especially with default parameters.In our manuscript we applied recommended settings like those described in the Bokulich paper. This paper extensively describes QIIME, the rationale behind the recommendations and the way that these choices impact the data. The scope of this manuscript was not to test QIIME under different settings. For NG-Tax analysis we also employed default and recommended settings so we thought that even if it is not optimal and has limitations, this could be a fair approach.We also analyzed the MCs with QIIME to show that this dataset is not an exceptional case with regards to the commonly reported problems (such as many un- or poorly classified OTUs, inflated richness and diversity, taxonomic profiles that do not match the expected ones, region dependent taxonomic classification and results which are highly dependent on minor changes in the experimental procedures) are also found in this dataset. So in our mind the QIIME analysis should primarily be seen as a performance comparison. In fact we encourage researchers to use more than one method, as this will increase the amount of information they can obtain from their datasets and determine the quality of their data. This will benefit their research and by extension the whole field.Nevertheless in an effort to increase comparability we also performed an additional analysis using QIIME with a 0.1% abundance threshold (which is conservative compared to the advised setting of 0.005%). Nevertheless this did still not reproduce the biological signal and the results obtained with 0.1% or 0.005% are consistent. These analyses have been added to the supplementary material as “Supplementary data. QIIME beta-div results all settings”.NG-Tax pipeline availability. Please include the pipeline on a Github or bitbucket repository.NG-Tax scripts were previously available as supplementary material, and as suggested by the reviewer they are now also available in Github (https://github.com/JavierRamiroGarcia/NG-Tax.git)References1. Bokulich NA, Subramanian S, Faith JJ, et al. Quality-filtering vastly improves diversity estimates from Illumina amplicon sequencing. Nat Methods 2013;10:57-9.2. Soergel DA, Dey N, Knight R, Brenner SE. Selection of primers for optimal taxonomic classification of environmental 16S rRNA gene sequences. ISME J 2012." } ] }, { "id": "15389", "date": "02 Aug 2016", "name": "Fiona Fouhy", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a novel and important piece of research. Extensive research is being conducted using next generation sequencing but researchers are becoming increasingly aware that many factors such as PCR bias, region of the 16S rRNA gene targeted etc. can impact on the results achieved. This has a negative impact on the ability to compare results across studies. This manuscript sets about to address this with their new analysis pipeline NG-Tax.\nThe title of the manuscript is good.\nThe abstract accurately summarises the research but the results section should have less methods and more results.\nFigure 1 is vague and fails to show the unique aspects of how NG-Tax differs from e.g. QIIME. More details would make this figure useful.\nI think greater details on the filtering and the classification used by this approach would benefit the reader. Perhaps a table showing the differences between this approach and e.g. RDP , QIIME etc. would improve the readers ability to interpret the novelty of the work.\nThis work was done only using HiSeq data. Do the authors feel that the approach would be equally successful on approaches e.g. Ion, MiSeq etc where longer reads are achieved? It would also be nice to test the approach with a real life data set and not a mock community and see how the results compare to those achieved using traditional analysis approaches.\nFigures 3 and 4 are difficult to interpret, perhaps remake as tables.", "responses": [ { "c_id": "4229", "date": "02 Jan 2019", "name": "Javier Ramiro-Garcia", "role": "Author Response", "response": "This is a novel and important piece of research. Extensive research is being conducted using next generation sequencing but researchers are becoming increasingly aware that many factors such as PCR bias, region of the 16S rRNA gene targeted etc. can impact on the results achieved. This has a negative impact on the ability to compare results across studies. This manuscript sets about to address this with their new analysis pipeline NG-Tax.The title of the manuscript is good.Thank you.The abstract accurately summarizes the research but the results section should have less methods and more results.We tried to find a proper balance between results and methods but taking into account that the paper describes a tool, a description of it should be included because the pipeline is both method and results at the same time. But we tried to include those results needed to prove that NG-Tax is suitable for 16S amplicon analysis:1) Taxonomy assignment using short reads should be comparable with the assignment using the complete 16S rRNA gene.2) Composition profiles based on sequencing data should resemble the real composition of the biological sample.3) α and β diversity should match the expected α and β diversity.4) Results should be reproducible and therefore robust against biological variation (different sample compositions) and technical (PCR and sequencing settings) biases.We consider that these requirements were met by NG-Tax and hope that they will convince readers of the actual improvements that were made, regarding robustness against methodological aspects as well as a more accurate reproduction of the MC compositions. In the new version of the manuscript we included more statistical tests to measure accuracy and reproducibility of NG-Tax.Figure 1 is vague and fails to show the unique aspects of how NG-Tax differs from e.g. QIIME. More details would make this figure useful.As suggested by the reviewer Figure 1 now includes those unique aspects of NG-Tax.I think greater details on the filtering and the classification used by this approach would benefit the reader. Perhaps a table showing the differences between this approach and e.g. RDP , QIIME etc. would improve the readers ability to interpret the novelty of the work.As suggested by the reviewer we detailed the filtering and the classification process in the user manual for those readers that want to dive into the technical details.“This script demultiplexes the raw data into samples using the information contained in the mapping file. It also generates an OTU table per sample after removing chimeras and assigns taxonomy to the OTUs. NG-Tax is designed for short reads, 70 nucleotides is the recommended read length. Reads can be trimmed to this length by the script. Longer length can be selected by the user but comparison with 70 nucleotide analysis is advisable.OTU picking: For each sample reads are ranked by abundance and OTUs are added to an OTU table starting from the most abundant sequence until the read abundance is lower than a percentage defined by the user (recommendeded is at is 0.1%). Subsequently, the discarded reads are clustered to the OTU table allowing one mismatch.Chimera removal: OTUs are subjected to non-reference based chimera checking according to the following principle: given three OTUs named A, B and C, C will be considered a chimera when the following conditions are satisfied: C and A 5’ reads are identical, C and B 3’ reads are identical and both OTUs, A and B, are at least twice as abundant as OTU C.Taxonomic assignment: For each OTU, usearch is used to retrieve hits for the forward and reverse reads against their respective trimmed reference database.Hits that are in common between both reads are divided in 6 identity thresholds 100, 98, 97, 95, 92, 90.A hit belongs to a certain level, for example 97, when both reads have at least a 97 percentage identity with that hit.Using the highest available identity threshold, NG-Tax assigns the consensus taxonomy to the OTU if this taxonomy is supported for at least half of the hits.Genus, Family or Order remains unassigned if the maximum identity percentage level is lower or equal to 97%, 95% and 92% respectively.”Now we also included the main characteristics of NG-Tax in Figure 1.The OTU generation is the main difference between NG-Tax and QIIME. NG-Tax uses a de novo generation approach without clustering. This increases the resolution and allows for the distinction between OTUs with one nucleotide distance. In addition, NG-Tax generates OTUs independently for each sample, which avoids problems associated to sample size heterogeneity. Another important feature of NG-Tax is the use of non-fixed thresholds for the taxonomic assignment, which results in more accurate classifications. To highlight those points we added the text “No clustering → max resolution” to Figure 1 and indicated in the workflow that OTUs are generated per sample. We also clearly state that NG-Tax does not use any clustering and explain that the taxonomic classification uses different identity levels.The authors of the RDP classifier stated that it does not perform well for short sequences, i.e. a length of 200 nt would give accurate family level classification but shorter reads will not, likely due to insufficient features 1. In contrast, NG-Tax has been specifically designed for short reads.This work was done only using HiSeq data. Do the authors feel that the approach would be equally successful on approaches e.g. Ion, MiSeq etc where longer reads are achieved? It would also be nice to test the approach with a real life data set and not a mock community and see how the results compare to those achieved using traditional analysis approaches.NG-Tax has been the reference method for 16S rRNA gene amplicon analysis in our lab for more than two years now, and has been used in more than 30 manuscripts that have been submitted or are in preparation. One of these manuscripts 1 was published before this manuscript. Since then another fifteen studies using NG-Tax have been published 2-16. These studies contain biological samples that belong to very different and specific environments and were sequenced both on MiSeq and HiSeq instruments. These will contribute to the assessment of NG-Tax’s performance, however these were not included in the current manuscript since the data is accessible in the aforementioned publications.Figures 3 and 4 are difficult to interpret, perhaps remake as tables. With Figure 3 and 4 we intended to visualize how close the observed sample composition resembled the expected composition at a glance and how many different taxa are found in the data. In order to further improve interpretation, we have now also added Table 1 to provide detailed information as to the number of misclassified reads at different taxonomic levels and Figure 5, which shows boxplots of the distances to the expected composition. We also performed statistical tests to quantitatively compare the performance of NG-Tax and QIIME. As suggested by the reviewer we included the tables with the taxonomical profiles as supplementary data, which can be used for evaluation of the results." } ] } ]	1	https://f1000research.com/articles/5-1791
https://f1000research.com/articles/7-1299/v1	15 Aug 18	{ "type": "Review", "title": "The evolution and diversity of the nonsense-mediated mRNA decay pathway", "authors": [ "James P. B. Lloyd" ], "abstract": "Nonsense-mediated mRNA decay is a eukaryotic pathway that degrades transcripts with premature termination codons (PTCs). In most eukaryotes, thousands of transcripts are degraded by NMD, including many important regulators of development and stress response pathways. Transcripts can be targeted to NMD by the presence of an upstream ORF or by introduction of a PTC through alternative splicing. Many factors involved in the recognition of PTCs and the destruction of NMD targets have been characterized. While some are highly conserved, others have been repeatedly lost in eukaryotic lineages. Here, I outline the factors involved in NMD, our current understanding of their interactions and how they have evolved. I outline a classification system to describe NMD pathways based on the presence/absence of key NMD factors. These types of NMD pathways exist in multiple different lineages, indicating the plasticity of the NMD pathway through recurrent losses of NMD factors during eukaryotic evolution. By classifying the NMD pathways in this way, gaps in our understanding are revealed, even within well studied organisms. Finally, I discuss the likely driving force behind the origins of the NMD pathway before the appearance of the last eukaryotic common ancestor: transposable element expansion and the consequential origin of introns.", "keywords": [ "RNA", "NMD", "evolution", "UPF1", "SMG1", "transposable element", "RNA decay" ], "content": "What is nonsense-mediated mRNA decay?\n\nGene expression is controlled by a variety of mechanisms, sometimes in unexpected ways. Early mutant screens identified mutations that introduced nonsense mutations, but surprisingly, these premature termination codons (PTCs) lead to a reduction in mRNA stability1,2. This increase in RNA decay is the result of an active translation-dependent process1,3. This pathway was termed nonsense-mediated mRNA decay (NMD) and is now known to regulate hundreds to thousands of transcripts in plants, animals and fungi4–8. Many of the NMD targeted transcripts are not the result of nonsense mutations, but are instead the result of alternative splicing events that introduce PTCs or the presence of an upstream open reading frame (uORF). Many such splicing events are not the result of splicing errors, but are in fact highly conserved events9,10. Therefore, NMD has a major role in shaping the transcriptome of diverse eukaryotes. However, the exact molecular nature of the NMD pathway varies between organisms. Most eukaryotes share the core NMD factors (see below), but an impressive number of modifications to the NMD pathway exist. In this review, I will examine the factors known to act in NMD, discuss the diversity of these factors in eukaryotes, and explore the different mechanisms that explain how a PTC is differentiated from an authentic stop codon. Finally, I will discuss how the NMD pathway may have evolved and some remaining key questions in our understanding of the NMD pathway.\n\n\nThe factors that read nonsense\n\nEarly mutant screens in baker's yeast and Caenorhabditis elegans identified three conserved factors that could suppress a nonsense mutation11,12. These factors were named UP-frameshift (UPF) 1, 2 and 3 in baker's yeast and suppressors with morphological defects on genitalia (SMG) 2, 3 and 4 in C. elegans. The baker's yeast names of these factors are used throughout this review. UPF1 is a highly conserved RNA helicase13 that interacts with UPF2, which is an MIF4G domain-containing protein14, that in turn binds to UPF3 (Figure 1)15,16. The initial mutant screens in C. elegans also revealed four additional factors: the kinase SMG1 and the 14-3-3-like domain proteins SMG5, SMG6 and SMG711,17. In animals, SMG1 is known to phosphorylate UPF1 after a PTC is been recognised (Figure 1)18–20. Initially, NMD factors were defined by their role in the phosphorylation of UPF1. UPF2 and UPF3 support the phosphorylation of UPF1 by creating a complex compatible for phosphorylation by SMG120, while also acting to activate the RNA helicase activity of UPF122. SMG5-7 bind to phosphorylated UPF123 and are active in the dephosphorylation of UPF1 by recruiting the PP2A phosphatase24–26. However, it is now clear that their primary role is in acting at various stages of RNA decay. SMG5-7 have a central role in recruiting the degradation machinery to degrade the NMD target (Figure 1). SMG5 and SMG7 act to recruit exonucleases27, while SMG6 is an endonuclease, cutting the transcript near the PTC28,29. Over time, many more NMD factors have been identified through further genetic and biochemical screens30–33. Of these, SMG8 and SMG9 are of particular interest. First identified in human cells as SMG1-interacting proteins, they act in the NMD pathway of humans and possibly C. elegans32,34 through the inhibition of the kinase SMG1. Curiously, studies in mammals have revealed that many NMD targets do not require the involvement of all NMD factors. Many NMD targets use “branches” of the NMD pathway that do not require UPF235 or UPF3b36. However, all branches do involve UPF1, highlightings its central importance to the NMD pathway.\n\nAt termination events, UPF1 and SMG1 are recruited to termination events by eRF1 and eRF3, leading to the formation of the SMG1-Upf1-eRF1-eRF3 (SURF) complex20. If an EJC, bound by UPF3 and UPF2, is present downstream of the terminating ribosome, then the decay-inducing (DECID) complex will form20,21. This will lead to the phosphorylation of UPF1 by SMG1. Then the ribosome will disassociate and SMG5-7 will be recruited to transcript through phoso-UPF1 binding. The transcript is degraded by nucleases.\n\nTogether these studies, mostly using animal systems, paint a picture where multiple factors (UPF2, UPF3, SMG1, SMG8, and SMG9) assist in the activation of UPF1, while other factors (SMG5-7) act to degrade an NMD target and dephosphorylate UPF1.\n\n\nVariations on a common pathway\n\nDespite the deduction of a basic schematic of the NMD pathway in animals (Figure 1), many of the factors involved in this classical model of NMD vary between different organisms (Figure 2 and Figure 3). The most highly divergent NMD pathways are those found in the excavata (Figure 2 and Figure 3). The excavata have been suggested to be the most basal group of eukaryotes37, although other work places them within the same supergroup as plants38,39. Although the NMD pathways of the parasites Giardia lamblia and Trypanosoma brucei have been studied, it is unclear if a functional NMD pathway exists in these organisms40,41. They contain heavily reduced compliments of NMD factors: the genome of G. lamblia only harbors UPF1, and the genome of T. brucei only harbors UPF1 and UPF240,41. Over-expression of UPF1 in G. lamblia caused an NMD reporter to further decrease, suggesting that G. lamblia might have an active NMD pathway40. In contrast, the knockdown of UPF1 in T. brucei did not increase NMD reporter construct expression, or endogenous genes41. However, tethering of UPF1 in T. brucei did decrease reporter expression41. Therefore, it is difficult to definitively conclude the status of the NMD pathway in excavata. However, it is worth noting that parasites are known to have reduced genomes relative to free-living relatives42, and that the excavata Naegleria gruberi does harbor the additional NMD factors of SMG1 and SMG943. This indicates that a complex NMD pathway involving the kinase SMG1 likely existed in the last eukaryotic common ancestor.\n\nThe distribution of the key NMD factors, UPF1, SMG1 and a member of the SMG5-7 family define the NMD pathway type. All NMD types have arisen multiple times within eukaryotic evolution. To date, no SMG5-7 family member has yet been identified in Naegleria gruberi but given the presence of SMG143, I am currently classifying it as a type 1 NMD pathway. The branch lengths do not reflect the relatedness of any species, but represent the order of separation between the lineages. The root of eukaryotes is unclear, so branches representing a Excavata early and late divergence are represent in grey, dashed-lines. *SMG1 appears to have been lost in other fungal lineages as well, representing repeated losses in multiple fungal lineages43. NMD pathways can be classified into four types (Figure 3).\n\n(A) Classical NMD, exemplified by humans (modified from Figure 1). (B) Recent SMG1-independent NMD, exemplified by A. thaliana. A. thaliana lost SMG1 within the last 5–10 million years43,44. A. thaliana requires SMG7 for a functional NMD pathway45, retains a S/TQ rich UPF143 and its UPF1 needs to be phosphorylated to function in NMD in tobacco leaves46,47. This suggests an alternative kinase may have replaced SMG1. (C) Ancient SMG1-independent NMD, exemplified by baker’s yeast. The NMD pathway of baker’s yeast was the first to be characterised. UPF1, UPF2 and UPF3 have central roles in this pathway. Reverse genetics revealed a potential lesser role for EBS1, a SMG7 homologue, in NMD48 but its UPF1 is depleted in S/TQ dipeptides43. (D) Heavily derived NMD, exemplified by T. brucei. It is unclear if a functional NMD pathway exists in these organisms. In T, it has been shown that UPF1 and UPF2 interact, but their interaction with the ribosome and potential NMD targets is unclear41. Tethering of UPF1 a transcript can decrease its abundance41.\n\nFurther support for this comes from the examination of plant NMD pathways. Plants, which diverged from animals and fungi early in eukaryotic evolution (Figure 2), do have functional homologues of the NMD holy trinity: UPF1-349,50. Plants also have homologues of SMG5-7, known as SMG7 and SMG7-like45, and SMG1 homologues43,44. SMG1 has been repeatedly lost throughout eukaryotic evolution, including two losses in land plants (Arabidopsis thaliana and Capsella rubella) and multiple losses in fungi (Figure 2)43,44. The repeated loss of SMG1 raises some interesting questions about the mechanism of NMD activation. In animals, and presumably in most plants, SMG1 phosphorylates SQ and TQ dipeptides at the N- and C-termini of UPF118,46,51. Species, such as baker’s yeast, with an ancient loss of SMG1 (Figure 2), have UPF1 sequences depleted of S/TQ dipeptides relative to species with SMG143. Species that lost SMG1 more recently, such as A. thaliana, have UPF1 proteins that are rich in S/TQ dipeptides43. The repeated losses of SMG1 in eukaryotes suggests that there is a genetic buffer, another factor/mechanisms that allows SMG1 to be lost but the NMD pathway to be activated43,44. In support of this notion, the experimental perturbation of SMG1 in fruit flies and zebrafish has little or no effect on the NMD pathway of these organisms52–54, suggesting that a backup UPF1-activation mechanism is already present in these species. Alternatively, different mechanisms have replaced SMG1 in each independent loss of SMG1, which might explain why some organisms have retained S/TQ richness within their UPF1 protein sequences while others have not43,44.\n\nThe SMG5-7 family split and diversified in the animal lineage, with the acquisition of the PIN domain in SMG5 and SMG626,55,56. The PIN domain of SMG6 gives it the ability to act as an endonuclease, cutting the NMD targeted transcript near the PTC28,29. The SMG5-7 family also have a role in regulating telomere length57. SMG5-7 homologues in plants are known as SMG7, given they lack the PIN domain of SMG5 and SMG645. SMG5-7 family members of baker’s yeast, EBS1 and ETS1, also lack the PIN domain48. In baker’s yeast, ETS1 is implicated in telomere regulation but not NMD, while a knockout of EBS1 reveals a mild NMD phenotype48. Given that baker’s yeast lacks SMG119,43,44, it is not clear why EBS1/SMG7 would be required for NMD. The UPF1 of baker’s yeast is depleted of S/TQ dipeptides43, which once phosphorylated by SMG1, normally act as binding site for SMG5-723. The lack of S/TQ dipeptides suggest that classical phosphorylation of UPF1 is not required for the activation of NMD in baker’s yeast. Tyrosine phosphorylation of UPF1 in baker’s yeast has been observed and appears to regulate the RNA helicase activity of UPF158. It is possible that these or other phosphorylated sites could act to recruit decay factors like S/TQ dipeptides do. However, given the differences between S and T residues from Y, it seems unlikely that EBS1/SMG7 would be involved. Recently, another member of the SMG5-7 family was characterized in the ciliate Tetrahymena thermophila59, despite the loss of the SMG1 kinase from T. thermophila43,59. The SMG5-7 family member of T. thermophila was named SMG6-like (SMG6L) due to the presence of the C-terminal NYN nuclease domain, potentially taking on the same role as the PIN domain of animal SMG6 proteins59. SMG6L appears to work with UPF1 in the NMD pathway of T. thermophila and is conserved in many other protozoa59. However, it is unclear how it is recruited to UPF1 given the lack of SMG1 and classical phosphorylation sites on UPF1.\n\nThe kinase activity of SMG1 is regulated in part by SMG8 and SMG932. These factors have been identified but not characterized outside of the animal kingdom43; a curious finding which indicates they may have a role in NMD in diverse eukaryotes. When SMG1 is lost from a genome, SMG8 and SMG9 are generally also lost43. Further work will be needed to reveal the extent of any conserved role in NMD for these factors.\n\nTaken together, a diverse set of NMD pathways with varying levels of classically defined NMD factors been identified. Generally speaking, these can be split into four major types (Figure 2 and Figure 3):\n\n1) Classical SMG1-dependent NMD (As exemplified by C. elegans, humans, and moss)\n\n2) Recent SMG1-independent NMD (As exemplified by A. thaliana)\n\n3) Ancient SMG1-independent NMD (As exemplified by baker’s yeast and T. thermophila)\n\n4) Heavily derived NMD (As exemplified by G. lamblia, T. brucei and Cyanidioschyzon merolae)\n\nType 1 NMD pathways (classical SMG1-dependent NMD; Figure 3A) are known to exist in both animals and plants18,19,44 and is likely to be the ancestral state of NMD43,44. However, even here, the dependence on SMG1 is not always clear: SMG1 mutants in fruit flies have much milder phenotypes than mutations in other NMD factors53,60 and knockdown of SMG1 in zebrafish revealed no phenotype54. It is possible that the NMD pathways of some species with a type 1 NMD pathway in appearance might better resemble type 2 NMD (Recent SMG1-independent NMD).\n\nType 2 NMD pathways (recent SMG1-independent NMD; Figure 3B), such as those of the land plants A. thaliana and C. rubella, appear very much like those of type 1, with the exception of SMG1 being absent from the genome, likely with the accompanying loss of SMG8 and SMG943,44. However, UPF1 still maintains the relatively high level of S/TQ dipeptide phosphorylation sites43, and phospho-UPF1 binding protein SMG78,45. It would be tempting to speculate that a kinase related to SMG1 replaced it in the NMD pathway44. ATM and ATR are two kinases from the same family as SMG1 that are conserved in plants and are involved in DNA repair. However, in A. thaliana, the reported mutant phenotypes of ATM and ATR61 do not overlap with the classical NMD phenotypes49, so this seems unlikely to be the case.\n\nA type 3 NMD pathway (ancient SMG1-independent NMD; Figure 3C), was the first to be characterized by a mutant screen in baker’s yeast12,62. These ancient losses of SMG1 lead to an NMD pathway without SMG1, without SMG8 and SMG943, with UPF1 depleted in S/TQ dipeptides43, and an unclear role for SMG5-7 proteins48,59. Future work (see below) will be needed to better understand the exact molecular role of SMG5-7 proteins in type 3 NMD pathways, and to understand how the NMD pathway functions without the SMG1 activating UPF1.\n\nType 4 NMD pathways (heavily derived NMD; Figure 3D) are the most variable group and are found throughout the eukaryotic tree of life. These pathways often lack SMG1, but also core NMD factors (UPF2 and UPF3). Although UPF3 is hard to identify with homology searches63, it does appear to be missing from the genomes of a number of species43. These include the excavata parasites G. lamblia and T. brucei40,41 but also the red algae C. merolae44. C. merolae has a very reduced genome, with only 27 introns in total64. C. merolae and G. lamblia also lack homologues of UPF2. It is certainly possible that the presence of these factors do not represent a fully functional form of an NMD pathway and instead reflect the molecular reminance of a former NMD pathway whose factors have not been co-opted for other functions.\n\nIn any of these species, additional NMD factors are likely to have arisen. PNRC2 is a vertebrate-specific NMD factor33. The only non-type 1 species to have had a forward genetics screen performed for is the baker’s yeast, so we have limited unbiased studies to draw from. More forward screens and biochemical studies are likely to yield more species-specific factors. This will be especially exciting in type 4 species, with the most heavily reduced NMD pathways. This framework of NMD types based on presence/absence of conserved NMD factor is aimed at aiding the comparison and discussion of NMD pathways from diverse organisms. Thinking of all NMD pathways as being fundamentally the same at the molecular level is wrong. There is certainly an overlap, but more focused studies are needed to understand when homologous NMD factors do have the same molecular role in NMD and do not.\n\n\nDefining NMD targets\n\nSo far I have discussed the molecular processes that link the recognition of a PTC to transcript destruction. However, a lot of work has also been focused on understanding the mechanism of how a PTC is differentiated from an authentic stop codon. Multiple models for how this is achieved have been proposed. One of the most well characterized models centres around the exon junction complex (EJC), a protein complex deposited on an mRNA after two exons are ligated together during splicing65,66. While most EJCs are removed from the transcript by the translating ribosome67, EJCs associated with exon-exon junctions ≥50 nt downstream of a stop codon are not removed and can elicit NMD68,69. Early work showed that the EJC was not involved in the NMD pathways of fruit flies56, but more recent work proved the contrary, revealing a role for the EJC in fruit fly NMD70. The EJC has been lost from baker’s yeast and so cannot have a role in its NMD pathway, but the EJC is involved in the fungi Neurospora crassa’s NMD pathway71. The EJC mode has even found support in plants, with reporter genes and transcriptome-wide studies supporting a role for exon-exon junctions in 3’ UTRs eliciting NMD50,72–74. These findings would suggest that the EJC mode is an ancient mechanism for targeting transcripts to NMD. A surprising version of the EJC mode is the finding that NMD in T. thermophila is dependent on splice junctions downstream of the stop codon, but not on the EJC itself59. Knockout of the core EJC component Mago nashi did not alter the expression levels of NMD targets identified by knockout of UPF1 and SMG6L59. This indicates that an alternative mechanism might maintain an EJC-like mode of NMD in T. thermophila.\n\nAnother well-explored system used in defining PTCs is the long 3’ UTR mode. Transcripts with abnormally long 3’ UTRs have been found in reporter genes50,72,75 and transcriptome-wide studies5,73,76 to target transcripts to NMD, although some recent transcriptome-wide studies found little to no trend across the transcriptome, when the presence of 3’ UTR introns were taken into account59,74,77,78. One proposed mechanism is the increased distance between the stop codon (PTC) and the polyA-binding protein, bound to the polyA tail79,80. This physical separation between the polyA tail and the terminating ribosome might lead to aberrant termination and the recruitment of NMD factors79,80. An alternative model posits that longer 3’ UTRs are able to recruit more UPF1 directly bound to the 3’ UTR81. It has been found that UPF1 coats transcripts, but translation displaces UPF1 from all regions, except the 3’ UTRs82. This model suggests that a higher level of UPF1 binding increases the chances of NMD being triggered during the termination of translation; Naturally long 3’ UTRs that are resistant to NMD have been observed to bind less UPF1 than susceptible long 3’ UTR transcripts81. In fact, some naturally long 3’ UTR transcripts appear to be protected from NMD by various features such as a recently identified cis-sequence element in the TRAM1 gene83 or the many genes found to bind PTBP1 near the stop codon to prevent NMD84. Such features protecting long 3’ UTR transcripts from NMD might explain why transcriptome-wide studies find so few long 3’ UTR transcripts that are targeted to NMD.\n\nIn baker’s yeast, a downstream sequence element (DSE) was identified85,86. When this sequence is downstream of a stop codon, NMD is elicited, likely through the recruitment of an RNA binding protein85,86. This mechanism is very similar to the way in which the EJC mode works in animals and plants. The existence of DSEs in other species are possible, but to date, none have been identified. Fission yeast also display an unusual PTC recognition mechanism, not yet observed in other species: splicing near a stop codon, either upstream or downstream of the stop codon, can induce NMD and is independent of the EJC87. Whether such a mechanism exists in other eukaryotes and what factors determine this remain to be seen.\n\nThe mechanisms used to define PTCs in the last eukaryotic common ancestor are unclear. While the EJC mode has been identified in both plants and animals, suggesting an ancient origin, there are many eukaryotic lineages where it has not been characterized, or does not function59,87. The long 3’ UTR mode of NMD has also been characterized in many diverse eukaryotes (plants, animals and fungi), but the mechanism underlying this mode, and failure to observe a strong signal for this feature in transcriptome-wide studies, does raise questions.\n\n\nThe origins of NMD\n\nToday, eukaryotes appear to utilize NMD in a variety of ways to achieve the same aim, degrading PTC-containing transcripts from a variety of sources. It is clear that a rather complex NMD pathway, belonging to the type 1 group, existed in the last eukaryotic common ancestor (see above). In extant diploid eukaryotes, NMD can prevent some mutations from being dominant, protecting heterozygous individuals by turning these alleles recessive11,88,89. However, NMD also increases the severity of some genetic disorders90, creating a double-edged sword: protecting some mutation-carrying individuals while exacerbating the conditions of others. Therefore, it is unlikely that protecting the genome from nonsense mutations was the driving force behind the origin of the NMD pathway. Early eukaryotes did face a particular selective pressure not present in prokaryotes: rapidly multiplying transposable elements (TEs). The origin of sex in eukaryotes allowed for TEs to expand in copy number, which is not possible in prokaryotes with their primarily asexual reproductive system91,92. With the advent of sex, eukaryotes faced the expansion of many TE classes, including the self-splicing (group II) introns. Expansion of group II introns has been proposed to have driven the evolution of the spliceosome to enhance the splicing of these selfish elements93, the nucleus evolved to physically separate the processes of transcription and translation and allow for intron removal before translation94, and NMD evolved to degrade intron-retaining transcripts that escaped the nucleus94. These adaptations ensure that retention of these efficiently-spliced introns would not be repeatedly translation. More recent expansions of introns in some eukaryotic lineages are due to the expansion of DNA transposons95, indicating the importance of these mechanisms to protect the genome from TE expansions in extant eukaryotes, and suggests multiple origins for introns from TEs throughout eukaryotic evolution. Once functional as a TE-intron protection pathway, NMD appears to have been co-opted to control gene expression. Today, in addition to repressing the expression of uORF-containing genes, pseudogenes and the products of alternative splicing, NMD may allow for the evolution of new introns. The presence of NMD may act as a buffer for novel introns with weak splice sites96. In fact, the red algae C. merolae only has 27 introns64 and is missing all of the classical NMD factors with the exception of a UPF1 homologue44. It is possible that C. merolae lacks a functional NMD pathway and this limits the acquisition of new introns, at least partly explaining its intron depleted genome.\n\n\nUnanswered questions\n\nMany years of study have revealed diverse NMD pathways, centering on UPF1. However, there are a number of fundamental questions remaining in the field regarding the mechanisms and evolution of NMD.\n\n1) Why is SMG1 repeatedly lost in different lineages? Is there a backup mechanism to activate UPF1 and is this conserved between the lineages that have recently lost (eg A. thaliana) and more anciently (eg baker’s yeast and T. thermophila) SMG1? Or are there multiple SMG1 replacement mechanisms?\n\n2) What recruits the RNA degradation machinery to UPF1 when SMG1 is lost and S/TQ dipeptides are depleted? Does this still depend on the SMG5-7 family and UPF1 phosphorylation?\n\n3) What precisely are the roles of SMG5-7 family members in lineages that have lost SMG1? Do their roles differ between species with type 2 (recent loss of SMG1) and type 3 (ancient loss of SMG1) NMD pathways?\n\n4) What is the molecular basis of an EJC mode of PTC recognition when the EJC is not involved, such as in T. thermophila?\n\n5) What is the precise mechanistic roles of UPF2/UPF3 in relation to EJC mode and non-EJC mode NMD pathways?\n\n6) To understand the discrepancies between transcriptome-wide and reporter construct approaches to the long 3’ UTR mode of NMD and to uncover the molecular mechanism(s) behind the long 3’ UTR mode.\n\nHopefully future research efforts can resolve these and other unknowns surrounding NMD.\n\n\nConclusion\n\nHere I have discussed the NMD pathway in the context of evolution and the many shapes the NMD pathways takes. I have proposed a classification system with four types of NMD pathway, based on the presence/absence of conserved NMD factors. I propose that the classical (type 1) NMD involves UPF1-3, the UPF1-kinase SMG1 and the SMG5-7 family. The recent (type 2) and ancient (type 3) loss of SMG1 define the next two types of NMD, while loss of all but UPF1 and perhaps UPF2 define the final type (type 4), where NMD might not actually function at all. It is highly likely that species specific NMD factors have been co-opted in many, if not all, of these types of NMD pathway and are waiting to be discovered. Discussing the evolution and mechanism of NMD within this framework will hopefully aid in the communication of ideas between different model systems used to study NMD and therefore help in knowledge acquisition. Finally, I outline key outstanding questions regarding the mechanism and evolution of the NMD pathway. Focused research efforts to address these issues will certainly help in our overall understanding of the NMD pathway and for us to at last appreciate the true fundamental nature of NMD.\n\n\nData availability\n\nNo data are associated with this article.", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by the Australian Research Council (ARC) Centre of Excellence program in Plant Energy Biology CE140100008.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nThanks to Barry Causier, Suruchi Roychoudhry and Joanna Franklin-Lloyd for critical feedback on this review. Thanks to the Centre of Excellence in Plant Energy Biology, Australian Research Council (CE140100008) for funding.\n\n\nReferences\n\nLosson R, Lacroute F: Interference of nonsense mutations with eukaryotic messenger RNA stability. Proc Natl Acad Sci U S A. 1979; 76(10): 5134–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMaquat LE, Kinniburgh AJ, Rachmilewitz EA, et al.: Unstable beta-globin mRNA in mRNA-deficient beta o thalassemia. Cell. 1981; 27(3 Pt 2): 543–53. PubMed Abstract \| Publisher Full Text\n\nGozalbo D, Hohmann S: Nonsense suppressors partially revert the decrease of the mRNA level of a nonsense mutant allele in yeast. Curr Genet. 1990; 17(1): 77–9. PubMed Abstract \| Publisher Full Text\n\nHe F, Li X, Spatrick P, et al.: Genome-wide analysis of mRNAs regulated by the nonsense-mediated and 5' to 3' mRNA decay pathways in yeast. Mol Cell. 2003; 12(6): 1439–52. PubMed Abstract \| Publisher Full Text\n\nMendell JT, Sharifi NA, Meyers JL, et al.: Nonsense surveillance regulates expression of diverse classes of mammalian transcripts and mutes genomic noise. Nat Genet. 2004; 36(10): 1073–8. PubMed Abstract \| Publisher Full Text\n\nRehwinkel J, Letunic I, Raes J, et al.: Nonsense-mediated mRNA decay factors act in concert to regulate common mRNA targets. RNA. 2005; 11(10): 1530–44. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKurihara Y, Matsui A, Hanada K, et al.: Genome-wide suppression of aberrant mRNA-like noncoding RNAs by NMD in Arabidopsis. Proc Natl Acad Sci U S A. 2009; 106(7): 2453–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRayson S, Arciga-Reyes L, Wootton L, et al.: A role for nonsense-mediated mRNA decay in plants: pathogen responses are induced in Arabidopsis thaliana NMD mutants. PLoS One. 2012; 7(2): e31917. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLareau LF, Inada M, Green RE, et al.: Unproductive splicing of SR genes associated with highly conserved and ultraconserved DNA elements. Nature. 2007; 446(7138): 926–9. PubMed Abstract \| Publisher Full Text\n\nLareau LF, Brenner SE: Regulation of splicing factors by alternative splicing and NMD is conserved between kingdoms yet evolutionarily flexible. Mol Biol Evol. 2015; 32(4): 1072–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPulak R, Anderson P: mRNA surveillance by the Caenorhabditis elegans smg genes. Genes Dev. 1993; 7(10): 1885–97. PubMed Abstract \| Publisher Full Text\n\nLeeds P, Wood JM, Lee BS, et al.: Gene products that promote mRNA turnover in Saccharomyces cerevisiae. Mol Cell Biol. 1992; 12(5): 2165–77. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCulbertson MR, Leeds PF: Looking at mRNA decay pathways through the window of molecular evolution. Curr Opin Genet Dev. 2003; 13(2): 207–14. PubMed Abstract \| Publisher Full Text\n\nPonting CP: Novel eIF4G domain homologues linking mRNA translation with nonsense-mediated mRNA decay. Trends Biochem Sci. 2000; 25(9): 423–6. PubMed Abstract \| Publisher Full Text\n\nSchell T, Köcher T, Wilm M, et al.: Complexes between the nonsense-mediated mRNA decay pathway factor human upf1 (up-frameshift protein 1) and essential nonsense-mediated mRNA decay factors in HeLa cells. Biochem J. 2003; 373(Pt 3): 775–83. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKadlec J, Izaurralde E, Cusack S: The structural basis for the interaction between nonsense-mediated mRNA decay factors UPF2 and UPF3. Nat Struct Mol Biol. 2004; 11(4): 330–7. PubMed Abstract \| Publisher Full Text\n\nCali BM, Kuchma SL, Latham J, et al.: smg-7 is required for mRNA surveillance in Caenorhabditis elegans. Genetics. 1999; 151(2): 605–16. PubMed Abstract \| Free Full Text\n\nYamashita A, Ohnishi T, Kashima I, et al.: Human SMG-1, a novel phosphatidylinositol 3-kinase-related protein kinase, associates with components of the mRNA surveillance complex and is involved in the regulation of nonsense-mediated mRNA decay. Genes Dev. 2001; 15(17): 2215–28. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGrimson A, O’Connor S, Newman CL, et al.: SMG-1 is a phosphatidylinositol kinase-related protein kinase required for nonsense-mediated mRNA Decay in Caenorhabditis elegans. Mol Cell Biol. 2004; 24(17): 7483–90. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKashima I, Yamashita A, Izumi N, et al.: Binding of a novel SMG-1-Upf1-eRF1-eRF3 complex (SURF) to the exon junction complex triggers Upf1 phosphorylation and nonsense-mediated mRNA decay. Genes Dev. 2006; 20(3): 355–67. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHug N, Cáceres JF: The RNA helicase DHX34 activates NMD by promoting a transition from the surveillance to the decay-inducing complex. Cell Rep. 2014; 8(6): 1845–56. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChamieh H, Ballut L, Bonneau F, et al.: NMD factors UPF2 and UPF3 bridge UPF1 to the exon junction complex and stimulate its RNA helicase activity. Nat Struct Mol Biol. 2008; 15(1): 85–93. PubMed Abstract \| Publisher Full Text\n\nFukuhara N, Ebert J, Unterholzner L, et al.: SMG7 is a 14-3-3-like adaptor in the nonsense-mediated mRNA decay pathway. Mol Cell. 2005; 17(4): 537–47. PubMed Abstract \| Publisher Full Text\n\nChiu SY, Serin G, Ohara O, et al.: Characterization of human Smg5/7a: a protein with similarities to Caenorhabditis elegans SMG5 and SMG7 that functions in the dephosphorylation of Upf1. RNA. 2003; 9(1): 77–87. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAnders KR, Grimson A, Anderson P: SMG-5, required for C.elegans nonsense-mediated mRNA decay, associates with SMG-2 and protein phosphatase 2A. EMBO J. 2003; 22(3): 641–50. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOhnishi T, Yamashita A, Kashima I, et al.: Phosphorylation of hUPF1 induces formation of mRNA surveillance complexes containing hSMG-5 and hSMG-7. Mol Cell. 2003; 12(5): 1187–200. PubMed Abstract \| Publisher Full Text\n\nUnterholzner L, Izaurralde E: SMG7 acts as a molecular link between mRNA surveillance and mRNA decay. Mol Cell. 2004; 16(4): 587–96. PubMed Abstract \| Publisher Full Text\n\nHuntzinger E, Kashima I, Fauser M, et al.: SMG6 is the catalytic endonuclease that cleaves mRNAs containing nonsense codons in metazoan. RNA. 2008; 14(12): 2609–17. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEberle AB, Lykke-Andersen S, Mühlemann O, et al.: SMG6 promotes endonucleolytic cleavage of nonsense mRNA in human cells. Nat Struct Mol Biol. 2009; 16(1): 49–55. PubMed Abstract \| Publisher Full Text\n\nLongman D, Plasterk RH, Johnstone IL, et al.: Mechanistic insights and identification of two novel factors in the C. elegans NMD pathway. Genes Dev. 2007; 21(9): 1075–85. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLongman D, Hug N, Keith M, et al.: DHX34 and NBAS form part of an autoregulatory NMD circuit that regulates endogenous RNA targets in human cells, zebrafish and Caenorhabditis elegans. Nucleic Acids Res. 2013; 41(17): 8319–31. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYamashita A, Izumi N, Kashima I, et al.: SMG-8 and SMG-9, two novel subunits of the SMG-1 complex, regulate remodeling of the mRNA surveillance complex during nonsense-mediated mRNA decay. Genes Dev. 2009; 23(9): 1091–105. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCho H, Kim KM, Kim YK: Human proline-rich nuclear receptor coregulatory protein 2 mediates an interaction between mRNA surveillance machinery and decapping complex. Mol Cell. 2009; 33(1): 75–86. PubMed Abstract \| Publisher Full Text\n\nRosains J, Mango SE: Genetic characterization of smg-8 mutants reveals no role in C. elegans nonsense mediated decay. PLoS One. 2012; 7(11): e49490. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGehring NH, Kunz JB, Neu-Yilik G, et al.: Exon-junction complex components specify distinct routes of nonsense-mediated mRNA decay with differential cofactor requirements. Mol Cell. 2005; 20(1): 65–75. PubMed Abstract \| Publisher Full Text\n\nChan WK, Huang L, Gudikote JP, et al.: An alternative branch of the nonsense-mediated decay pathway. EMBO J. 2007; 26(7): 1820–30. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHe D, Fiz-Palacios O, Fu CJ, et al.: An alternative root for the eukaryote tree of life. Curr Biol. 2014; 24(4): 465–70. PubMed Abstract \| Publisher Full Text\n\nRichards TA, Cavalier-Smith T: Myosin domain evolution and the primary divergence of eukaryotes. Nature. 2005; 436(7054): 1113–8. PubMed Abstract \| Publisher Full Text\n\nDerelle R, Torruella G, Klimeš V, et al.: Bacterial proteins pinpoint a single eukaryotic root. Proc Natl Acad Sci U S A. 2015; 112(7): E693–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChen YH, Su LH, Sun CH: Incomplete nonsense-mediated mRNA decay in Giardia lamblia. Int J Parasitol. 2008; 38(11): 1305–17. PubMed Abstract \| Publisher Full Text\n\nDelhi P, Queiroz R, Inchaustegui D, et al.: Is there a classical nonsense-mediated decay pathway in trypanosomes? PLoS One. 2011; 6(9): e25112. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPoulin R, Randhawa HS: Evolution of parasitism along convergent lines: from ecology to genomics. Parasitology. 2015; 142 Suppl 1: S6–15. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCausier B, Li Z, De Smet R, et al.: Conservation of Nonsense-Mediated mRNA Decay Complex Components Throughout Eukaryotic Evolution. Sci Rep. 2017; 7(1): 16692. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLloyd JP, Davies B: SMG1 is an ancient nonsense-mediated mRNA decay effector. Plant J. 2013; 76(5): 800–10. PubMed Abstract \| Publisher Full Text\n\nRiehs N, Akimcheva S, Puizina J, et al.: Arabidopsis SMG7 protein is required for exit from meiosis. J Cell Sci. 2008; 121(Pt 13): 2208–16. PubMed Abstract \| Publisher Full Text\n\nKerényi F, Wawer I, Sikorski PJ, et al.: Phosphorylation of the N- and C-terminal UPF1 domains plays a critical role in plant nonsense-mediated mRNA decay. Plant J. 2013; 76(5): 836–48. PubMed Abstract \| Publisher Full Text\n\nMérai Z, Benkovics AH, Nyikó T, et al.: The late steps of plant nonsense-mediated mRNA decay. Plant J. 2013; 73(1): 50–62. PubMed Abstract \| Publisher Full Text\n\nLuke B, Azzalin CM, Hug N, et al.: Saccharomyces cerevisiae Ebs1p is a putative ortholog of human Smg7 and promotes nonsense-mediated mRNA decay. Nucleic Acids Res. 2007; 35(22): 7688–97. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nArciga-Reyes L, Wootton L, Kieffer M, et al.: UPF1 is required for nonsense-mediated mRNA decay (NMD) and RNAi in Arabidopsis. Plant J. 2006; 47(3): 480–9. PubMed Abstract \| Publisher Full Text\n\nKerényi Z, Mérai Z, Hiripi L, et al.: Inter-kingdom conservation of mechanism of nonsense-mediated mRNA decay. EMBO J. 2008; 27(11): 1585–95. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOkada-Katsuhata Y, Yamashita A, Kutsuzawa K, et al.: N- and C-terminal Upf1 phosphorylations create binding platforms for SMG-6 and SMG-5:SMG-7 during NMD. Nucleic Acids Res. 2012; 40(3): 1251–66. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChen Z, Smith KR, Batterham P, et al.: Smg1 nonsense mutations do not abolish nonsense-mediated mRNA decay in Drosophila melanogaster. Genetics. 2005; 171(1): 403–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMetzstein MM, Krasnow MA: Functions of the nonsense-mediated mRNA decay pathway in Drosophila development. PLoS Genet. 2006; 2(12): e180. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWittkopp N, Huntzinger E, Weiler C, et al.: Nonsense-mediated mRNA decay effectors are essential for zebrafish embryonic development and survival. Mol Cell Biol. 2009; 29(13): 3517–28. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nClissold PM, Ponting CP: PIN domains in nonsense-mediated mRNA decay and RNAi. Curr Biol. 2000; 10(24): R888–90. PubMed Abstract \| Publisher Full Text\n\nGatfield D, Unterholzner L, Ciccarelli FD, et al.: Nonsense-mediated mRNA decay in Drosophila: at the intersection of the yeast and mammalian pathways. EMBO J. 2003; 22(15): 3960–70. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAzzalin CM, Reichenbach P, Khoriauli L, et al.: Telomeric repeat containing RNA and RNA surveillance factors at mammalian chromosome ends. Science. 2007; 318(5851): 798–801. PubMed Abstract \| Publisher Full Text\n\nLasalde C, Rivera AV, León AJ, et al.: Identification and functional analysis of novel phosphorylation sites in the RNA surveillance protein Upf1. Nucleic Acids Res. 2014; 42(3): 1916–29. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTian M, Yang W, Zhang J, et al.: Nonsense-mediated mRNA decay in Tetrahymena is EJC independent and requires a protozoa-specific nuclease. Nucleic Acids Res. 2017; 45(11): 6848–63. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFrizzell KA, Rynearson SG, Metzstein MM: Drosophila mutants show NMD pathway activity is reduced, but not eliminated, in the absence of Smg6. RNA. 2012; 18(8): 1475–86. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCulligan K, Tissier A, Britt A: ATR regulates a G2-phase cell-cycle checkpoint in Arabidopsis thaliana. Plant Cell. 2004; 16(5): 1091–104. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLeeds P, Peltz SW, Jacobson A, et al.: The product of the yeast UPF1 gene is required for rapid turnover of mRNAs containing a premature translational termination codon. Genes Dev. 1991; 5(12A): 2303–14. PubMed Abstract \| Publisher Full Text\n\nAronoff R, Baran R, Hodgkin J: Molecular identification of smg-4, required for mRNA surveillance in C. elegans. Gene. 2001; 268(1–2): 153–64. PubMed Abstract \| Publisher Full Text\n\nMatsuzaki M, Misumi O, Shin-I T, et al.: Genome sequence of the ultrasmall unicellular red alga Cyanidioschyzon merolae 10D. Nature. 2004; 428(6983): 653–7. PubMed Abstract \| Publisher Full Text\n\nLe Hir H, Moore MJ, Maquat LE: Pre-mRNA splicing alters mRNP composition: evidence for stable association of proteins at exon-exon junctions. Genes Dev. 2000; 14(9): 1098–108. PubMed Abstract \| Free Full Text\n\nLe Hir H, Izaurralde E, Maquat LE, et al.: The spliceosome deposits multiple proteins 20-24 nucleotides upstream of mRNA exon-exon junctions. EMBO J. 2000; 19(24): 6860–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGehring NH, Lamprinaki S, Kulozik AE, et al.: Disassembly of exon junction complexes by PYM. Cell. 2009; 137(3): 536–48. PubMed Abstract \| Publisher Full Text\n\nGehring NH, Neu-Yilik G, Schell T, et al.: Y14 and hUpf3b form an NMD-activating complex. Mol Cell. 2003; 11(4): 939–49. PubMed Abstract \| Publisher Full Text\n\nFerraiuolo MA, Lee CS, Ler LW, et al.: A nuclear translation-like factor eIF4AIII is recruited to the mRNA during splicing and functions in nonsense-mediated decay. Proc Natl Acad Sci U S A. 2004; 101(12): 4118–23. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSaulière J, Haque N, Harms S, et al.: The exon junction complex differentially marks spliced junctions. Nat Struct Mol Biol. 2010; 17(10): 1269–71. PubMed Abstract \| Publisher Full Text\n\nZhang Y, Sachs MS: Control of mRNA Stability in Fungi by NMD, EJC and CBC Factors Through 3’UTR Introns. Genetics. 2015; 200(4): 1133–48. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKertész S, Kerényi Z, Mérai Z, et al.: Both introns and long 3’-UTRs operate as cis-acting elements to trigger nonsense-mediated decay in plants. Nucleic Acids Res. 2006; 34(21): 6147–57. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDrechsel G, Kahles A, Kesarwani AK, et al.: Nonsense-mediated decay of alternative precursor mRNA splicing variants is a major determinant of the Arabidopsis steady state transcriptome. Plant Cell. 2013; 25(10): 3726–42. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLloyd JPB, Lang D, Zimmer AD, et al.: The loss of SMG1 causes defects in quality control pathways in Physcomitrella patens. Nucleic Acids Res. 2018; 46(11): 5822–5836, [cited 2018 Mar 28]. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYepiskoposyan H, Aeschimann F, Nilsson D, et al.: Autoregulation of the nonsense-mediated mRNA decay pathway in human cells. RNA. 2011; 17(12): 2108–18. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHurt JA, Robertson AD, Burge CB: Global analyses of UPF1 binding and function reveal expanded scope of nonsense-mediated mRNA decay. Genome Res. 2013; 23(10): 1636–50. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLindeboom RG, Supek F, Lehner B: The rules and impact of nonsense-mediated mRNA decay in human cancers. Nat Genet. 2016; 48(10): 1112–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nColombo M, Karousis ED, Bourquin J, et al.: Transcriptome-wide identification of NMD-targeted human mRNAs reveals extensive redundancy between SMG6- and SMG7-mediated degradation pathways. RNA. 2017; 23(2): 189–201. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAmrani N, Ganesan R, Kervestin S, et al.: A faux 3’ -UTR promotes aberrant termination and triggers nonsense-mediated mRNA decay. Nature. 2004; 432(7013): 112–8. PubMed Abstract \| Publisher Full Text\n\nBehm-Ansmant I, Gatfield D, Rehwinkel J, et al.: A conserved role for cytoplasmic poly(A)-binding protein 1 (PABPC1) in nonsense-mediated mRNA decay. EMBO J. 2007; 26(6): 1591–601. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHogg JR, Goff SP: Upf1 senses 3'UTR length to potentiate mRNA decay. Cell. 2010; 143(3): 379–89. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZünd D, Gruber AR, Zavolan M, et al.: Translation-dependent displacement of UPF1 from coding sequences causes its enrichment in 3′ UTRs. Nat Struct Mol Biol. 2013; 20(8): 936–43. PubMed Abstract \| Publisher Full Text\n\nToma KG, Rebbapragada I, Durand S, et al.: Identification of elements in human long 3’ UTRs that inhibit nonsense-mediated decay. RNA. 2015; 21(5): 887–97. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGe Z, Quek BL, Beemon KL, et al.: Polypyrimidine tract binding protein 1 protects mRNAs from recognition by the nonsense-mediated mRNA decay pathway. eLife. 2016; 5: pii: e11155. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPeltz SW, Brown AH, Jacobson A: mRNA destabilization triggered by premature translational termination depends on at least three cis-acting sequence elements and one trans-acting factor. Genes Dev. 1993; 7(9): 1737–54. PubMed Abstract \| Publisher Full Text\n\nGonzález CI, Ruiz-Echevarría MJ, Vasudevan S, et al.: The yeast hnRNP-like protein Hrp1/Nab4 marks a transcript for nonsense-mediated mRNA decay. Mol Cell. 2000; 5(3): 489–99. PubMed Abstract \| Publisher Full Text\n\nWen J, Brogna S: Splicing-dependent NMD does not require the EJC in Schizosaccharomyces pombe. EMBO J. 2010; 29(9): 1537–51. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCali BM, Anderson P: mRNA surveillance mitigates genetic dominance in Caenorhabditis elegans. Mol Gen Genet. 1998; 260(2–3): 176–84. PubMed Abstract \| Publisher Full Text\n\nKhajavi M, Inoue K, Lupski JR: Nonsense-mediated mRNA decay modulates clinical outcome of genetic disease. Eur J Hum Genet. 2006; 14(10): 1074–81. PubMed Abstract \| Publisher Full Text\n\nBhuvanagiri M, Schlitter AM, Hentze MW, et al.: NMD: RNA biology meets human genetic medicine. Biochem J. 2010; 430(3): 365–77. PubMed Abstract \| Publisher Full Text\n\nHickey DA: Selfish DNA: a sexually-transmitted nuclear parasite. Genetics. 1982; 101(3–4): 519–31. PubMed Abstract \| Free Full Text\n\nZeyl C, Bell G, Green DM: Sex and the spread of retrotransposon Ty3 in experimental populations of Saccharomyces cerevisiae. Genetics. 1996; 143(4): 1567–77. PubMed Abstract \| Free Full Text\n\nCavalier-Smith T: Intron phylogeny: a new hypothesis. Trends Genet. 1991; 7(5): 145–8. PubMed Abstract \| Publisher Full Text\n\nMartin W, Koonin EV: Introns and the origin of nucleus-cytosol compartmentalization. Nature. 2006; 440(7080): 41–5. PubMed Abstract \| Publisher Full Text\n\nHuff JT, Zilberman D, Roy SW: Mechanism for DNA transposons to generate introns on genomic scales. Nature. 2016; 538(7626): 533–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFarlow A, Meduri E, Dolezal M, et al.: Nonsense-mediated decay enables intron gain in Drosophila. PLoS Genet. 2010; 6(1): e1000819. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "37408", "date": "30 Aug 2018", "name": "Wei Miao", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this manuscript, the author discussed the diversity of the NMD pathway according to the variation of NMD factors in different eukaryotes, and briefly summarized some popular NMD models. Furthermore, the author also discussed the relationship between intron gain/loss and NMD. Several interesting and unsolved questions in this field were also mentioned by the end of the manuscript. Interestingly, the author came up with a novel classification system (although it needs to be further modified, see below) to classify NMD mechanisms into different types based on the new criterion. Namely, it classifies NMD mechanisms according to the presence or absence of key factors that related to the well-characterized Upf1 C-terminus phosphorylation events. The manuscript is clearly written, and this work is likely to be of general interest to the NMD field. My main criticism is about the novel classification system.\nMajor\nAs it is still unclear whether excavates have functional NMD, the type 4 NMD may not need to be taken into account. Or question marks should be included in the figure 2 and figure 3D, and in the main text.\n\nType 1 NMD seems to be SMG1-dependent and EJC-dependent (see figure 3A). Hence it cannot be exemplified by C. elegans (EJC-independent NMD). The author could modify the figure a little. For example, one could draw EJC with a dashed line to indicate it is dispensable for NMD in some organisms, like C. elegans. Besides, the presence of Smg1 and Upf1 orthologs in N. gruberi does not necessarily mean it possesses a functional NMD pathway, thus I suggest not consider it has type 1 NMD. Or, as I mentioned above, excavates need not be discussed too much.\n\nAs mentioned by the author, some organisms (e.g., D. melanogaster and D. rerio) are considered to have SMG1-dependent type 1 NMD, yet their SMG1 proteins are dispensable for NMD activation and therefore similar to the type 2 NMD. I am wondering whether this issue can be solved by defining another type of NMD, namely in between type 1 and type 2. Besides, an additional figure (e.g., figure 3 in (Lareau and Brenner, 20151) could be provided to show the evolutionary relationship between different types of NMD.\n\nFigure 2 needs to be modified. Clearly, solely based on the pattern of these icons (NMD proteins), readers may get confused why organisms with the same pattern are not classified into the same group. For example, like C. rubella, Tetrahymena also has a red square and a blue triangle. However, they are classified into different types. Although, in this case, it can be easily solved by adding another icon to indicate Upf1 with phosphorylatable S/TQ motifs. Additional modifications are required to let readers understand, for example, why Dikarya and Mirosporia are considered as type 1, but not type 2.\n\nMinor\nPage 2, left panel, lane 10. Ciliates should be mentioned here as well, because NMD was also proved to be required for regulating many of their transcripts (Jaillon et al. 20082, Tian et al. 20173).\n\nPage 2, right panel, lanes 19 – 20. I suggest deleting this sentence. Or, I would rather say that, even in animals, NMD factors were simply defined by their requirement for NMD. Because phosphorylation of Upf1 seems to be not essential for eliciting NMD in some animals (e.g., SMG1-independent NMD in fruit flies and zebrafish).\n\nPage 2, figure 1, the bottom panel. I suggest moving Upf2, Upf1 and its associated proteins (especially the endonuclease Smg6) to the left side of the EJC, close to the endonucleolytic cleavage site.\n\nPage 2, figure 1, the title of figure legend. The word “Animals” is probably too general, hence the author may want to replace it with another word (e.g., vertebrates). Because it is known that EJC is not required or dispensable for NMD in some invertebrates (Longman et al. 20074, Gatfield et al. 20035).\n\nPage 2, figure 1, the last sentence of the figure legend. Replace “phoso-UPF1” with “phosphor-UPF1” or “phos-UPF1”.\n\nPage 3, right panel, lane 8. Smg7 recruits CCR4–NOT deadenylase complex instead of exonuclease (Loh et al. 20136).\n\nPage 3, left panel, lane 14. Reference 34 (Rosains and Mango, 20127) suggests that Smg8 is not required for NMD in C. elegans.\n\nPage 3, left panel, lane 28. Change the word “animals”, see comments #3.\n\nPage 3, left panel, the first paragraph of the section “Variations on a common pathway”. Since it is still unclear whether NMD exists in excavates, it is inappropriate to say that they have “the most divergent NMD pathway”. Some following sentences in this paragraph are also against the existence of NMD pathway in excavates. To support the hypothesis “a complex NMD pathway involving… in the last eukaryotic common ancestor”, the author could start the discussion from plant NMD mechanism, and then use the existence of orthologs of NMD core factors in excavates as a supporting evidence.\n\nPage 4, figure 3 legend. “In T, it has been shown” should be “In T. brucei …”.\n\nPage 5, left panel, lanes 2 – 3 and lanes 20 – 21. The author should mention that the interaction between Upf1 and Smg6 can also be phosphorylation-independent (Chakrabarti et al. 20148).\n\nPage 5 and many other places. It is better to use a unified way to indicate different organisms (groups). For example: change “C. elegans, humans, and moss” to “C. elegans, H. sapiens and P. patens” or “worms, humans and moss”.\n\nPage 5, right panel, lane 5. Is there experimental evidence to support that S/TQ sites have undergone phosphorylation? If not, I suggest replacing “S/TQ dipeptide phosphorylation sites” with “phosphorylatable S/TQ motifs”.\n\nPage 5, right panel, “PNRC2 is a vertebrate-specific NMD factor”. The author may want to remove this sentence because a recent study has shown that PNRC2 may not be required for NMD (Nicholson et al. 20189).\n\nPage 6, left panel, lanes 19 – 21. The author should modify this sentence a little. Despite the splice junctions downstream of the stop codon (DSJ) are enriched in potential Tetrahymena NMD targets (Tian et al. 20173), it is still insufficient to conclude that “NMD in T. thermophila is dependent on DSJ”.\n\nPage 6, left panel, bottom. The description of DSE model could be removed because evidence from a few studies are against this model (for example (Meaux et al. 200810).\n\nPage 6, right panel, lane 12. The fungus N. crassa should also be mentioned here, due to the requirement of EJC for its NMD pathway.\n\nPage 6, right panel, section “The origins of NMD”, the second sentence. The word “clear” is too strong here.\n\nPage 6, right panel, section “The origins of NMD”. Firstly, the author may want to change the subtitle of this section. In this section, a clear answer to origins of NMD is not given, and discussions are mainly about the relationship between NMD and the intron evolution. Secondly, when discussing the relationship between intron evolution and (the evolution of) NMD, an earlier study from Michael Lynch and Avinash Kewalramani is deserved to be mentioned here (Lynch et al. 200311).\n\nPage 7, left panel, section “unanswered questions”. Regarding the question 2, the author should consider the existence of phosphorylation-independent interaction between Upf1 and Smg6. The question 4 should be modified as the existence of “an EJC mode of PTC recognition” in Tetrahymena has not been proven yet.\n\nIs the topic of the review discussed comprehensively in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Partly\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? Partly", "responses": [ { "c_id": "4219", "date": "12 Nov 2018", "name": "James Lloyd", "role": "Author Response", "response": "Thank you for reading my review and giving so many insights into how I can improve my manuscript. I have now submitted a version 2 of this article that updated the text and figures with many of the insights you and the other referees offered. Below are some highlights of things that I address in the updated manuscript:As it is still unclear whether excavates have functional NMD, the type 4 NMD may not need to be taken into accountCertainly and I have tried to make this clear in the submitted version 2 of this review. Type 1 NMD seems to be SMG1-dependent and EJC-dependent (see figure 3A). Hence it cannot be exemplified by C. elegans (EJC-independent NMD).I did not mean to give that impression so I have removed the EJC from Figure 3 of the submitted version 2 of this review. Also, while no reports of EJC involvement in C. elegans are published, I do not think we have enough data to state whether NMD in C. elegans is really EJC-independent and I look forward to future work that might add to this. Figure 2 needs to be modified. Clearly, solely based on the pattern of these icons (NMD proteins), readers may get confused why organisms with the same pattern are not classified into the same group. For example, like C. rubella, Tetrahymena also has a red square and a blue triangle. However, they are classified into different types. Although, in this case, it can be easily solved by adding another icon to indicate Upf1 with phosphorylatable S/TQ motifs. Additional modifications are required to let readers understand, for example, why Dikarya and Mirosporia are considered as type 1, but not type 2. Thank you for raising some issues with Figure 2. I have now corrected some of my mistakes of misclassification in this figure. I was not satisfied with any of my attempts to differentiate between UPF1 proteins of Type 2 and 3 so that I have left unchanged but I was a good point that you raise. “PNRC2 is a vertebrate-specific NMD factor”. The author may want to remove this sentence because a recent study has shown that PNRC2 may not be required for NMDThank you for raising this point about PNRC2, as did other referees and I have removed mention of it from my review for the sake of simplicity. The description of DSE model could be removed because evidence from a few studies are against this modelThis is a great point and I have now removed the section of DSE from the submitted version 2 of this review." } ] }, { "id": "37786", "date": "21 Sep 2018", "name": "Damien Garcia", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this review, the author suggests a novel NMD classification system based on the evolutionary conservation of known NMD factors among different species. The author also discusses a possible driving force for the apparition of NMD. This is a well-written and structured manuscript considering NMD under an interesting evolutionary point of view. It is of broad interest for the NMD community, and it mentions several fundamental questions currently unanswered in the NMD field. Nevertheless, a few modifications of the figures and the addition of some very important references and concepts could make it easier to read and broaden the general interest of the review, as detailed in the following paragraphs.\n\nMajor comments:\n\nIn Figure 1, the model of mammalian NMD indicates cleavage of the NMD target downstream of the PTC. It doesn’t indicate the distinct roles of SMG6 and SMG5/7; it would be important here to indicate the distinct involvement of SMG6 in the cleavage activity and the role of SMG5/7 in decay factors recruitment. These two routes for decay should clearly appear in Figure 1.\n\nIn Figure 2, the definition of Type 1/2/3/4 NMD should already be mentioned in the figure legends (e.g. Type 1: classical SMG1 dependent NMD, Type 2: recent loss of SMG1 with conserved phosphorylation, Type 3: ancient loss of SMG1 with loss of UPF1 S/Q phosphorylation, Type 4: Heavily derived NMD).\n\nIn the paragraph 'Defining NMD targets', the author cites the 2016 paper on the NMD protection effect of PTB1; an earlier study describing similar protection effect of Pub1 in yeast should also be cited (Ruiz-Echevarria and Peltz, 20001).\n\nIn the same paragraph, the author focuses on Baker’s yeast DSE, stimulating NMD. The author should also mention a recent paper in yeast describing that poor translation efficiency is a major criteria for NMD targeting (Celik et al., 20172). This major result could potentially explain not yet understood deregulations observed in several other species upon NMD knockdown.\n\nIt was proposed that invading RNAs of external origin, including TEs and viruses, could be a driving force for NMD apparition and evolution - this should be mentioned in the paragraph on the possible origin of NMD (Hamid and Makeyev, 20163).\n\nMinor comments:\n\nUPF1 might have other essential functions beyond NMD, as observed for UPF3b in mammals, involved in translation termination (Neu-Yilik et al., 20174), which could explain its presence in some species without any other known NMD factors. This could be mentioned in the corresponding section.\n\nAs branches of NMD exist without the need of UFP2/UPF3, it suggests that NMD could be active with only UPF1. This possibility should be discussed/mentioned when describing Type 4 species depleted of UPF2 and UPF3.\n\nIn addition to ATR/ATM, the author could mention TOR or TRRAP kinases as described in (Lloyd and Davies, 20135), as possible kinase replacements for SMG1.\n\nIn Figure 3, the author should add precisions on endonucleolytic cleavage and decay factor recruitment for (A), and decay factor recruitment only for the others. The different types of NMD defined in Figure 2 should be mentioned again here in Figure 3, namely: Type 1, Type 2, Type 3 and Type 4.\n\nIs the topic of the review discussed comprehensively in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? Partly", "responses": [ { "c_id": "4220", "date": "12 Nov 2018", "name": "James Lloyd", "role": "Author Response", "response": "I am thankful for you carefully reading my review and giving great suggestions for its improvement. I have now submitted a version 2 of this article that updated the text and figures with many of the insights you and the other referees offered. Below are some highlights of things that I address in the updated manuscript:involvement of SMG6 in the cleavage activity and the role of SMG5/7 in decay factors recruitment. These two routes for decay should clearly appear in Figure 1.This is a great point and I have tried to address this in the submitted version 2 of the review. In Figure 2, the definition of Type 1/2/3/4 NMD should already be mentioned in the figure legendsI have now added this in the submitted version 2 of this review. In the paragraph 'Defining NMD targets', the author cites the 2016 paper on the NMD protection effect of PTB1; an earlier study describing similar protection effect of Pub1 in yeast should also be citedThank you for bringing this to my attention and it is now included in the submitted version 2 of this review. It was proposed that invading RNAs of external origin, including TEs and viruses, could be a driving force for NMD apparition and evolution - this should be mentioned in the paragraph on the possible origin of NMD (Hamid and Makeyev, 20163).I have now cited this work in the submitted version 2 of this review. UPF1 might have other essential functions beyond NMD, as observed for UPF3b in mammals, involved in translation termination (Neu-Yilik et al., 20174), which could explain its presence in some species without any other known NMD factors. This could be mentioned in the corresponding section.Great point and I have now mentioned this in the submitted version 2 of this review. As branches of NMD exist without the need of UFP2/UPF3, it suggests that NMD could be active with only UPF1. This possibility should be discussed/mentioned when describing Type 4 species depleted of UPF2 and UPF3.Agreed and this is now mentioned in the submitted version 2 of this review. In addition to ATR/ATM, the author could mention TOR or TRRAP kinases as described in (Lloyd and Davies, 20135), as possible kinase replacements for SMG1.Good point, I have now added TOR to the submitted version 2 of this review. I did not add TRRAP because TRRAP has been reported to be a kinase dead member of the family (https://www.sciencedirect.com/science/article/pii/S0092867400814798?via%3Dihub)." } ] }, { "id": "38016", "date": "24 Sep 2018", "name": "Niels H. Gehring", "expertise": [ "Reviewer Expertise mRNA turnover (nonsense-mediated mRNA decay)" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this review article James P. B. Lloyd summarizes the evolution of NMD in various eukaryotic organisms. He describes the process of NMD in general as well as the detection of NMD substrates by the NMD machinery and discusses the functions of several NMD factors. As a novelty, he introduces a new classification of NMD pathways based on the presence of NMD factors in different organisms.\nIn my view, this is a well-written review that takes a refreshing new look at evolutionary aspects of NMD. However, one should keep in mind that the absence of a certain NMD pathway does not necessarily mean that it really does not exist in a particular organism. A good example for this is NMD in Drosophila, which was thought to be EJC-independent based on initial publications. However, there is now evidence that certain NMD substrates in Drosophila are degraded in an EJC-dependent manner. It is therefore possible that the classification of NMD pathways will change with future publications.\nAdditional comments:\nPage 4, last paragraph on the right side: The author probably does not mean ETS1, but EST1 (Reichenbach et al., 20031). Furthermore, a second SMG6 homologue in yeast may be NMD4, which has been originally identified in a yeast-2-hybrid screen (He and Jacobson, 19952) and was recently described by Dehecq et al. (Dehecq et al., 20183) to be associated with UPF1.\nPage 4, last paragraph on the right side and page 5, first paragraph on the left side: Wang et al. (Wang et al., 20064) and de Pinto et al. (de Pinto et al., 20045) have reported that UPF1 and UPF2 are phosphoproteins in yeast. However, it is not clear which kinase is responsible for the phosphorylation of UPF1 and UPF2 and whether the phosphorylation plays a functional role during NMD in yeast.\nPage 5, penultimate paragraph on the right side: It is strongly debated whether PNRC2 is an NMD factor (Nicholson et al., 20186). Therefore, PNRC2 is not a good example for a vertebrate-specific NMD factor.\nPage 6, last paragraph on the right side: the meaning of the sentence “These adaptations ensure that retention of these efficiently-spliced introns would not be repeatedly translation” is not clear to me.\nPage 7, question 2: It is known that SMG6 can interact with UPF1 also in a phosphorylation-independent manner (Nicholson et al., 20147 and Chakrabarti et al., 20148). This could explain how NMD substrates are degraded in the absence of SMG1 or when S/TQ dipeptides are depleted.\n\nIs the topic of the review discussed comprehensively in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? Partly", "responses": [ { "c_id": "4221", "date": "12 Nov 2018", "name": "James Lloyd", "role": "Author Response", "response": "I am truly grateful for your careful reading and thoughtful feedback of my review. I have now submitted a version 2 of this article that updated the text and figures with many of the insights you and the other referees offered. Below are some highlights of things that I address in the updated manuscript:one should keep in mind that the absence of a certain NMD pathway does not necessarily mean that it really does not exist in a particular organism. A good example for this is NMD in Drosophila, which was thought to be EJC-independent based on initial publications. However, there is now evidence that certain NMD substrates in Drosophila are degraded in an EJC-dependent manner. It is therefore possible that the classification of NMD pathways will change with future publications.I completely agree with this point and I hope that the readers will take this message away with them after reading my review. The author probably does not mean ETS1, but EST1Yes, thank you for catching this. Furthermore, a second SMG6 homologue in yeast may be NMD4, which has been originally identified in a yeast-2-hybrid screen (He and Jacobson, 19952) and was recently described by Dehecq et al. (Dehecq et al., 20183) to be associated with UPF1.Thank you for bringing Dehecq et al. to my attention! I think that this article represents very important work and I have now included much discussion of it and its implications in the submitted version 2 of my article. penultimate paragraph on the right side: It is strongly debated whether PNRC2 is an NMD factorThank you for raising this point about PNRC2, as did other referees and I have removed mention of it from my review for the sake of simplicity. It is known that SMG6 can interact with UPF1 also in a phosphorylation-independent manner (Nicholson et al., 20147 and Chakrabarti et al., 20148). This could explain how NMD substrates are degraded in the absence of SMG1 or when S/TQ dipeptides are depleted.Yes, I think that this is a really great point and I discuss this at length in the submitted version 2 of the review. Thank you for raising this point." } ] }, { "id": "37784", "date": "25 Sep 2018", "name": "J. Robert Hogg", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe nonsense-mediated mRNA decay pathway is found throughout eukaryotes, where it performs important quality control and regulatory functions. How the pathway originally emerged and has subsequently adapted to diverse eukaryotic transcriptomes is an important but poorly understood question. This manuscript represents a helpful summary of the evidence for compositionally- and mechanistically-distinct NMD pathways in diverse eukaryotes and raises interesting questions that deserve future study.\nMajor points:\nThe review raises the question of how RNA decay machinery is recruited to NMD substrate mRNAs in organisms that have lost SMG1. Work from yeast and human cells give two possible answers to this question. First, the Jacobson and Parker labs have presented evidence that budding yeast UPF1 directly interacts with components of the decapping complex, enabling SMG1- and SMG6-independent decay 1,2. Second, the Conti and Muhlemann groups have 3,4 shown that SMG6 can interact with UPF1 in a phosphorylation-independent manner. Their data suggest that UPF1 phosphorylation may be dispensable for SMG6 recruitment but may contribute to activation of its endonucleolytic activity, raising the possibility that a SMG-1 independent pathway could rely on phosphorylation-independent SMG6-UPF1 interactions coupled with a distinct mechanism for activation of SMG6.\n\nFigure 3: The schematics in the figure suggest that there is a universal step of mRNP remodeling that involves ribosome displacement from target mRNAs prior to initiation of decay, but it is not clear that this is the case. Previous work from the Baker lab has indicated that budding yeast NMD can initiate on polysome-bound mRNAs, rather than those stripped of ribosomes 5. In addition, this figure and Figure 1 should acknowledge that there is evidence that “classical” NMD can also proceed through deadenylation, decapping, and exonucleolytic decay, not just SMG6-mediated cleavage. For other organisms, “cleavage” implies an endonucleolytic step, which in organisms lacking SMG6 is not known to occur. As referenced above, yeast NMD proceeds through decapping, and this should be made clear in the figure.\n\nIt would be helpful to note that in Drosophila, a much more significant role for SMG1 is uncovered in SMG5 mutants 6. This is consistent with the idea that organisms such as Drosophila have developed redundant pathways for NMD. However, this also means that the extent to which “the dependence on SMG1 is not always clear” may be overstated. The fact that SMG1 has been maintained in this organism and can function in at least some contexts should carry greater weight than the failure to observe a strong phenotype in the limited experimental contexts in which it has been examined.\n\nPage 6, second paragraph: It is somewhat misleading to state that “transcriptome-wide studies find so few long 3’UTR transcripts that are targeted to NMD.” It is true that there have been differing reports of the extent to which 3’UTR length correlates with decay susceptibility transcriptome-wide, but this is not the same as saying that these studies did not find evidence that substantial numbers of long 3’UTR-containing transcripts are subject to NMD. In addition, it is important to recognize that the Lindeboom et al. study cited here examined apparent NMD susceptibility of mRNAs with nonsense mutations, not the scope of long 3’UTR-mediated decay among normal transcripts 7.\n\nPage 6, second paragraph, continued: The poly-A binding protein-centric model and the UPF1 length-sensing model are not necessarily exclusive. It has been reported that Pab1 and poly-A tails are dispensable for accurate NMD target discrimination in yeast 8, but it is possible that UPF1 binding contributes to competition between poly-A binding protein and NMD factors for binding to release factors at the terminating ribosome. Another emerging possibility is that PABP antagonizes UPF1 binding to 3’UTRs, as proposed by Lee et al. 9.\n\nMinor points:\nIt would be helpful to the reader to reference recent evidence that PNRC2 may function in general decapping but contribute minimally to NMD in vertebrates 10.\n\nPage 2, second paragraph: Since UPF2 and UPF3 were identified and initially characterized in yeast, in which UPF1 phosphorylation is not known to play a key role in decay, the basis for the statement that “initially, NMD factors were defined by their role in the phosphorylation of UPF1” is unclear.\n\nPage 2, first paragraph: The Maquat et al., 1981 paper was not based on a “mutant screen” but instead observations in human genetic disease 11.\n\nFigure 1 implies that the roles of UPF2 and UPF3 are dependent on the EJC, but this is not the case — these proteins have been found to be required for decay of NMD substrates lacking 3’UTR introns 12,13.\n\nAt several points, it is not clear whether the authors use the nomenclature 'SMG5-7' to refer to SMG5, 6, and 7, or just SMG5 and 7, as is the more standard usage. For example, on Page 3, first paragraph, it should be made clear that SMG6 does not function in “recruiting the degradation machinery” but is itself an endonuclease.\n\nPage 5, fifth paragraph. It is possible that other PIKK-type kinases other than ATM and ATR are re-purposed to phosphorylate UPF1.\n\nDiscussion of the Tetrahymena data should acknowledge that at this point the evidence for a role for exon junctions is correlative and remains to be mechanistically investigated. This is an important caveat to the classification system offered by the author 14.\n\nPage 6, third paragraph: More recent investigations of yeast NMD have not uncovered evidence for a downstream sequence element that contributes significantly to decay target discrimination 8,15.\n\nIs the topic of the review discussed comprehensively in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Partly\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? Partly", "responses": [ { "c_id": "4222", "date": "12 Nov 2018", "name": "James Lloyd", "role": "Author Response", "response": "Thank you for your thoughtful and insightful reading of my review. You raised a number of important points that I have tried to address in version 2 of my review article. Some of the points I directly address are:Second, the Conti and Muhlemann groups have 3,4 shown that SMG6 can interact with UPF1 in a phosphorylation-independent manner.It would be helpful to note that in Drosophila, a much more significant role for SMG1 is uncovered in SMG5 mutants 6.This is a very important point and I think that this points to a possible mechanism to explain how some species can have an active NMD pathway without SMG1. When we combine this with recent yeast work finding direct interactions between EBS1 and NMD4, and UPF1, this suggests that in many cases, phosphorylation is not needed for NMD. This supports a model that in mammals phosphorylation of UPF1 increases if NMD is stalled, suggesting that it acts to increase the recruitment of degradation factors (https://www.nature.com/articles/ncomms12434), potentially in NMD limiting conditions or for tricky to degrade transcripts. The poly-A binding protein-centric model and the UPF1 length-sensing model are not necessarily exclusive. It has been reported that Pab1 and poly-A tails are dispensable for accurate NMD target discrimination in yeast 8, but it is possible that UPF1 binding contributes to competition between poly-A binding protein and NMD factors for binding to release factors at the terminating ribosome. Another emerging possibility is that PABP antagonizes UPF1 binding to 3’UTRs, as proposed by Lee et al.9.This is a fair point and I did not mean to suggest that the two models were mutually exclusive in the first version. Also I think that Lee et al. is a great reference for me to include, thank you for bringing it to my attention! It would be helpful to the reader to reference recent evidence that PNRC2 may function in general decapping but contribute minimally to NMD in vertebrates 10.Thank you for raising this point about PNRC2, as did other referees and I have removed mention of it from my review for the sake of simplicity. Figure 1 implies that the roles of UPF2 and UPF3 are dependent on the EJC, but this is not the case — these proteins have been found to be required for decay of NMD substrates lacking 3’UTR introns 12,13.Yes this is a very important point and I did not mean to give that impression with this figure. I have removed the EJC from Figure 3 and I have mentioned in the legend of Figure 1 that NMD can happen without an EJC present in the now submitted version 2 of this manuscript. It is possible that other PIKK-type kinases other than ATM and ATR are re-purposed to phosphorylate UPF1.The only other kinase active PIKK in Arabidopsis is TOR, which I have now included in the submitted version 2 of this review. More recent investigations of yeast NMD have not uncovered evidence for a downstream sequence element that contributes significantly to decay target discrimination 8,15.I have now removed the section of DSE from the submitted version 2 of this review." } ] } ]	1	https://f1000research.com/articles/7-1299
https://f1000research.com/articles/7-1042/v1	10 Jul 18	{ "type": "Method Article", "title": "Identifying communities from multiplex biological networks by randomized optimization of modularity", "authors": [ "Gilles Didier", "Alberto Valdeolivas", "Anaïs Baudot", "Alberto Valdeolivas" ], "abstract": "The identification of communities, or modules, is a common operation in the analysis of large biological networks. The Disease Module Identification DREAM challenge established a framework to evaluate clustering approaches in a biomedical context, by testing the association of communities with GWAS-derived common trait and disease genes. We implemented here several extensions of the MolTi software that detects communities by optimizing multiplex (and monoplex) network modularity. In particular, MolTi now runs a randomized version of the Louvain algorithm, can consider edge and layer weights, and performs recursive clustering.\nOn simulated networks, the randomization procedure clearly improves the detection of communities. On the DREAM challenge benchmark, the results strongly depend on the selected GWAS dataset and enrichment p-value threshold. However, the randomization procedure, as well as the consideration of weighted edges and layers generally increases the number of trait and disease community detected.\nThe new version of MolTi and the scripts used for the DMI DREAM challenge are available at: https://github.com/gilles-didier/MolTi-DREAM.", "keywords": [ "Biological Networks", "Multiplex", "Multi-layer", "Community identification", "Clustering", "DREAM challenge" ], "content": "Introduction\n\nBiological macromolecules do not act isolated in cells, but interact with each other to perform their functions, in signaling or metabolic pathways, molecular complexes, or, more generally, biological processes. Thanks to the development of experimental techniques and to the extraction of knowledge accumulated in the literature, biological networks are nowadays assembled on a large-scale. A common feature of biological networks is their modularity, i.e., their organization around communities - or functional modules - of tightly connected genes/proteins implicated in the same biological processes1,2.\n\nThe Disease Module Identification (DMI) DREAM challenge aims at investigating different algorithms dedicated to the identification of communities, in a biomedical context3. The challenge has been divided into two sub-challenges, to identify communities either i) from six biological networks independently, or ii) from all these networks jointly. The clustering approaches proposed by the participants are assessed regarding their capacity to reveal disease communities, defined as communities significantly associated with genes implicated in diseases in GWAS studies3,4. The challengers proposed various strategies and clustering approaches, including kernel clustering, random walks or modularity optimization. We competed with an enhanced version of MolTi, a modularity-based software that we recently developed5. MolTi was initially developed to cluster multiplex networks, i.e., networks composed of different layers of interactions. It extended the modularity measure to multiplex networks and adapted the Louvain algorithm to optimize this multiplex-modularity. We have demonstrated that this multiplex approach better identifies the communities than approaches merging the networks, or performing consensus clusterings, both on simulated and real biological datasets5.\n\nGrounded on these initial results, we here extended and tested our MolTi software, both on simulated data and on the DMI challenge framework. We improved MolTi with the implementation of a randomization procedure, the consideration of edge and layer weights, and a recursive clustering of the classes larger than a given size.\n\nWith simulated data, we observed that considering more than one network layer improves the detection of communities, as already noted5, but also that communities are better detected with the randomization procedure. With the DMI benchmark, we pointed to a great dependence on the GWAS dataset used for the evaluation and on the FDR threshold defined, but, overall, randomizations and edge and layer weights increase the number of detected disease communities.\n\n\nMethods\n\nWe detected communities with an extended version of MolTi5, a modularity-based software. Although MolTi was specifically designed for multiplex networks, it deals with monoplex networks by considering them as multiplexes composed of a single layer. All the networks are here considered undirected. The new version of MolTi, MolTi-DREAM, and the scripts used for the DMI DREAM challenge are available at https://github.com/gilles-didier/MolTi-DREAM.\n\nModularity. Network modularity was initially designed to measure the quality of a partition into communities6, and subsequently used to find such communities. Since finding the partition optimizing the modularity is NP-complete, we applied the meta-heuristic Louvain algorithm7. Louvain starts from the community structure that separates all vertices. Next, it tries to move each vertex from its community to another, picks the move that increases the most modularity, and iterates until no change increases the modularity anymore. It then replaces the vertices by the detected communities and performs the same operations on the newly obtained graph, until the modularity cannot be increased anymore. In order to handle multiplexes, we use a multiplex-adapted modularity and an adaptation of the Louvain algorithm for optimizing this multiplex-modularity.\n\nEdge and layer weights Modularity approaches can deal with weighted networks8, and we modified MolTi to handle weighted networks. We also added the possibility to weight each layer of the multiplex network: the contribution of each layer in Equation (1) is multiplied by its weight when computing the multiplex modularity.\n\nMultiplex modularity The modularity measure to detect communities in a multiplex network (X(g))g can be written as\n\n\n\nwhere X(g) denotes the (monoplex) network of the layer g, w(g) is the user-defined weight associated to the network g, m(g) is the sum of the weights of all the edges of X(g), Xi,j(g) is the weight of the edge {i, j} in X(g), Ki(g) is the sum of the weights of all the edges involving vertex i in X(g), δci,cj is equal to 1 if i and j belong to a same community and to 0 otherwise.\n\nRandomization. We implemented a randomized version of the Louvain algorithm, similar to the one in GenLouvain9. Rather than updating the current partition by picking the move leading to the greatest increase of the modularity, we randomly pick a move among those leading to an increase of the modularity. Different runs of the randomized Louvain generally return different partitions, even if the results are often close. MolTi-DREAM runs the randomized Louvain algorithm a user-defined number of times (from one to ten in this work, four by default), and returns the partition with the highest modularity.\n\nWe simulated random multiplex networks with a fixed known community structure, namely 1,000 vertices split into 20 balanced communities, and various topological properties (i.e., dense/sparse/mixed, with/without missing data)5. Multiplex networks are simulated by drawing each layer according to this structure and fixed intra/inter community edge probabilities (0.1/0.01 for sparse layers and 0.5/0.2 for dense ones). We also generated multiplex networks with missing data in which we randomly withdrawn vertices of each layer with probability 0.5. The relevance of a community structure is assessed by computing the adjusted Rand index10 between the detected communities and the ones used to simulate the multiplex networks.\n\nBiological Networks. The DMI challenge provided six biological networks: two protein-protein interactions, one signaling, one co-expression, one network linking genes essential for the same cancer types, and one network connecting evolutionary-related genes. These six networks have various sizes and edge densities (Table 1). All networks have weighted edges, and all networks but the signaling network are undirected. However, we considered the signaling network as undirected.\n\nEvaluations with GWAS data. The communities identified by the different challengers were evaluated according to the associations of their member genes with GWAS data, following the PASCAL tool4. The procedure leverages the SNP-based p-value statistics obtained from 180 GWAS datasets, covering common diseases and traits. It is to note that an important parameter is the FDR threshold used to define the significant associations3,4, and to get the number of significant disease communities. We used three datasets: the “Leaderboard” (76 GWASs) and “Final” (104 GWASs), which were used during the challenge, and their union in a “Total” dataset (180 GWASs).\n\nObtaining modules in a given size range. The DMI challenge set up two constraints on the submitted communities: no overlap and a size ranging from 3 to 100 nodes. We tested different pre-filters (pruning leaves), parameters (resolution parameter, recursions, combination of graph weights for multiplexes) and post-filters (density, size, pruning leaves) in each leaderboard round. We took into account both the number of significant communities and the total number of submitted communities to evaluate the pertinence of each combination. All partitions were post-filtered to keep only classes containing from 7 to 100 nodes.\n\nResolution parameter Modularity-based clustering approaches are often associated to a resolution parameter γ to tune the size of the obtained communities. We tested different values of this parameters (γ = 1, γ = 5, γ = 10, γ = 110), but the leaderboard tests showed clearly better results for the recursive approach. We chose to keep the default γ = 1 and focused on this recursive procedure.\n\nRecursion procedure We re-clustered all the communities above a certain size (here 100 vertices) by extracting the corresponding subgraphs from the networks and applying recursively the MolTi algorithm. We iterated the process until obtaining only communities with less than 100 vertices, if possible (some communities with more than 100 vertices cannot be split by considering modularity).\n\n\nResults\n\nTo evaluate the accuracy of the community structures detected from the initial MolTi and its improved version that includes the randomization procedure, we simulated random multiplex networks with a fixed, known community structure, and various features5. Considering a greater number of layers always improves the inference of communities, as already observed5 (Figure 1). In addition, communities are better detected from sparse multiplexes than from dense ones. We also observed that the randomizations improve the accuracy of the detected communities, in particular for dense multiplex networks, with or without missing data. Increasing the number of randomization runs improves the results, but to a limited extend after more than four runs.\n\nMultiplex networks contain from 1 to 9 graph layers. The indexes are averaged over 2,000 random multiplex networks of 1,000 vertices and 20 balanced communities. Each layer of sparse (resp. dense) multiplex networks is simulated with 0.1/0.01 (resp. 0.5/0.2) internal/external edge probabilities. Mixed multiplex networks are simulated by uniformly sampling each layer among these two pairs of edge probabilities. Multiplex networks with missing data (right column) are generated by withdrawing vertices from each layer with probability 0.5.\n\nWe applied the improved MolTi to the networks provided by the DMI challenge (Methods). We focused on the sub-challenge 2, which was dedicated to the identification of communities from multiple networks. We considered the six DMI biological networks as layers of a multiplex network, and applied the recursion procedure to obtain communities in the required size range. The significant disease communities were selected regarding their enrichments in GWAS-associated genes (Methods). We observed first that the number of detected disease communities varies in a non-trivial way depending on the GWAS dataset and FDR threshold used (Figure 2). However, we can observe that the number of detected significant disease modules slightly increases after randomization, in particular when the FDR threshold is higher (Figure 2).\n\n“Leaderboard” and “Final” datasets were used during the training and final evaluation of the challenge, respectively, whereas the “Total” dataset is the union of the two previous ones.\n\nMultiplex versus monoplex. We next evaluated the added value of the multiplex approach as compared to the identification of modules from the individual networks. When analyzing the significant disease modules obtained for a FDR threshold of 0.1, we observed that combining biological networks in a multiplex generally increases the number of significant modules (Figure 3). However, this does not stand for the cancer and/or homology networks, which lower the number of significant modules retrieved when added as layers of the multiplex. We hypothesize that the community structures of these networks (if they exist) are so unrelated that it is pointless to seek for a common structure by integrating them.\n\nTen randomization have been applied, and the FDR threshold is set to 0.1.\n\nThese observations are consistent with the DMI challenge observations, in which the top-scoring team in the sub-challenge 2 handled only the two protein-protein interaction networks. Our algorithm also performs well with the two protein-protein interactions networks, but the highest number of disease modules is retrieved by considering network combinations that exclude the cancer and homology network layers (Figure 3).\n\nEvaluation of the edge and layer weighting. All the six biological networks used in the DMI challenge have weighted edges. We compared the number of disease modules obtained by considering or not these weights in the MolTi partitioning, for different FDR thresholds (Table 2). We observed that intra-layer edge weights only has a slight effect on the number of identified significant disease modules, except for the very low significance threshold of 0.01, where it seems pertinent to use these weights.\n\nMolTi-DREAM allows assigning weights to each layer of the multiplex network, for instance to emphasize the layers known to contain more relevant biological information. Given the results of the DMI challenge and our first analyses, we decided to test a combination of weights that would lower the importance of the 5-cancer and 6-homology network layers. We observed that this led to detecting more disease modules (Figure 4). Conversely, less disease modules are detected when higher weights are given to these networks (Figure 4).\n\n\nDiscussion and conclusion\n\nWe applied here the MolTi software and various extensions to identify disease-associated communities following the DMI challenge benchmark. The new version of MolTi, MolTi-DREAM, runs a randomization procedure, takes into account edge and layer weights, and performs a recursive clustering of the classes that are larger than a given size. We finished tied for second in the challenge. However, even if we obtained higher scores than monoplex approaches, the difference was not significant and the organizers of the DREAM challenge declared the sub-challenge 2 vacant.\n\nIn the simulations, all the networks are randomly generated from the same community structure. These networks can thereby be seen as different and partial views of the same underlying community structure. Combining their information in a suitable way is thereby expected to recover the original structure more accurately. In contrast, combining networks with unrelated community structures (or no structure at all) is rather likely to blur the signal carried by each network. The DMI biological networks are constructed from different biological sources that might correspond to unrelated community structures.\n\nThis may explain the results of the sub-challenge 2, in which the top-performer used only the two protein-protein interaction networks, and the fact that the highest number of modules retrieved by our approach was not obtained from a multiplex containing all the six networks. From a biological perspective, the protein-protein networks and the pathway networks are expected to contain mainly physical or signaling interactions between proteins. It has been shown that interacting proteins tend to be co-expressed11, which could explain why the co-expression network also provides complementary information. In contrast, both the cancer and the homology networks are determined from processes operating at a very different level.\n\nEvaluating the relevance of the community structure detected from real-life datasets is a very complicated problem since the actual structure is hidden and generally unknown. In this context, the only possibility for assessing the detected communities is to consider indirect evidence provided by some independent biological information. Different teams are thereby developing proxies to evaluate the communities, mainly based on testing the enrichment of genes contained in each community in Pathways or Gene Ontology annotations. The approach followed by the DMI DREAM challenge is based on GWAS data. This GWAS-based evaluation is specific in the sense that it considers p-value-weighted annotations rather than usual binary ones, i.e., “annotated/not annotated”. This probably contributed to the volatility of the results observed with the DMI DREAM challenge framework.\n\n\nData and software availability\n\nMolTi-DREAM and the scripts used for the DMI DREAM challenge: https://github.com/gilles-didier/MolTi-DREAM\n\nArchived scripts and source code for MolTi-DREAM as at time of publication: http://doi.org/10.5281/zenodo.130120912\n\nLicense for MolTi-DREAM: GNU 3\n\nGD designed MolTi and its extensions, AB and AV applied MolTi during and after the challenge. AV and AB are currently at AixMarseille Univ, Inserm, MMG, France. All authors participated to the design of the study, the interpretation of the results and the writing of the manuscript.", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe project leading to this publication has received funding from the Centre National de la Recherche Scientifique (PEPS BMI IMFMG), the French “Plan Cancer 2009–2013”, and the Excellence Initiative of Aix-Marseille University - AMIDEX, a French “Investissements d’Avenir” programme.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nHartwell LH, Hopfield JJ, Leibler S, et al.: From molecular to modular cell biology. Nature. 1999; 402(6761 Suppl): C47–52. PubMed Abstract \| Publisher Full Text\n\nMitra K, Carvunis AR, Ramesh SK, et al.: Integrative approaches for finding modular structure in biological networks. Nat Rev Genet. 2013; 14(10): 719–32. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChoobdar S, Ahsen ME, Crawford J, et al.: Open community challenge reveals molecular network modules with key roles in diseases. bioRxiv. 2018. Publisher Full Text\n\nLamparter D, Marbach D, Rueedi R, et al.: Fast and Rigorous Computation of Gene and Pathway Scores from SNP-Based Summary Statistics. PLoS Comput Biol. 2016; 12(1): e1004714. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDidier G, Brun C, Baudot A: Identifying Communities from Multiplex Biological Networks. PeerJ. 2015; 3: e1525. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNewman ME, Girvan M: Finding and evaluating community structure in networks. Phys Rev E Stat Nonlin Soft Matter Phys. 2004; 69(2 Pt 2): 026113. PubMed Abstract \| Publisher Full Text\n\nBlondel VD, Guillaume J-L, Lambiotte R, et al.: Fast unfolding of communities in large networks. Journal of Statistical Mechanics: Theory and Experiment. 2008; 2008(10): P10008. Publisher Full Text\n\nNewman ME: Analysis of weighted networks. Phys Rev E Stat Nonlin Soft Matter Phys. 2004; 70(5 Pt 2): 056131. PubMed Abstract \| Publisher Full Text\n\nMucha PJ, Richardson T, Macon K, et al.: Community structure in time-dependent, multiscale, and multiplex networks. Science. 2010; 328(5980): 876–8. PubMed Abstract \| Publisher Full Text\n\nSantos JM, Embrechts M: On the use of the adjusted rand index as a metric for evaluating supervised classification. In C. Alippi, M. Polycarpou, C. Panayiotou, and G. Ellinas, editors, Artificial Neural Networks – ICANN2009, 175–184, Berlin, Heidelberg, Springer Berlin Heidelberg, 2009Publisher Full Text\n\nRual JF, Venkatesan K, Hao T, et al.: Towards a proteome-scale map of the human protein-protein interaction network. Nature. 2005; 437(7062): 1173–1178. PubMed Abstract \| Publisher Full Text\n\nDidier G: gilles-didier/MolTi-DREAM: MolTi-DREAM (Version v1.0). Zenodo. 2008. Data Source" }	[ { "id": "36392", "date": "03 Aug 2018", "name": "Emre Guney", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nDidier and colleagues present MolTi-DREAM, an update to their previous software for community detection in multiplex networks and its application to the disease module discovery using synthetic and DREAM challenge data. The new version of the software adds the possibility of edge and layer weights, the randomization of the underlying Louvain algorithm and partitioning of large modules into smaller modules based on user defined parameters. Based on their analysis of simulated and biological interaction data, they reaffirm that using multiple layers of networks and randomized repetition of module discovery could improve the accuracy. Overall the article is clearly written and the technical details are clearly explained with several exceptions I mention below:\n1. The multiplex modularity formula lacks the resolution parameter (which should appear before ki kj / 2m in the summation). The authors are encouraged to provide data / figures with respect to the reasoning on the selection of the default resolution parameter (currently they mention verbally that it showed better results on leaderboard).\n2. Randomization procedure is unclear. I believe the cost function used is the multiplex modularity given in Eq 1 (L) but the “move” leading to an increase of the modularity is not formally stated (sth like i, j: random_pick(argmax ci, cj L)). The choice of 4 by default is also somewhat arbitrary given that the difference between 3,4,5 are all very small. The significance of the difference in Rand index across runs in Figure 1 could be used to decide the number after which the change is not significant.\n3. The simulation of the multiplex networks could be explained better. Potentially owing to the previous publication, there is no detail in regards to which model was used to generate these networks. A brief description of SBMs and the underlying assumptions of the model it relies on would help. Are these simulated networks are more ER-like? Could this be the reason why the simulated results do not reflect the results on the real networks? Also, it would certainly be useful to add the standard deviation across 2000 networks in Figure 1. I had difficulties following how exactly the mixed and noisy networks are generated (did the mixed ones still have 1000 vertices, 500 from sparse and 500 from dense, for the missing data if the vertices were drawn half of the time, did they have 500 vertices?).\n4. The coverage / precision of disease module discovery is missing. What is the number of all modules identified by the algorithm, how many disease modules are there (or could there be) in total? The latter question could be tricky to answer but I believe a lower and upper bound could be provided (e.g., number of diseases and a number based on the GWAS hits for that disease divided by the min module size).\n5. “combining biological networks in a multiplex generally increases the number of significant modules”. The Figure 3 does not exactly reflect this, it seems the number of significant modules for multiplex(ppi1, ppi2) < monoplex(ppi1) + monoplex(ppi2). Maybe the authors could show the numbers for the union of the modules identified using ppi1and ppi2 and put that into the figure as to what to expect when the modules are combined using these networks separately.\n6. The choice of the for each biological network seems rather arbitrary. Could these weights be optimized (i.e. using leaderboard data) and their effect be tested on final data?\n7. “the community structure of these networks (if they exist) are so unrelated that it is pointless to seek for a common structure by integrating them” I feel that this statement should be supported by further evidence. Could the authors provide some measures in regards to the modularity of each of these networks and the overlap of nodes / edges between them. This would also help to justify the argument on the “DMI biological networks are constructed from different biological sources that might correspond to unrelated community structures”.\n\nMinor:\nExplain what monoplex network means at its first occurrence for readers not familiar with terminology. “hande mulxtiplexes” multiplex networks “following the PASCAL tool” identified using “Considering a greater number of layers...” lacks the main clause / verb “after more than four runs” upto four runs / repetitions “varies in a non-trivial way” non-trivial?\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Partly", "responses": [ { "c_id": "4086", "date": "22 Nov 2018", "name": "Anais Baudot", "role": "Author Response", "response": "1. The multiplex modularity formula lacks the resolution parameter (which should appear before ki kj / 2m in the summation). The authors are encouraged to provide data / figures with respect to the reasoning on the selection of the default resolution parameter (currently they mention verbally that it showed better results on leaderboard).Indeed, the resolution parameter was lacking in the formula. We now have corrected this. Concerning the choice of the default resolution parameter, we made extensive tests that are provided below, but we lack space in the manuscript to detail them. These tests show that the recursion approach provides overall a higher number of significant disease modules in the Dream Module Identification (DMI) challenge framework.Total number of modules obtained when running Molti with 10 randomizations on the 6 DMI challenge network for different values of the resolution parameter:Table 1Number of significant Modules obtained with variations of Gamma:Table 2 Table 3Table 4 Comparison with recursion (apply Molti in a recursive way to communities larger than 100):Table 5 For FDR= 0.1Figure 1For FDR = 0.05Figure 2For FDR = 0.01Figure 3 2. Randomization procedure is unclear. I believe the cost function used is the multiplex modularity given in Eq 1 (L) but the “move” leading to an increase of the modularity is not formally stated (sth like i, j: random_pick(argmax ci, cj L)). The choice of 4 by default is also somewhat arbitrary given that the difference between 3,4,5 are all very small. The significance of the difference in Rand index across runs in Figure 1 could be used to decide the number after which the change is not significant.The cost function is indeed the multiplex-modularity given in Eq 1. At each step, all the moves of an element from a community to another that lead to an increase of the multiplex-modularity are considered, and one of them is randomly picked with a uniform probability. The final number of randomizations was arbitrarily chosen, but several were tested during the Challenge.The choice of 4 randomizations by default was indeed arbitrary, and we removed this sentence in the manuscript to just state that this number is user-defined. The significance of the difference in Rand indexes can also be used to select a correct number of randomizations. However, since the adjusted Rand index is computed with regard to the real communities, it could be used only when the real communities are known, i.e. in our random network simulations. For the real biological networks provided by the challenge, the actual community structure is unknown.Increasing the number of randomizations cannot harm (except with regard to the computation time) since the modularity of the detected communities increases with this number. A possibility could be to test the significance of the modularity increase between two successive numbers of randomizations. However, computing this significance is complex and time-consuming for large networks. 3. The simulation of the multiplex networks could be explained better. Potentially owing to the previous publication, there is no detail in regards to which model was used to generate these networks. A brief description of SBMs and the underlying assumptions of the model it relies on would help. Are these simulated networks are more ER-like? Could this be the reason why the simulated results do not reflect the results on the real networks? Also, it would certainly be useful to add the standard deviation across 2000 networks in Figure 1. I had difficulties following how exactly the mixed and noisy networks are generated (did the mixed ones still have 1000 vertices, 500 from sparse and 500 from dense, for the missing data if the vertices were drawn half of the time, did they have 500 vertices?).Indeed, the simulations have been described in our previous publication, but the current version was lacking details. We extended the method section to describe more precisely how the simulated networks with a known community structure, and their variations are generated.As detailed in the revision, a key property of the SBMs is that, given a community structure, each edge is drawn independently with a probability that depends only the community to which its nodes belong. These property is certainly not granted in biological networks but random graph models taking into account dependance between edges (e.g., exponential random graph models) are difficult to estimate and to simulate. We considered simple SBMs parameterized with a probability for edges between nodes in a same community (intra-community edges) and a probability for edge between nodes belonging to two different communities (inter-community edges). Networks obtained from SBMs are generally not homogeneous, thus quite different from ER ones. The software performing the simulation of multiplex networks from SBMs, called “simul” is now available in the GitHub repository.We tried to display the standard deviation of the adjusted Rand index, but the result was quite confusing since the the error bars tend to overlap themselves (see below for the simulated sparse and dense networks).Figure 4Figure 54. The coverage / precision of disease module discovery is missing. What is the number of all modules identified by the algorithm, how many disease modules are there (or could there be) in total? The latter question could be tricky to answer but I believe a lower and upper bound could be provided (e.g., number of diseases and a number based on the GWAS hits for that disease divided by the min module size).Indeed, the total number of discovered modules was not detailed in the manuscript. It was however used by the PASCAL tool to compute the number of significant disease modules, as the total number of submitted modules is used to correct for multiple testing. We added the total number of discovered valid modules (of size [7-100]) to the figure and table captions.Number of total modules detected by Molti for the results presented on: Figure 2. The numbers used in the Figure 2 have been added to the caption. Table 6 Figure 3: Table 7 Table 2: (nRandom = 10). These numbers have been added to the Table 2 caption No Weights <- 615Weights <- 585 Figure 4: (nRandom = 10).These numbers have been added to the Figure caption. No Weights <- 585Confidence Weights <- 555Inverse Confidence Weights <- 648 The question of how many disease modules exist or could exist is indeed tricky to answer, and is partly discussed in the DMI challenge consortium paper. In this challenge manuscript, the full detail on the 180 GWAS disease/trait datasets, their classification, as well as the scoring method are detailed. Briefly, a module is considered significant if it is enriched in at least one GWAS-associated trait or disease according to the PASCAL approach. A sampling approach is also used to account for robustness in comparing the challengers’ results. The figures 1, 3 and 4 of the consortium paper are in particular and attempt to dig in the reviewer’s proposed direction.https://www.biorxiv.org/content/early/2018/02/15/265553 5. “combining biological networks in a multiplex generally increases the number of significant modules”. The Figure 3 does not exactly reflect this, it seems the number of significant modules for multiplex(ppi1, ppi2) < monoplex(ppi1) + monoplex(ppi2). Maybe the authors could show the numbers for the union of the modules identified using ppi1and ppi2 and put that into the figure as to what to expect when the modules are combined using these networks separately.The modules identified by the multiplex approach cannot be directly compared to the sum of the modules identified independently in the monoplex networks, as the communities might be the sames or very similar. In principle, the two PPI networks should detect similar communities, since they are supposed to contain biological information of the same nature. However, both networks are built from different sources, each one with different features or missing data. Therefore, their combination into a multiplex network should lead to a closer representation of the “real PPI network”, and is expected to result in the detection of a larger number of more accurate communities. In order to compare the communities detected by the two monoplex PPI networks individually with the communities identified when these networks are integrated into a 2-layer multiplex network, we computed the adjusted Rand index [1] between the three corresponding partitions. The adjusted Rand index is the corrected-for-chance version of the Rand index, which is a measure of the similarity between two data clustering [2]. It is usually applied to compare partitions obtained by different clustering algorithms on the same network. Nevertheless, it can be applied to our problem by considering the common nodes between every pair of monoplex network (i.e., considering the partitions reduced to the intersection of the two sets of elements).The results are displayed below in a table and a figure:PPI_1 PPI_2 PPI1_+_PPI2PPI_1 1.0000000 0.1075211 0.2645176PPI_2 0.1075211 1.0000000 0.2906837PPI1_+_PPI2 0.2645176 0.2906837 1.0000000 Figure 6 These results show that the community structure obtained from the multiplex approach is closer to the ones found with both single PPIs. In other words, the communities obtained from single PPI networks are closer to the multiplex than to the communities obtained from the other single network. In addition, 2RandIndex(PPI_1,PPI_2) < RandIndex(PPI_1, PPI1_+_PPI2) and 2RandIndex(PPI_1,PPI_2) < RandIndex(PPI_2, PPI1_+_PPI2). Therefore, it seems that the multiplex approach can reflect better the “real” community structure of the “real PPI” than the addition of both single networks. Finally, the community partition of the multiplex is closer to the one of the PPI_2 than to the one of the PPI_1.6. The choice of the for each biological network seems rather arbitrary. Could these weights be optimized (i.e. using leaderboard data) and their effect be tested on final data?We indeed considered different optimizations of the weights. This point was not clearly stated in the manuscript, but the weights were chosen based on the number of significant modules obtained in the application of Molti on the isolated monoplex networks. We rephrased the corresponding sentence to clarify this.However, we do not think a posteriori that the results of the leaderboard can be used to choose the best weights in the final data : the results obtained using the Leaderboard and the Final GWAS datasets (as well as using the union in the Total GWAS dataset) are not clearly correlated, making difficult to optimize the weights from the Leaderboard results. In addition, the evaluation of the significance of the results by a discrete number of significant disease modules is strongly sensitive to the number of submitted modules, because of the correction for multiple testing. This volatility makes the benchmark difficult to use to evaluate subtle changes in the methods. That’s why we implemented and tested our approach on the simulated networks. Unfortunately, the simulated networks cannot be used to evaluate weights.7. “the community structure of these networks (if they exist) are so unrelated that it is pointless to seek for a common structure by integrating them” I feel that this statement should be supported by further evidence. Could the authors provide some measures in regards to the modularity of each of these networks and the overlap of nodes / edges between them. This would also help to justify the argument on the “DMI biological networks are constructed from different biological sources that might correspond to unrelated community structures”.The modularity is a measure that needs to be computed on a community structure, hence we need first to partition the networks to compute the modularity. To evaluate the modularity of the considered biological networks, we applied Molti with default parameters on the individual networks (as in Figure 3) to obtain the network partitions. We then computed the adjusted Rand index [1] in between the 6 network partitions. Here also, the adjusted Rand index is applied to our problem by considering the common nodes between every pair of monoplex network (i.e., considering the partitions reduced to the intersection of the two sets of elements).The Adjusted Rand Index computed for every pair of partitions of the monoplex networks provided by Molti (10 randomizations) are displayed in the following table and figure: PPI_1 PPI_2 PATH_3 COEX_4 CANCER_5 HOMO_6PPI_1 1.000000000 0.107521064 0.082823431 0.012937084 0.004922555 0.051958946PPI_2 0.107521064 1.000000000 0.091792998 0.010104981 0.004755694 0.023984747PATH_3 0.082823431 0.091792998 1.000000000 0.007815447 0.001958814 0.024252282COEX_4 0.012937084 0.010104981 0.007815447 1.000000000 0.002040522 0.006049318CANCER_5 0.004922555 0.004755694 0.001958814 0.002040522 1.000000000 0.002211310HOMO_6 0.051958946 0.023984747 0.024252282 0.006049318 0.002211310 1.000000000Figure 7 The two PPI networks are the most similar in terms of communities structure, as expected. The community structure of the Pathways network is also quite similar to both PPIs, which makes sense since proteins in the same pathway are more likely to interact physically than those belonging to different pathways. In addition, the Homology network is more similar to those three networks than the Co-expression network. This might come from the fact that this homology network is built by expanding known pathways (Kegg in particular) with the addition of new member related by eukaryotic homology, as detailed in the challenge consortium manuscript. The partition the most different is the one obtained from the cancer network. [1] Santos J, Embrechts M. 2009. On the use of the adjusted rand index as a metric for evaluating supervised classification. In: Alippi C, Polycarpou M, Panayiotou C, Ellinas G, eds. Artificial neural networks—ICANN 2009. Lecture notes in computer science, vol. 5769. Berlin Heidelberg: Springer, 175–184. [2] W. M. Rand (1971). \"Objective criteria for the evaluation of clustering methods\". Journal of the American Statistical Association. American Statistical Association. 66 (336): 846–850. doi:10.2307/2284239. JSTOR 2284239. Minor:Explain what monoplex network means at its first occurrence for readers not familiar with terminology.OK, done in the method section “hande mulxtiplexes” multiplex networksOK, all occurrences of “multiplexes” have been changed to “multiplex networks” “following the PASCAL tool” identified usingChanged “following” to “using” “Considering a greater number of layers...” lacks the main clause / verbOK, changed to “We observed that considering a greater …”“after more than four runs” upto four runs / repetitionsOK, changed to “Increasing the number of randomization runs improves the results up to four runs.”“varies in a non-trivial way” non-trivial?OK, we changed to “is strongly dependent“. This expression was related to the fact that we do not observe a clear correlation between the results obtained on the leaderboard and final GWAS datasets." } ] }, { "id": "35949", "date": "07 Aug 2018", "name": "Lenore J. Cowen", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors explain the method that underlies their submission to the 2016 DREAM Disease Module Identification challenge. The authors only discuss their results from subchallenge 2; they should either say this is what they are going to do up front or also mention their performance on subchallenge 1.\nThe paper is not self-contained, it already assumes some familiarity with the setup of the challenge; since this is being published in a collection on F1000 related to the dream challenge, perhaps that is appropriate, however I would have preferred if the authors had spent more time describing the challenge: i.e. what does it mean in more detail to identify communities from the six biological networks jointly, rather than independently?\nWhile the authors release their code, which is commendable, there are not sufficient details in the text to completely understand their methods without returning to the code. For example, they talk about a resolution parameter γ but this parameter is defined nowhere in their paper: is this a parameter of the GenLouvain method to which they refer?\nThe authors present a randomized algorithm which they run from 1 to 10 times, and return the partition with highest modularity, but figure 1 uses only 1 to 5 randomization runs. Furthermore, the authors do not explain how the parameter \"5\" or \"10\" is selected. There are other method details that involve parameters that are missing, for example in their set of simulated networks, it is impossible to discover how they set their parameters from just this writeup: They write \"we simulated random multiplex networks with a fixed known community structure,... and various topological properties (i.e. dense/sparse/mixed, with/without missing data)\" Is this the same collection of simulated networks that they generated in a previous paper (reference 5 that they cite?) if so, please say this explicitly. Even the writeup in reference 5 is somewhat sketchy, but it's better than what is here.\nThe authors conclude that the power of including multiplex networks is dependent perhaps on the networks being generated from partial views of the same underlying community structure; when networks 5 and 6, which were very different were included, the performance of their method degraded, and this was found across the challenge in general.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? No\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4085", "date": "22 Nov 2018", "name": "Anais Baudot", "role": "Author Response", "response": "The authors explain the method that underlies their submission to the 2016 DREAM Disease Module Identification challenge. The authors only discuss their results from subchallenge 2; they should either say this is what they are going to do up front or also mention their performance on subchallenge 1. We indeed focused on subchallenge 2. During the challenge leaderboard phase, our consortium tested many very different approaches on subchallenge 1, and we did not evaluate Molti thoroughly in this subchallenge. In addition, Molti was specifically designed to leverage multiplex networks. We added a sentence (highlighted in blue) to the manuscript to state this point more clearly. The paper is not self-contained, it already assumes some familiarity with the setup of the challenge; since this is being published in a collection on F1000 related to the dream challenge, perhaps that is appropriate, however I would have preferred if the authors had spent more time describing the challenge: i.e. what does it mean in more detail to identify communities from the six biological networks jointly, rather than independently? We agree that our manuscript was specifically written for the F1000 dream challenge channel: it has strong size constraints and requires some familiarity with the challenge. The consortium manuscript is available in bioRxiv, and is still under consideration for publication in a journal. We wanted to avoid any potential overlap in the interpretation of the results in between the consortium paper and our manuscript. We finally also checked other papers published in DREAM channels and tried to give the same level of details. We however extended the description of the DREAM sub-challenge 2 in the main text. While the authors release their code, which is commendable, there are not sufficient details in the text to completely understand their methods without returning to the code. For example, they talk about a resolution parameter γ but this parameter is defined nowhere in their paper: is this a parameter of the GenLouvain method to which they refer? Thanks for pointing this out. Indeed, we had forgotten the γ parameter and its description in the method section. The γ parameter is a resolution parameter allowing to tune the size of the obtained communities. Increasing this parameter allows reducing the size of the obtained communities. We now present the results obtained by tuning this parameter instead of applying the recursion in the answer to another referee. We also extended the description of the code in the github repo and the description of the simulated networks with SBM to answer the comment below. The authors present a randomized algorithm which they run from 1 to 10 times, and return the partition with highest modularity, but figure 1 uses only 1 to 5 randomization runs. Furthermore, the authors do not explain how the parameter \"5\" or \"10\" is selected. There are other method details that involve parameters that are missing, for example in their set of simulated networks, it is impossible to discover how they set their parameters from just this writeup: They write \"we simulated random multiplex networks with a fixed known community structure,... and various topological properties (i.e. dense/sparse/mixed, with/without missing data)\" Is this the same collection of simulated networks that they generated in a previous paper (reference 5 that they cite?) if so, please say this explicitly. Even the writeup in reference 5 is somewhat sketchy, but it's better than what is here. Number of randomizations We agree that the different number of randomizations used along the manuscript can be confusing. We tried to make this point clearer in the revised version. The number of randomizations needed for a given dataset is not easy to set a priori. Since increasing the number of randomizations improves the modularity - and hopefully the accuracy - of the communities detected, we basically increased the number of randomization until obtaining no significant change in the communities detected. For the simulations, we selected 5 randomizations because the adjusted Rand index plot for 5 randomizations is confounded with that for 4 randomizations. For the Dream Challenge dataset, we considered 10 randomizations because increasing this number was too heavy in computation time. Simulated networks with a community structure The method section has been extended to describe SBM network simulations more precisely. These random networks are generated with the same protocol as the one described in reference 5, but were re-generated for this project. In addition, the code to generate the multiplex networks with SBM (and detect and compare the communities with adjusted Rand index) is now available on the github repository." } ] }, { "id": "36848", "date": "17 Aug 2018", "name": "Yasir Suhail", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nOverview\nThe paper presents: 1. a couple of improvements over the method in the authors' previous 2015 paper on multiplexed modularity, and 2. the application of this improved method to the Disease Module Identification DREAM challenge.\nMethod\nThe paper adequately describes the incremental improvement of the general method over the previous paper regarding the following. 1. The algorithm randomly selects one of the moves improving the modularity at every stage. I assume this helps the method to arrive at different local minima for each run, thereby improving its performance on the nonlinear modularity surface. 2. The incorporation of edge and graph weights into the definition of modularity.\nBoth of these points are well justified, and presented in adequate detail. The only minor detail that is missing is that Equation 1 does not include the resolution parameter.\nResults and Performance Evaluation\nThe improvement in performance due to randomization is presented in Figure 1, while Figure 4 presents evidence of improvement due to the graph (layer) weights.\nThe improvement due to edge weights is not presented, but it logically follows that any network model wherein edges with larger weights are likely to form within modules will show improved performance under a weighted definition of modularity.\nThe authors did not have control over the DREAM challenge evaluation, but a reader might have a few questions regarding the performance evaluation. For example, if the judging criterion is the number of significantly GWAS associated modules, does the score improve by dividing one large disease associated module into two, if they both are still significant?\n\nPresentaion\nOther than the missing resolution parameter, the method is adequately presented. There were a few minor issues with language. For example: 1. The second paragraph of the abstract should end with \"the number of trait and disease communities detected.\" 2. While describing the method of simulating the missing data, another form of the word should be used instead of \" ... we randomly withdrawn vertices of each layer ...\"\nThe authors provide source code for both the latest version, and also the version that was used for the DREAM challenge. It would be helpful if the authors also provide, if those are readily available, the scripts that were used to run the DREAM and simulated data for the results presented. This will help answer minor questions regarding details like filtering etc.\nConclusion and suggestions\nI think the method described by the paper is scientifically valid and its presentation is adequate, other than the minor issues raised above.\nSome of the important questions that may be pursued further are related to the selection of the graph weights. Currently, the weights are selected arbitrarily using the relative performance of the individual networks on predicting the disease modules. Could there be a more systematic manner of weight selection based on both the congruency with what's known of the true modules and the relative redundancy in the various layers?\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Partly\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4087", "date": "22 Nov 2018", "name": "Anais Baudot", "role": "Author Response", "response": "The paper presents:1. a couple of improvements over the method in the authors' previous 2015 paper on multiplexed modularity, and2. the application of this improved method to the Disease Module Identification DREAM challenge. Method The paper adequately describes the incremental improvement of the general method over the previous paper regarding the following.1. The algorithm randomly selects one of the moves improving the modularity at every stage. I assume this helps the method to arrive at different local minima for each run, thereby improving its performance on the nonlinear modularity surface.2. The incorporation of edge and graph weights into the definition of modularity. Both of these points are well justified, and presented in adequate detail. The only minor detail that is missing is that Equation 1 does not include the resolution parameter.Thanks for pointing out this omission. The resolution parameter has been included to the Equation 1 in the revised version of the manuscript. Results and Performance Evaluation The improvement in performance due to randomization is presented in Figure 1, while Figure 4 presents evidence of improvement due to the graph (layer) weights. The improvement due to edge weights is not presented, but it logically follows that any network model wherein edges with larger weights are likely to form within modules will show improved performance under a weighted definition of modularity.The tests of edge weights is indeed not presented as a figure, but some data are summarized as a table (Table 2). The authors did not have control over the DREAM challenge evaluation, but a reader might have a few questions regarding the performance evaluation. For example, if the judging criterion is the number of significantly GWAS associated modules, does the score improve by dividing one large disease associated module into two, if they both are still significant?This is a tricky question. Briefly, the DREAM evaluation uses the PASCAL tool to compute a score for each modules according to their enrichments in GWAS-significant genes. The number of significant disease module in a given partition is then computed according to an enrichment threshold, and considering correction for multiple testing. In this context, the total number of submitted valid communities is important, as submitting more modules will induce a stronger correction for multiple testing. It is thereby not straightforward to evaluate if dividing a large disease module into two will lead to two submodules both still significant, as this will also increase the correction for multiple testing. In other words, submitting more modules can affect badly the results even if without multiple testing correction it would have increased the number of significant disease modules. The PASCAL tool also takes into account the GWAS statistics to compute the significance, it is thus overall very hard to use the obtained discrete results (i.e., the number of significant disease modules) to tune/optimize slight variations of the approach, as the variability in the results can come from many factors. Some of these points are discussed in the DMI challenge manuscript as some of the top-performers of the subchallenge 1 selected the modules based on their sizes without working on the improvement of the algorithms per se.Presentaion Other than the missing resolution parameter, the method is adequately presented. There were a few minor issues with language. For example:1. The second paragraph of the abstract should end with \"the number of trait and disease communities detected.\"Corrected2. While describing the method of simulating the missing data, another form of the word should be used instead of \" ... we randomly withdrawn vertices of each layer ...\"We changed the instances of “withdraw” to “remove”The authors provide source code for both the latest version, and also the version that was used for the DREAM challenge. It would be helpful if the authors also provide, if those are readily available, the scripts that were used to run the DREAM and simulated data for the results presented. This will help answer minor questions regarding details like filtering etc.The scripts used to run the DREAM challenge approaches are all available from the challenge webpages as a community resource.Conclusion and suggestions I think the method described by the paper is scientifically valid and its presentation is adequate, other than the minor issues raised above. Some of the important questions that may be pursued further are related to the selection of the graph weights. Currently, the weights are selected arbitrarily using the relative performance of the individual networks on predicting the disease modules. Could there be a more systematic manner of weight selection based on both the congruency with what's known of the true modules and the relative redundancy in the various layers?This is a very interesting question. As noted by the reviewer, we selected the weights based on the relative performance on the individual networks. However, this could not be done in a more systematic way, as true modules are not known (and in fact the challenge goal was to defined a benchmark as a proxy for these true modules). Another approach could be to weight the different layers based on our a priori knowledge/trust of the different layers. However, we have difficulties understanding how it would be possible in practice to weight according to the relative redundancy of the various layers." } ] }, { "id": "36849", "date": "17 Aug 2018", "name": "Arda Halu", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn their manuscript entitled “Identifying communities from multiplex biological networks by randomized optimization of modularity,” Didier et al. apply a network clustering method to identify communities that are significantly enriched in disease signatures. They build on their previously published community detection method, which is based on the greedy optimization (following the algorithm by Louvain et al.) of multiplex modularity. The study is submitted as part of the DMI DREAM Challenge, which aims to evaluate different clustering algorithms by how well the resulting clusters are associated with GWAS-implicated disease genes. The authors test their approach on simulated datasets as well as DMI benchmark datasets. They have three improvements on their original method: randomization, edge and layer weights, and recursive clustering for large clusters.\n\nThe DMI DREAM Challenge itself is an important exercise addressing the non-trivial task of identifying biologically meaningful communities in molecular networks. This paper by Didier et al. takes on this challenge by treating the benchmark datasets as a multiplex network. While limited by the constraints of the challenge, I think the paper is an important contribution that offers an insight as to how multiplex methods fare on real-world biological networks of diverse and not necessarily related sources.\n\nThe paper is well written, but it has room for improvement, especially in the concise way information is presented in the Methods and Results section. Most of my major comments relate to the clarification of details I thought to be missing from Methods and Results. In various places in the manuscript, the reader is assumed to be familiar with the DREAM Challenge and the previous paper [5] on which this paper is based. The Methods section should thus be expanded to render the study more accessible and self-contained. In some places, the results could be interpreted better. Below are my suggestions meant to help the authors improve the presentation of their study:\n\nMajor comments:\n\n1) Randomly generated synthetic networks: Even if previously published in [5], it would be helpful if this were described a little more in detail. Providing details about how the community structure is defined or what exactly balanced communities means would be helpful to the reader.\n\n2) More information is needed for the network types, even if they’re described elsewhere under the umbrella of the DMI challenge. What were the sources for each network? Which organism are the PPI networks derived from? Are the co-expression networks tissue-specific or not? Details like these would be informative when interpreting the results of Figure 3 from a biological standpoint.\n\n3) It is important to clarify how the multiplex network was constructed out of the six networks. For it to be a multiplex network, the same set of nodes should be represented on each layer, which likely requires the pruning of the six benchmark networks. How many nodes did the final multiplex consist of? How many edges were on each layer? Is there any specific inter-layer link structure?\n\n4) The authors use PASCAL to determine the disease-associated genes from GWAS, which are in turn used to determine the communities that are significantly enriched in disease signals. It is crucial at this point to convey to the reader very clearly what PASCAL does and how it’s translated to “significant disease modules.” The authors should, with a couple of sentences, describe how PASCAL does the SNP-to-gene mapping, and then, more importantly, how the p-value weighted gene annotations from PASCAL are used in the enrichment to determine the significant disease modules. A good description of this would facilitate the interpretations of Figure 2-4.\n\n5) The Methods section on controlling the size of modules could be more informative. “We tested different pre-filters (pruning leaves), parameters (resolution parameter, recursions, combination of graph weights for multiplexes) and post-filters (density, size, pruning leaves) in each leaderboard round.” This sentence contains a lot of information without really getting into details. The authors could elaborate on each item. For instance, the resolution parameter does not really seem to come into play (it is even omitted from the definition of multiplex modularity since it is set to 1). Recursion was used to limit the size of the clusters, and while the authors mention tests conducted in leaderboard rounds, no data here is shown as to how varying the resolution parameter changes the results. Perhaps the authors could either omit the section about the resolution parameter altogether or provide some supplementary figures on it.\n\n6) The 100 cutoff is set by the challenge, but in other settings, is there a way to set this ad hoc limit more concretely? Could the authors comment on this?\n\n7) In Figure 1, the authors note that the randomizations improve the accuracy of the detected communities, in particular for dense multiplex networks. This is interesting. May that suggest that dense networks have more possibilities whereby local maxima in the modularity landscape can lead to better results than the best solution? Can the authors comment on possible reasons why this could be the case?\n\n8) What Figure 2 tells me is that if we simply relax the association criterion (FDR cutoff), we’ll simply get more enriched communities, which is not entirely surprising. Here I think the distinction between the different GWAS datasets is important to discuss, if the authors are saying the results are sensitive to the datasets. What is it that makes the “Leaderboard” and “Final” datasets different? What are the diseases and traits included in the GWAS datasets? Perhaps the authors could comment on this.\n\n9) Also for Figure 2 (I may be missing something obvious here) – are these results based on multiplex or monoplex?\n\n10) “We hypothesize that the community structures of these networks (if they exist) are so unrelated that it is pointless to seek for a common structure by integrating them.” This is an interesting point that should remind us that the multiplex biological network should be constructed with a hypothesis in mind (e.g. various molecular levels from transcripts to proteins to metabolites) rather than piling up different networks into a multiplex structure. There is not much reason to think that just adding different sources of biological information should increase the performance of detecting disease modules. The tradeoff between signal and noise with the addition of diverse biological layers can be added as a point of discussion.\n\n11) Multiplex versus monoplex: The notion that multiplex networks help identify a greater number of significant disease modules than monoplex networks combined is an exciting prospect. However, the evidence shown in Figure 3 is a little scant. How was the comparison done exactly? Were the number of significant modules from each separate network summed up and compared to the one from multiplex? These kinds of details should be included.\n\n12) I found the Discussion a bit too DREAM Challenge-oriented. I think the parts comparing the results with those of the top performer are unnecessary (but this may be a requirement of the challenge so ignore if that's the case). The authors should revise the discussion to recapitulate and interpret their results, and to highlight the advantages of using the multilayer approach. I think it is commendable that the authors chose to include all of the six datasets, regardless of how related they are. Even the fact that the addition of some (possibly orthogonal) layers is detrimental to the outcome is an important finding.\n\nMinor points: 1) In the definition of multiplex modularity: K^(g)_i should be lowercase, i.e. k^(g)_i.\n\n2) Also in the definition of multiplex modularity, I would suggest the authors use s_i instead of k_i if they’re dealing with weights to denote the strength of the node rather than the degree. This is an optional point, but it would make it more aligned with the network science literature and nomenclature.\n\nTypos: - “biological networks are nowadays assembled on a large-scale.” --> on a larger scale - “picks the move that increases the most modularity” --> picks the move that increases modularity the most - “data in which we randomly withdrawn vertices” --> data in which we randomly withdraw (or just say remove) vertices - “but to a limited extend after more than four runs” --> to a limited extent - “obtained by considering or not these weights in the MolTi partitioning” --> obtained by considering and not considering these weights in the MolTi partitioning\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Partly\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4088", "date": "22 Nov 2018", "name": "Anais Baudot", "role": "Author Response", "response": "In their manuscript entitled “Identifying communities from multiplex biological networks by randomized optimization of modularity,” Didier et al. apply a network clustering method to identify communities that are significantly enriched in disease signatures. They build on their previously published community detection method, which is based on the greedy optimization (following the algorithm by Louvain et al.) of multiplex modularity. The study is submitted as part of the DMI DREAM Challenge, which aims to evaluate different clustering algorithms by how well the resulting clusters are associated with GWAS-implicated disease genes. The authors test their approach on simulated datasets as well as DMI benchmark datasets. They have three improvements on their original method: randomization, edge and layer weights, and recursive clustering for large clusters.The DMI DREAM Challenge itself is an important exercise addressing the non-trivial task of identifying biologically meaningful communities in molecular networks. This paper by Didier et al. takes on this challenge by treating the benchmark datasets as a multiplex network. While limited by the constraints of the challenge, I think the paper is an important contribution that offers an insight as to how multiplex methods fare on real-world biological networks of diverse and not necessarily related sources.The paper is well written, but it has room for improvement, especially in the concise way information is presented in the Methods and Results section. Most of my major comments relate to the clarification of details I thought to be missing from Methods and Results. In various places in the manuscript, the reader is assumed to be familiar with the DREAM Challenge and the previous paper [5] on which this paper is based. The Methods section should thus be expanded to render the study more accessible and self-contained. In some places, the results could be interpreted better. Below are my suggestions meant to help the authors improve the presentation of their study:We agree that our manuscript requires familiarity with the DMI DREAM challenge data and consortium manuscript: It is indeed not independant, and was specifically written as a companion paper to be published with papers from other participants for the F1000 dream challenge channel. We apologize for it’s strong size constraints, and the fact that we tried to avoid any overlap with the main consortium paper, available in bioRxiv and considered for publication in a journal. We checked and tried to give the same amount of details as the other papers published in DREAM channels, in particular in the DMI channel.Major comments: 1) Randomly generated synthetic networks: Even if previously published in [5], it would be helpful if this were described a little more in detail. Providing details about how the community structure is defined or what exactly balanced communities means would be helpful to the reader.This point was also stated by another reviewer, and we now have extended the method section describing the generation of random network with a known community structure. We simulated networks with 20 balanced communities, meaning that the communities have the same size (regarding the number of nodes).2) More information is needed for the network types, even if they’re described elsewhere under the umbrella of the DMI challenge. What were the sources for each network? Which organism are the PPI networks derived from? Are the co-expression networks tissue-specific or not? Details like these would be informative when interpreting the results of Figure 3 from a biological standpoint.We agree that more information about the individual networks can help deepening the understanding of some of the results presented in the article, but the networks are an important point and deeply described in the main DMI challenge consortium paper available on biorXiv. This challenge paper is still considered for publication in a journal, and we are expected to avoid any overlap. We however now state in the method section that all the data are human.3) It is important to clarify how the multiplex network was constructed out of the six networks. For it to be a multiplex network, the same set of nodes should be represented on each layer, which likely requires the pruning of the six benchmark networks. How many nodes did the final multiplex consist of? How many edges were on each layer? Is there any specific inter-layer link structure?This is a very important and interesting point. The integration of biological networks obtained from different sources into a multiplex network is indeed not straightforward, as the different networks contain different data. In the multiplex framework, it is not mandatory to prune the networks to get the same set of nodes: multiplex networks share the same set of vertices (i.e., the union of all the nodes from all layer), as defined in [1]. Each layer provides information for a subset of the nodes, and the information for the remaining nodes is unknown and considered as missing data. In this context, the number of nodes to be clustered is the union of nodes in each layer, not the intersection. It is also in this context that we tested network layers with missing data in the simulations.The multiplex modularity is equal to the sum of the individual modularities of the graphs of the multiplex with regard to the same community partition. Therefore, there is no need for an inter-layer link structure when constructing the multiplex network. [1] Kivelä M, Arenas A, Barthelemy M, Gleeson JP, Moreno Y, Porter MA. 2014. Multilayer networks. Journal of Complex Networks 2:203–271 4) The authors use PASCAL to determine the disease-associated genes from GWAS, which are in turn used to determine the communities that are significantly enriched in disease signals. It is crucial at this point to convey to the reader very clearly what PASCAL does and how it’s translated to “significant disease modules.” The authors should, with a couple of sentences, describe how PASCAL does the SNP-to-gene mapping, and then, more importantly, how the p-value weighted gene annotations from PASCAL are used in the enrichment to determine the significant disease modules. A good description of this would facilitate the interpretations of Figure 2-4.We rephrased the corresponding method section to mention p-value and multiple testing corrections. However, we have for this companion paper a strong size constraints. The PASCAL tool is a complex approach deeply described in a publication (Lamparter et al. Plos Computational Biology), and its application and use in the context of the DMI dream challenge is fully described in the consortium paper. A precise and correct description of the evaluation procedure, from SNP mapping to significant modules would be too long and overlapping with the consortium and PASCAL papers.5) The Methods section on controlling the size of modules could be more informative. “We tested different pre-filters (pruning leaves), parameters (resolution parameter, recursions, combination of graph weights for multiplexes) and post-filters (density, size, pruning leaves) in each leaderboard round.” This sentence contains a lot of information without really getting into details. The authors could elaborate on each item. For instance, the resolution parameter does not really seem to come into play (it is even omitted from the definition of multiplex modularity since it is set to 1). Recursion was used to limit the size of the clusters, and while the authors mention tests conducted in leaderboard rounds, no data here is shown as to how varying the resolution parameter changes the results. Perhaps the authors could either omit the section about the resolution parameter altogether or provide some supplementary figures on it.Indeed, the resolution parameter was missing in the equation, and we now have included it. In addition, we removed from the manuscript the sentence describing the various test we made as this was indeed possibly confusing.Concerning the choice of the default resolution parameter, we made extensive tests that are provided below, but we lack space in the manuscript to detail them. These tests show that the recursion approach provides overall a higher number of significant disease modules in the Dream Module Identification (DMI) challenge framework.Total number of modules obtained when running Molti with 10 randomizations on the 6 DMI challenge network for different values of the resolution parameter:Table 1Number of significant Modules obtained with variations of Gamma:Table 2Comparison with recursion (apply Molti in a recursive way to communities larger than 100): Table 3 For FDR= 0.1Figure 1For FDR = 0.05Figure 2For FDR = 0.01Figure 3 6) The 100 cutoff is set by the challenge, but in other settings, is there a way to set this ad hoc limit more concretely? Could the authors comment on this?Defining the optimal module size is a complex task. One could imagine using the challenge benchmark to find the optimal module size. However, this is not feasible in practice as 1) the computation of PASCAL scores are time-consuming and cannot really be used to optimize the algorithms, 2) the sensitivity/volatility of the results to the number of submitted modules and GWAS data used make the interpretation on a fine-grained scale difficult.The consortium paper discuss this point and identify, in particular, that the average module size do not correlate with the number of significant modules identified by the different approaches (Figure 3C in bioRxiv). It is also interesting to note that some modules are submodules of larger modules.7) In Figure 1, the authors note that the randomizations improve the accuracy of the detected communities, in particular for dense multiplex networks. This is interesting. May that suggest that dense networks have more possibilities whereby local maxima in the modularity landscape can lead to better results than the best solution? Can the authors comment on possible reasons why this could be the case?This is a very interesting point. We were actually surprised by the extent to which randomization improves the community detection on simulated networks. We initially thought that the greater increase in the accuracy of the communities detected on dense multiplex networks came from the fact that there was more room for improvement in this case. Your point makes sense since one can expect a greater number of local maxima for the modularity for dense network than for sparse ones, in addition to more possibilities of moving vertices during the Louvain execution. This point definitely deserves to be further investigated.8) What Figure 2 tells me is that if we simply relax the association criterion (FDR cutoff), we’ll simply get more enriched communities, which is not entirely surprising. Here I think the distinction between the different GWAS datasets is important to discuss, if the authors are saying the results are sensitive to the datasets. What is it that makes the “Leaderboard” and “Final” datasets different? What are the diseases and traits included in the GWAS datasets? Perhaps the authors could comment on this.The number of significant communities increases when the FDR cutoff is larger, as expected. However, the results are also quite sensitive to the dataset considered as pointed out by the reviewer. The different GWAS datasets were chosen by the DREAM challenge organizers, and splited into “Leaderboard” and “Final” manually. A detailed discussion of the diseases and traits included in the GWAS datasets, as well as their comparison, is proposed in the DMI consortium manuscript (in particular in Figures 1, 3 and 4) and cannot be reproduced here.9) Also for Figure 2 (I may be missing something obvious here) – are these results based on multiplex or monoplex?In order to clarify this point, we now state in the figure caption that the results are based on the multiplex networks. 10) “We hypothesize that the community structures of these networks (if they exist) are so unrelated that it is pointless to seek for a common structure by integrating them.” This is an interesting point that should remind us that the multiplex biological network should be constructed with a hypothesis in mind (e.g. various molecular levels from transcripts to proteins to metabolites) rather than piling up different networks into a multiplex structure. There is not much reason to think that just adding different sources of biological information should increase the performance of detecting disease modules. The tradeoff between signal and noise with the addition of diverse biological layers can be added as a point of discussion.Thanks for this interesting remark. We agree that multiplex biological networks should be constructed with a hypothesis in mind, and that the network that are piled up in a multiplex are expected to be realisations of the true underlying biological modules. This is the case for the two PPI network and the pathway network that are expected to represent cellular processes. In addition, pairs of proteins belonging to the same pathway or co-expressed are more likely to physically interact than pairs of random proteins [1]. Therefore, even if the biological information contained in co-expression or pathways networks is completely different to the one contained in PPI networks, they still can improve the detection of the communities by adding more signal to the system and reducing the bias and noise of individual networks as we previously demonstrated [2]. The cancer network, contrarily, might contain very different biological information.However, concerning the bias and noise present in the different networks, as the bias and noise are not the same, we see the combination of the different sources in a more positive way, as not adding noise but complementing each other to reduce noise. For instance, if we take individual PPI networks obtained from different sources and we integrate them into a multiplex PPI network, we expect to obtain better clustering results than with the individual networks alone. All these monoplex PPI networks try to reflect the real PPI network containing all the human genes, all their interactions and their real community structure. However, they have missing data and they strongly depend on the technologies used to get the interactions (bias and noise). Therefore, since all of them attend to reflect the same system, the biological signal increases and the noise and missing data tend to reduce. We added a sentence to the discussion concerning this point. [1] Rual JF, Venkatesan K, Hao T, et al.: Towards a proteome-scale map of the human protein-protein interaction network. Nature. 2005; 437(7062): 1173–1178.[2] Didier G, Brun C, Baudot A: Identifying Communities from Multiplex Biological Networks. PeerJ. 2015; 3: e1525.11) Multiplex versus monoplex: The notion that multiplex networks help identify a greater number of significant disease modules than monoplex networks combined is an exciting prospect. However, the evidence shown in Figure 3 is a little scant. How was the comparison done exactly? Were the number of significant modules from each separate network summed up and compared to the one from multiplex? These kinds of details should be included.The comparison between the different approaches is done following the DMI challenge benchmark, that is by computing the number of significant disease modules (with a FDR threshold of 0.1). We computed this number of significant modules identified from the 6 biological networks independently, and then from combination of networks thanks to the multiplex framework . We used combinations of 2 to 6 networks, considered as multiplex networks of 2 to 6 layers.12) I found the Discussion a bit too DREAM Challenge-oriented. I think the parts comparing the results with those of the top performer are unnecessary (but this may be a requirement of the challenge so ignore if that's the case). The authors should revise the discussion to recapitulate and interpret their results, and to highlight the advantages of using the multilayer approach. I think it is commendable that the authors chose to include all of the six datasets, regardless of how related they are. Even the fact that the addition of some (possibly orthogonal) layers is detrimental to the outcome is an important finding.The manuscript is indeed DREAM challenge oriented as it is not an independent manuscript but a companion paper. Regarding the interpretation of the multiplex approach, we think it is important to state that the top-performers selected a subset of the 6 networks, reflecting also the reviewer’s comment on the choice of the network to be piled up. We agree that a lot could still be discussed regarding our (and others) obtained results in this challenge, and extended a bit the discussion. This point is also discussed in the challenge consortium manuscript.Minor points:1) In the definition of multiplex modularity: K^(g)_i should be lowercase, i.e. k^(g)_i.Thanks for pointing this out, it has been changed to s following your next comment.2) Also in the definition of multiplex modularity, I would suggest the authors use s_i instead of k_i if they’re dealing with weights to denote the strength of the node rather than the degree. This is an optional point, but it would make it more aligned with the network science literature and nomenclature.DoneTypos:Thanks for pointing all these typos !- “biological networks are nowadays assembled on a large-scale.” --> on a larger scaleDone- “picks the move that increases the most modularity” --> picks the move that increases modularity the mostDone- “data in which we randomly withdrawn vertices” --> data in which we randomly withdraw (or just say remove) verticesDone- “but to a limited extend after more than four runs” --> to a limited extentThis sentence has been removed after other reviewer’s comment.- “obtained by considering or not these weights in the MolTi partitioning” --> obtained by considering and not considering these weights in the MolTi partitioningDone" } ] } ]	1	https://f1000research.com/articles/7-1042
https://f1000research.com/articles/7-1838/v1	22 Nov 18	{ "type": "Research Article", "title": "Alcoholic bitters modulates sex hormones and some biochemical parameters of testicular function in male Wistar rats", "authors": [ "Omowumi T. Kayode", "Abolanle A.A. Kayode", "Charles O. Nwonuma", "Abolanle A.A. Kayode", "Charles O. Nwonuma" ], "abstract": "Background: Alcoholic bitters have been acclaimed to boost sexual function and fertility in animals but there is no reported scientific evidence that evaluated its effects on the normal functioning of the testes. This study was therefore conducted to assess the effect of some alcoholic bitters on testicular function indices of male Wistar rats. Methods: A total of 25 male Wistar rats were assigned into five groups of five animals each and treated with distilled water, ethanol, Alomo, Striker and Orijin Alcoholic Bitters at 0.2, 0.2, 0.2, 0.16 and 0.3 ml/kg body weight respectively for 28 days. The animals were thereafter sacrificed and the serum obtained was used for the determination of sex hormones. Assessment of testosterone, luteinizing hormone (LH), follicle stimulating hormone (FSH), cholesterol, and malondialdehyde (MDA) concentrations as well as the activity of β-Hydroxy β-methylglutaryl-CoA reductase (HMG-CoA reductase), superoxide dismutase (SOD) and catalase (CAT) were carried out using standard methods. Results: There were significant (p < 0.05) increases in protein, cholesterol, testosterone, FSH and LH, as well as in the activity of HMG-CoA reductase, SOD and CAT in all the groups of animals administered the alcoholic bitters, whereas concentration of MDA was significantly reduced (p<0.05). Concentration of triglycerides was not significantly different (p>0.05) from those of the control animals. Conclusion: The alcoholic bitters enhanced the normal functioning of the testes, the antioxidant enzymes and the release of the reproductive hormones. This may partly explain its use in boosting sexual function and fertility in male rats.", "keywords": [ "Alcoholic bitters", "Cholesterol", "Testosterone", "Antioxidant enzymes", "Testicular function." ], "content": "Introduction\n\nInfertility describes prolonged difficulty at achieving success with conception between a male and female having unhindered access to sexual activity1.\n\nA male’s inability to fertilize the female’s ovum for upwards of twelve months of unrestricted intercourse categorizes him as infertile. This challenge is represented globally across continents where up to about fifteen percent of heterosexual couples are infertile and forty percent of these represents male factor infertility and sexual inadequacies1\n\nFor centuries, plants and plant-based products have been utilized as an important and safe natural source of medicines for treating different health challenges2. Various plant extracts have been acclaimed to have aphrodisiac potentials and utilized traditionally to enhance sexual performance and fertility3,4. Some of these extracts are produced, packaged and sold as bitters drink, which has gained global popularity in recent times.\n\nBitters are liquid preparations (regularly alcoholic) produced from herb and root concentrates of tropical and subtropical plants5. They have been utilized over several decades for their acclaimed set of medicinal benefits, such as treatment of kidney and bladder dysfunction, regulation of blood pressure, management of indigestion, menstrual cramps, ulcers, gastritis, insomnia, stress, depression, excessive weight and sexual inadequacies4,6. The most commonly consumed alcoholic bitters in Omu aran, Nigeria are Alomo, Orijin and Striker bitters. The rising popularity of these bitters despite their severe bitter taste and appalling scent could be attributed to the alleged preferences of being effective medicinally, cheap and readily available.\n\nThis study therefore was aimed to study the effects of these alcoholic bitters on some testicular indices in order to justify its traditionally acclaimed use as an aphrodisiac and pro-fertility agent.\n\n\nMethods\n\nBottles of Alomo, Orijin and Striker were purchased from the main market in Omu Aran, Kwara State, Nigeria.\n\nTwenty five healthy, adult male Wistar rats (110–130 g) which made up five groups of five animals each, were obtained from the Animal Holding Unit of the Department of Biochemistry, University of Ilorin, Ilorin, Kwara State. The animals were housed in plastic rat cages placed in well-ventilated rooms (photo-period: 12 h light and 12 h dark cycle; temperature: 25–27°C; humidity: 45–55%) and fed with standard rat pellet and water ad libitum.\n\nReagent kits used for the assay for total cholesterol (Lot no: 416569) and triglyceride (Lot no: 435212) were products of Randox Laboratory Limited, UK. The assay kits for testosterone, follicle-stimulating hormone (FSH) and leutinizing hormone (LH) were products of Monobind Inc, Lake Forest, CA, USA. All other reagents used were of analytical grade and were prepared in volumetric flask using glass-distilled water.\n\nThe Wistar rats were assigned into one of five groups consisting of five animals each as shown below:\n\nGroup A is the control group which received 0.2ml/kg body weight distilled water;\n\nGroup B received 0.2ml/kg body weight ethanol\n\nGroup C received 0.2ml/kg body weight Alomo bitters\n\nGroup D received 0.16ml/kg body weight Strikers bitters\n\nGroup E received 0.3ml/kg body weight Orijin bitters\n\nAdministration of the herbal bitters, ethanol and water was performed orally once daily for 28 days, after which the rats were sacrificed by placing them in a jar containing diethyl ether. Testicular tissue was excised and used for the biochemical assays.\n\nBlood samples were collected by venal puncture into plain sample bottles subsequent to sacrifice of the rats, then centrifuged at 5000rpm and the serum was stored at 20°c prior to the subsequent hormonal assay.\n\nOne half of the testes for each rat were homogenized in ice-cold 0.25M sucrose solution. The homogenates were diluted using a dilution factor 1:9 w/v ratio. The diluted homogenates were further centrifuged at 5000 rpm for 10 minutes and then stored for biochemical assays.\n\nTotal lipid was extracted from the second half of the testicular tissue by a mixture of chloroform methanol (2:1, v/v) as described by Folch et al.7. Aliquots of total lipid extract were used for quantitative analysis of cholesterol and triglycerides8.\n\nThe serum concentrations of testosterone, luteinizing hormone (LH) and follicle stimulating hormone (FSH) were quantitatively determined from the serum by adopting the procedure highlighted by Tietz9.\n\nThe total protein concentration was estimated in the testes homogenates using the Biuret method10. Twenty microlitres of serum were mixed with 1000 μl of Biuret reagent (6 mmol/L potassium iodide, 21 mmol/L potassium sodium tartarate, 6 mmol/L copper sulphate, and 58 mmol/L sodium hydroxide). Thereafter, the mixture was incubated for 10 min at 37°C and the absorbance taken at 546 nm. Bovine serum albumin was used as the standard protein and the total protein was subsequently calculated using the formula:\n\nTotal protein concentration (g/dL)= (Absorbance of sample/Absorbance of standard) × 6\n\nDetermination of total cholesterol and triglyceride concentration were carried out using enzymatic kits procured from Randox laboratories11,12.\n\nHMG- CoA reductase activity. HMG-CoA reductase activity was measured in testes homogenate using the procedure of Rao and Ramakrishnan13. The ratio of HMG-CoA to mevalonate was taken as an index of enzyme activity which catalyzes the conversion of HMG to mevalonate. The lower the ratio, the higher the enzyme activity. Equal volumes (0.5 mL each) of tissue homogenate and diluted perchloric acid (50 mL/L) were mixed together. This was allowed to stand for 5 minute and centrifuged at 2000 rpm for 10 min. This was filtered and 1 mL of filtrate was mixed with 0.5 mL of freshly prepared hydroxyl amine (2 mol/L) reagent of pH 5.5 for HMG-CoA and with 0.5 mL of freshly prepared hydroxyl amine (2 mol/L) reagent of pH 2.1 for mevalonate. After 5 minutes, 1.5 mL of ferric chloride reagent (prepared by dissolving 5.2 gm of trichloroacetic acid and 10 gm of ferric chloride in 50 mL of 0.65 mol/L HCl) was added to each of the test tube for HMG-CoA and mevalonate. The tubes were well shaken. Absorbance was read after 10 min at 540 nm versus a similarly treated arsenate blank using double beam UV-Visible spectrophotometer.\n\nSuperoxide dismutase (SOD). SOD was assayed using the standard method of Misra and Fridovich et al.14. Based on the inhibition of Epinephrine-Adrenochrome transition by the enzyme. A 0.5ml of testes homogenate was taken in a reaction vessel and then 0.5ml of distilled water was added to dilute the sample. To this 0.25ml of ice cold ethanol and 0.15ml of chloroform were added to precipitate the reaction mixture. The reaction mixture was well shaken for about 5minutes at 4°C and then centrifuged. The enzyme activity in the supernatant solutions was determined using spectrophotometer. The Adreno-Chrome produced in the reaction mixture, contains 0.2ml of EDTA (0.6mM), 0.4ml of Na2CO3 (0.25M) and 0.2ml of Epinephrine (3.0mM), final volume was adjusted to 2.5ml and then the absorbance readings were measured at 420nm in a UV-Visible recording Spectrophotometer. The Transition of Epinephrine to Adreno-Chrome was determined by the addition of the required quantity of enzyme to assess the enzyme activity expressed in terms of units/minute/mg protein.\n\nCatalase (CAT). CAT was assayed using standard protocol15. The breakdown of H2O2 on addition of the enzyme is followed by absorbing the decrease in light absorption of peroxide solution in the UV region was determined. A 3ml of reaction mixture containing 1.9ml of phosphate buffer, [(0.05M) of pH 7.0], 1.0ml of substrate H2O2 and 0.1ml of testes homogenate was used in this assay. The activity was measured as change in optical activity/density at 240nm at 30sec interval for about 3 minutes. The CAT activity was expressed in terms of mole of H2O2 consumed/minute/mg protein.\n\nMalondialdehyde (MDA) concentration. Thiobarbituric acid reacts with MDA giving rise to a high absorptivity adduct which was assessed with a spectrophotometer at 531 nm. A standard graph was plotted, and concentration of MDA was expressed as nmol/ml16.\n\nData were expressed as mean ± SEM of five replicates. The data were subjected to statistical analysis using Analysis of Variance and complemented with Duncan post hoc. The analyses were done with IBM SPSS statistics, version 22.0 (SPSS Inc; Chicago USA). Differences were considered statistically significant at p < 0.05.\n\nAll procedures performed on the animal subjects were in accordance with the ethical standards of Landmark University Animal Care Committee (approval number LUAC-0031A). All efforts were made to ameliorate any suffering to the animals in the course of treatment by following careful intubation procedures and also by anaesthetising the animals prior to sacrifice to prevent their experiencing any form of associated pain.\n\n\nResults\n\nThe serum level of testosterone, FSH and LH significantly increased (p<0.05) in all bitters administered groups compared to the control group and the ethanol group (Table 1).\n\nData are mean of five determinations ± SEM. Values with superscripts different from the control for each parameter are significantly different (p<0.05)\n\nThe administration of the Alomo, Striker and Orijin alcoholic bitters as well as ethanol to the animals resulted in a significant increase (p<0.05) in protein concentration, HMG-CoA reductase activity, and cholesterol levels when compared to the control group. There was however no significant increases (p>0.05) in the levels of the triglycerides (Table 2).\n\nCholesterol and triglycerides were determined from total lipid extracted from testicular tissue; HMG-CoA reductase and protein concentration were determined from homogenized testicular tissue.\n\nHMG-CoA: β -Hydroxy β-methylglutaryl-CoA reductase. Data are mean of five replicates ± SEM . Values with superscripts different from the control for each parameter are significantly different (p<0.05)\n\nThere was a significant increase (p<0.05) in MDA concentration in the ethanol control group compared to the control group. However, the concentration of MDA in the bitters treated groups significantly reduced (p<0.05) compared to the ethanol group (Table 3). Furthermore, there was a significant increase in the activities of SOD and catalase in the ethanol and bitters administered groups compared to the control group. The upregulation of enzyme activity was however significantly higher (p<0.05) in the bitters group compared to the ethanol group.\n\nData are mean of five replicates ± SEM . Values with superscripts different from the control for each parameter are significantly different (p<0.05)\n\n\nDiscussion\n\nThe administration of alcoholic bitters to the rats for the period of 28 days did not result in obvious physical toxicity to the animals. However, their behavioural pattern after the consumption of these bitters showed there was a sharp increase in their physical activity and agility, which was noticeable within the first few hours of the bitters and ethanol administration to the animals. This may be associated with the alcoholic content in these bitters. Another noticeable change was the improved appetite of these animals post-administration of the bitters, which supports a previous report that suggests that bitters promote appetite and digestion17.\n\nThe increased testicular protein concentration in the bitters-treated animals may indicate the upregulation of protein synthesis in the testes, which may positively enhance testicular function and spermatogenesis18.\n\nMalondialdehyde (MDA), which is a by-product of lipid peroxidation19 was only elevated in the animals administered ethanol and not alcoholic bitters in the testicular tissue, suggesting that alcohol treatment alone may foster generation of reactive oxygen species leading to oxidative stress, a condition that precedes the onset of several diseases in humans20. Of interest is the involvement of oxidative stress in the aetiology of testicular perturbation and altered spermatogenic function, which may adversely affect male fertility and sexual function21. The observed reduction of MDA for all the groups administered alcoholic bitters might indicate an antioxidative potential of the bitters.\n\nAntioxidant enzymes SOD and catalase reduces oxidative stress by scavenging oxygen radical and converting it to hydrogen peroxide and oxygen and further reduction of the hydrogen peroxide to minimize tissue damage by the radicals22. The alcoholic bitters treatment resulted in increased activity of these enzymes compared to the group administered ethanol. The observed increase in the activity of the antioxidant enzymes SOD and catalase in this study might be related to the upregulation of proteins (a biomolecule class to which SOD and CAT belongs) by the bitters stated earlier which will in turn lead to effective mop-of circulating free radicals and hence reducing oxidative stress and protecting the tissue from highly reactive hydroxyl radical22. This further corroborates the inherent antioxidant property of these bitters.\n\nIncreased level of sex hormones (testosterone, LH and FSH) in the alcoholic bitters administered animals is probably a further indication of the ability of the bitters to mediate improved sexual function and fertility23. Testosterone being the major male sex hormone will direct fortification of sexual prowess and function in animals. Furthermore, increased Luteinizing Hormone will likely contribute to the increased production of testosterone by the interstitial cells besides the other mechanism for its production via increased cholesterol synthesis24. While increased FSH observed from this study by the bitters will enhance spermatogenesis by the sertoli cells leading to increased production of viable and motile spermatozoa\n\nThe observed increased cholesterol levels and the activity of HMG-CoA reductase in the testes after administration of the alcoholic bitters may shed light on the mechanism by which the bitters improved the synthesis of testosterone in the animal. HMG-CoA reductase is a major enzyme in the synthetic pathway for cholesterol and its increased activity will also lead to production of increased levels of testicular cholesterol, which is a precursor for synthesis of steroid sex hormones18.\n\n\nConclusion\n\nThe acclaimed aphrodisiac and fertility enhancement property of herbal bitters is supported by this study as evidenced by its ability to combat oxidative stress and enhance synthesis of sex hormones through cholesterol upregulation in the testes.\n\n\nData availability\n\nF1000Research: Dataset 1. Data for assessment of various hormones in the serum and biochemical parameters in the testes of Wistar rats treated with: Group A (1) control group which received 0.2ml/kg body weight distilled water; Group B (2) received 0.2ml/kg body weight ethanol; Group C (3) received 0.2ml/kg body weight Alomo bitters; Group D (4) received 0.16ml/kg body weight Strikers bitters; Group E (5) received 0.3ml/kg body weight Orijin bitters, https://doi.org/10.5256/f1000research.16648.d22511725.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nMahat RK, Arora M, Bhale DV, et al.: Risk Factors and Causes of Male Infertility-A Review. Biochem Anal Biochem. 2016; 5: 271. Publisher Full Text\n\nNoumi E, Zollo PHA, Lontsi D: Aphrodisiac plants used in Cameroon. Fitoterapia. 1998; 69(2): 125–134. Reference Source\n\nD’Cruz SC, Vaithinathan S, Jubendradass R, et al.: Effects of plants and plant products on the testis. Asian J Androl. 2010; 12(4): 468–479. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKamatchouing P, Mbongue Fandio GY, Dimo T, et al.: Evaluation of androgenic activity of Zingiber officinale and Pentadiplandra brazzeana in male rats. Asian J Androl. 2002; 4(4): 299–301. PubMed Abstract\n\nAnionye JC: Department of Medical Biochemistry, College of Medical Sciences, University of Benin, Benin City, Nigeria. Bio-Research. 2011; 11.\n\nOgbonnia SO, Mbaka GO, Igbokwe NH, et al.: Antimicrobial evaluation, acute and subchronic toxicity studies of Leone bitters, a Nigeria prolyherbal formulation in rodent. Agric Biol JN Am. 2010; 1: 366–376. Reference Source\n\nFolch J, Lees M, Sloane-stanley GH: A simple method for the isolation and purification of total lipides from animal tissues. J Biol Chem. 1957; 226(1): 497–509. PubMed Abstract\n\nOmodeo Salè F, Marchesini S, Fishman PH, et al.: A sensitive enzymatic assay for determination of cholesterol in lipid extracts. Anal Biochem. 1984; 142(2): 347–350. PubMed Abstract \| Publisher Full Text\n\nTietz NW: Clinical Guide to Laboratory Tests. W.B. Saunders Company, Philadelphia. 1995; (3).\n\nGornall AC, Bardawill CJ, David MM: Determination of serum proteins by means of the biuret reaction. J Biol Chem. 1949; 177(2): 751–766. PubMed Abstract\n\nFriedewald WT, Levy RI, Fredrickson DS: Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem. 1972; 18(6): 499–502. PubMed Abstract\n\nJacobs NJ, Vandenmark PJ: Colorimetric method for determination of triglycsrides. Arch Biochem Biophys. 1960; 88: 250–55.\n\nRao GR, Ramakrishnan T, Sirsi M: Enzymes in Candida albicans. I. Pathways of glucose dissimilation. J Bacteriol. 1960; 80(5): 654–8. PubMed Abstract \| Free Full Text\n\nMisra HP, Fridovich I: The role of superoxide anion in the autoxidation of epinephrine and a simple assay for superoxide dismutase. J Biol Chem. 1972; 247(10): 3170–3175. PubMed Abstract\n\nBeers RF Jr, Sizer IW: A spectrophotometric method for measuring the breakdown of hydrogen peroxide by catalase. J Biol Chem. 1952; 195(1): 133–140. PubMed Abstract\n\nSatoh K: Serum lipid peroxide in cerebrovascular disorders determined by a new colorimetric method. Clin Chim Acta. 1978; 90(1): 37–43. PubMed Abstract \| Publisher Full Text\n\nAnionye JC, Onyeneke EC: Composition and in vitro antioxidant capacity of super bitters. Europ J Appl Sci. 2016; 8(5): 289–296. Reference Source\n\nNurudeen QO, Ajiboye TO: Aqueous root extract of Lecaniodiscus cupanioides restores the alterations in testicular parameters of sexually impaired male rats. Asian Pac J Reprod. 2012; 1(2): 120–124. Publisher Full Text\n\nde Zwart LL, Meerman JH, Commandeur JN, et al.: Biomarkers of free radical damage applications in experimental animals and in humans. Free Radic Biol Med. 1999; 26(1–2): 202–226. PubMed Abstract \| Publisher Full Text\n\nMatés JM: Effects of antioxidant enzymes in the molecular control of reactive oxygen species toxicology. Toxicology. 2000; 153(1–3): 83–104. PubMed Abstract \| Publisher Full Text\n\nOyewopo AO, Olaniyi SK, Oyewopo CI, et al.: Cell Phone Alters Oxidative Status and Impairs Testicular Function of Male Wistar Rats. J Reproductive Endocrinol Infert. 2017; 2(1): 22. Reference Source\n\nRagini V, Prasad KVSRG, Bharathi K: Antidiabetic and antioxidant activity of Shorea tumbuggaia Rox. International Journal of Innovative Pharmaceutical Research. 2011; 2(2): 113–121. Reference Source\n\nShah J: Erectile dysfunction through the ages. BJU Int. 2002; 90(4): 433–441. PubMed Abstract \| Publisher Full Text\n\nKayode OT, Yakubu MT: Parquetina nigrescens leaves: chemical profile and influence on the physical and biochemical indices of sexual activity of male Wistar rats. J Integr Med. 2017; 15(1): 64–76. PubMed Abstract \| Publisher Full Text\n\nKayode OT, Kayode AAA, Nwonuma CO: Dataset 1 in: Alcoholic bitters modulates sex hormones and some biochemical parameters of testicular function in male Wistar rats. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16648.d225117" }	[ { "id": "48915", "date": "03 Jun 2019", "name": "Gerhard Haidl", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is, in principle, an interesting article on the effect of alcoholic bitters on male rat sexual hormones and fertility. The authors found an increase in the hormonal concentrations of FSH, LH and testosterone and conclude, that sexual desire and fertility have increased and improved, respectively. In this context, the antioxidant effect may indeed contribute to a better testicular function. However, for interpretation of changes in fertility evaluation of seminal parameters or – in this case even better –testicular biology is necessary. From the clinical experience we know that FSH and LH as well as testosterone can rise after stimulating treatment with antiestrogens, but normally elevated levels of FSH and LH indicate a testicular damage. As alcohol is not only toxic to the liver but also to the testis it is well imaginable that fertility has not improved but deteriorated.\nTherefore, the conclusion that administration of alcoholic bitters may be helpful for fertility problems is than questionable. Without providing data on semen variables and testicular biopsy before and after treatment the manuscript should not be indexed.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] }, { "id": "56032", "date": "04 Dec 2019", "name": "Quadri O. Nurudeen", "expertise": [ "Reviewer Expertise Reproductive Biochemistry and Toxicology." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe conclusion that \"The acclaimed aphrodisiac and fertility enhancement property of herbal bitters is supported by this study as evidenced by its ability to combat oxidative stress and enhance synthesis of sex hormones through cholesterol upregulation in the testes\". The results in the study shows promising aphrodisiac potentials of the bitters, but not supporting the fertility enhancement property of the herbal bitters.\nHowever, the article described what the authors achieved accurately, and clearly stated the problem being investigated. There is sufficient information presented for replication of the research. The article is interestingly novel and may be recommended for indexing after the minor corrections.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/7-1838
https://f1000research.com/articles/7-1834/v1	21 Nov 18	{ "type": "Research Article", "title": "Behavioural and emotional issues among primary school pupils with congenital colour vision deficiency in the Federal Territory of Kuala Lumpur, Malaysia: A case-control study", "authors": [ "Belina Anne William M Thomas", "Sharanjeet Kaur", "Mohd Izzuddin Hairol", "Mahadir Ahmad", "Lei Hum Wee", "Belina Anne William M Thomas", "Mohd Izzuddin Hairol", "Mahadir Ahmad", "Lei Hum Wee" ], "abstract": "Background: Congenital colour vision deficiency (CCVD) is an untreatable disorder which has lifelong consequences. Increasing use of colours in schools has raised concern for pupils with CCVD. This case-control study was conducted to compare behavioural and emotional issues among age, gender and class-matched pupils with CCVD and normal colour vision (NCV). Methods: A total of 1732 pupils from 10 primary schools in the Federal Territory of Kuala Lumpur were screened, of which 46 pupils (45 males and 1 female) had CCVD. Mothers of male pupils with CCVD (n=44) and NCV (n=44) who gave consent were recruited to complete a self-administered parent report form, Child Behaviour Checklist for Ages 4-18 (CBCL/ 4-18) used to access behavioural and emotional problems. The CBCL/ 4-18 has three broad groupings: Internalising, Externalising and Total Behaviour Problems. Internalising Problems combines the Withdrawn, Somatic Complaints and Anxiety/ Depression sub constructs, while Externalising Problems combines the Delinquent and Aggressive Behaviour sub constructs. Results: Results from CBCL/ 4-18 showed that all pupils from both groups had scores within the normal range for all constructs. However, results from the statistical analysis for comparison, Mann-Whitney U test, showed that pupils with CCVD scored significantly higher for Externalising Problems (U=697.50, p=0.02) and Total Behaviour Problems (U=647.00, p= 0.01). Significantly higher scores were observed in Withdrawn (U=714.00, p=0.02), Thought Problems (U=438.50, p<0.001) and Aggressive Behaviour (U=738.00, p=0.04). Odds ratios, 95% CI, showed significant relative risk for high Total Behaviour Problem (OR:2.39 ,CI:1.0-5.7), Externalising Problems (OR:2.32, CI:1.0-5.5), Withdrawn (OR:2.67, CI:1.1-6.5), Thought Problems (OR:9.64, CI:3.6-26.1) and Aggressive Behaviour (OR:10.26, CI:3.4-31.0) scores among pupils with CCVD. Conclusion: Higher scores among CCVD pupils indicates that they present more behavioural and emotional problems compared to NCV pupils. Therefore, school vision screenings in Malaysia should also include colour vision to assist in the early clinical management of CCVD children.", "keywords": [ "behavioural problem", "emotional problem", "Child Behaviour Checklist/ ages 4-18", "colour vision", "congenital colour vision deficiency", "colour blind", "primary school pupil", "quantitative method" ], "content": "Introduction\n\nNormal colour vision (NCV), in which, all of the three types of retinal cones are functioning well, is a vital attribute of visual perception. The importance of NCV is observed even in the early years of a person life as colour plays a vital role in teaching, which helps to improve memory and increase pupil’s interest in the learning process (Bhoopal & Arya, 2016). However, the excessive use of colours in teaching and learning at primary schools may create confusion and a less favourable learning environment to pupils with congenital colour vision deficiency (CCVD), whose absorption spectra of cone pigments are defective.\n\nA cross-sectional study conducted by Reddy & Hassan (2006) among 1214 primary school pupils aged between seven and 12 years old in Petaling Jaya, Malaysia, showed an overall CCVD prevalence rate of 2.60% with a significant male predominance of 4.80% and 0.20% for female. This indicates that in every classroom of 40 pupils, one or two male pupils is/are expected to have CCVD. In Malaysia, school vision screenings are conducted at primary one, which is at the age of 7. However, they do not provide screening for colour vision. This may cause pupils with CCVD to be unaware of their condition. Hence, they may not be able to adapt well with their surroundings, which may lead to behavioural, emotional and social issues. This may subsequently result in a decline in overall academic performance (Espinda, 1971) as well as having a negative impact on the individual’s self-confidence (Bailey, 2013).\n\nThe main objective of this case-control study was to compare behavioural and emotional issues among primary school pupils with CCVD (as case) and age, gender and class-matched pupils with normal colour vision (NCV) (as controls) in the Federal Territory of Kuala Lumpur, Malaysia using the Malay language (Bahasa Melayu) adapted version of Child Behaviour Checklist/ages 4-18 (CBCL/ 4-18). The CBCL/ 4-18 is used to access behavioural and emotional problems (also known as behavioural syndromes). It has three broad groupings of syndromes scale: Internalising Problems, Externalising Problems and Total Problems. Internalising Problems combines the sub constructs of Withdrawn, Somatic Complaints, and Anxiety/Depression sub constructs, while Externalising problems combines the Delinquent Behaviour and Aggressive Behaviour sub constructs (Achenbach, 1991). The Total Problems combines the scores of all the sub constructs (Brown & Achenbach, 1992). The Child Behaviour Checklist/ages 4-18 (CBCL/ 4-18) has been translated and validated into 75 languages, and it is widely used for clinical diagnosis and in research (Ivanova et al., 2010). To date this questionnaire has not been used to study the behavioural and emotional problems among pupils with CCVD. The CBCL/ 4-18 has been widely used previously in various studies among children in Malaysia, and it was found that this checklist was a good screening tool for the maladjusted (Normah & Shalisah, 1999; Talib et al., 2011; Teoh, 2010). Studying the significance of this issue is important to assist in the early clinical management such as adaptive strategies and early counselling for the children with CCVD.\n\n\nMethods\n\nThe study is reported in accordance with the STROBE case-control reporting guidelines (von Elm et al., 2007).\n\nA case-control research survey using a parent-completed questionnaire was used to carry out this study.\n\nTen national primary schools were selected by means of simple random sampling method from a list primary schools in four districts in the Federal Territory of Kuala Lumpur, Malaysia which are Kuala Lumpur (n=4), Sentul (n=2), Cheras (n=2) and Sungai Besi (n=2).\n\nIn the beginning, we did not specify which parent should complete the CBCL/ 4-18, but the returned consent form were mostly fill out by mother of the pupils (n=1655, 95.6%). Thus, based on majority, we decided to recruit only mother of pupils as respondents in this study. A purposive sampling method was used whereby the participants were recruited based on particular consideration that is, the mother of pupils with CCVD, and age, gender and class-matched pupils with NCV (Etikan, 2016). Inclusion and exclusion criteria in this study includes voluntary participation from mothers of pupils with CCVD and NCV (control group), pupils are between the age of eight and 11 years, and should have no other vision, physical, or cognitive disability.\n\nThe required sample size of pupil with CCVD in this study was calculated based on the prevalence rate of CCVD among the population of primary school pupil in Petaling Jaya, Selangor which is at 2.60% (Reddy & Hassan, 2006). Hence, the required sample size was calculated using Morgan’s simple sample size calculation formula (Daniel, 1999).\n\nn=Ζ2P(1-P)d2 n= 1.962(0.026)(1-0.026)0.052n=38.91n≈40\n\nWhere, Z is the standard value for level of confidence at 95.00% (1.96); P is the prevalence rate of primary school pupil with CCVD in Petaling Jaya, Malaysia at 2.60% and d is the margin of error set at 5.00%.\n\nThe required sample size was 40 primary school pupils from each group. Taking into consideration of dropouts, an additional 10% (4 pupils) was added to the sample size calculation. Thus, making the actual sample size to 44 pupils with CCVD. A total of 1732 (male=849 (49.0%), female=883 (51.0%) primary school pupils underwent the colour vision screening. The screening revealed a total of 46 (2.7%) pupils had CCVD. The prevalence rate of CCVD is higher in males (5.3%, n=45) than females (0.1%, n=1). The participants of this study consisted of mother of the pupils with CCVD (case group) and NCV as a control group. Based on the sample size calculation, only 44 mothers of pupils with CCVD (all male) who gave consent and agreed to participate in this study were recruited. The control group was equal to the size of the cases (case:control ratio of 1:1). Therefore, the control group included 44 mothers of pupils with NCV, where all the pupils that were age, gender and class-matched were selected by means of purposive sampling based their presence at the study location. All the pupils were given written consent to obtain their mother’s agreement to be recruited in this study.\n\nThe questionnaire used in this study was the adapted Malay language version of the Child Behaviour Checklist/ages 4-18 (CBCL/ 4-18) completed by the mothers of pupils with CCVD and NCV to illustrate their children’s behavioural and emotional issues (Achenbach & Rescorla, 2001) (Questionnaire is available from the Achenbach System of Empirically Based Assessment (ASEBA) website). This questionnaire comprised of 113 questions measuring eight sub constructs: Withdrawn, Somatic Complaints, Anxiety/Depression, Social Problems, Thought Problems, Attention Problems, Delinquent Behaviour and Aggressive Behaviour (Achenbach, 1991). Respondents read and gave their views using a three-point Likert scale: scale of '0' indicated 'not true', the '1' scale indicated 'sometimes true’ and the ‘2’ scale indicated ‘very true’. This version, designed for 4 to 18 year old children, had clear instructions, and the mother of the pupils could complete it within 15 to 20 minutes without the need for supervision or guidance from the researcher.\n\nConsent forms and information sheets were given to all pupils to obtain their parent’s consent to allow them to participate in this study. Only those with their parent’s consent were recruited. Participant’s parents were informed that all collected data would remain confidential.\n\nFirstly, visual acuity and colour vision screening were conducted by the researcher at the selected schools from February 2018 till May 2018 to identify pupils with CCVD. Pupils who passed the inclusion criteria, underwent the visual acuity measurement which was conducted at both distant (6 metres) and near (40 cm) using the Early Treatment Diabetic Retinopathy Study (EDTRS) chart and colour vision screening using Ishihara 24-Pseudoisochromatic Plates (Kanehara Trading Co. Ltd, Tokyo, Japan) and followed by Farnsworth D-15 test (Munsell Color Company, Inc., Baltimore, MD, USA). The Ishihara 24-plates were performed by holding them under daylight at a distance of 50 cm and tilted so that each plate was at right angles to the line of vision. The time allocated to read each plate was less than 5 seconds. Those who failed to read four or more plates, were then asked to arrange 15 coloured caps on the Farnsworth D-15 test. To perform this test, pupils were instructed to arrange 15 randomly ordered coloured caps in order of hue with a reference coloured cap placed at the starting point for them to arrange the rest of the caps in an order. When the pupil had completed the test, the cap sequence is plotted and based on the number and direction of major crossovers on the plot, the type of colour vision deficiency were determined.\n\nThen, pupils who were identified having CCVD and an equal number of age, gender and class-matched pupils with NCV were required to obtain consent of their mother to be respondents in this study. Only mothers of pupil who agreed were recruited and given the self-administered CBCL/ 4-18 questionnaire to be completed to illustrate any issues in their child’s behaviour and emotional state. Demographic data of all the pupils, which included age and type of colour vision deficiency (for pupil with CCVD), as well as their mothers’ which included age, race, education level, family income (monthly), marital status, medical and ocular history of the child, and awareness of their child’s colour vision problem were also recorded. A flow diagram of the participant recruitment procedure is as shown in Figure 1.\n\nThe CBCL/ 4-18 template for hand-scoring was used to transfer of data from the questionnaire forms completed by the respondents to the hand-scored problem scales profile for the pupil. The raw scores for each sub construct are converted to age-standardized T-scores with the aid of the template to statistically analyse the data (T-score, µ = 50 and σ = 10). The quantitative data collected was analysed using the IBM SPSS Statistical 22.0 software.\n\nThe data was analysed for normality using the Shapiro-Wilks test, and the Descriptive Statistics test was carried out to measure the behaviour and emotional of the pupil by looking at the skewness of the graph, whether positive or negative. The tests revealed that the data was not normally distributed. Thereafter, the non-parametric statistical analysis, Mann-Whitney U test, was used to compare and determine whether there was a significant difference in behaviour and emotions between pupils with CCVD (case) and with NCV (control). Odds ratios were also calculated with 95% confidence intervals (CI) as estimates of the relative risk for high symptom scores among children with CCVD compared with the control group (Liljenfeldt & Pettersson, 2017). A two-tailed p value below 0.05 was considered statistically significant.\n\n\nResults\n\nThe demographic data of mothers and their children with CCVD and NCV who agreed to be recruited as respondents in this study are shown in Table 1. The majority (86.36% of mothers of pupils with CCVD, and 72.72% of mother of pupils with NCV), were within 31 to 50 years old and mostly are married. About 45.45% of mother of pupils with CCVD and 38.64% of mother of pupils with NCV had undergraduate degrees, and most of them had a monthly family income of RM (Malaysian ringgit) 4,001-RM 6,000. Only 15.91% of mothers of pupils with CCVD stated that they were aware of their child’s condition. All aged-matched pupil in both the CCVD group and NCV group were within the age group of 8 to 11 years old (µ =9.47, σ =1.04). Among those with CCVD, 33 (75.00%) were identified as deuteranomalous trichromats and 11 (25.00%) as protanomalous trichromats. All mother of pupils with CCVD and NCV ruled out any known medical and ocular history of the child.\n\nCCVD – congenital colour vision deficiency, NCV – normal colour vision\n\nA normality test was conducted on each sub construct to determine the distribution of data. The Shapiro-Wilk, W, test for all eight sub constructs and combination of sub constructs, revealed a p-value <0.05 which shows that all data were not normally distributed. Descriptive statistical analysis results for the CBCL/ 4-18 questionnaire are presented based on raw score which had been converted into standard scores (T-score, µ =50 and σ =10). The conversion to T-score allows the comparison of scores obtained with normative data from other pupils of the same age range as shown in Table 2 (Achenbach, 1991). Based on the frequency analysis, the scores for all eight sub constructs and the three broad groupings of behavioural syndrome were within the normal range of both groups as shown in Figure 2.\n\nHorizontal broken lines = Borderline clinical range.\n\nAs the data were not normally distributed, non-parametric statistical analysis, Mann-Whitney U test, was chosen to compare the results for both groups. As illustrated in Table 3, significantly higher scores were observed for the CCVD group in Externalising Problems (U=697.50, p=0.02) and Total Behaviour Problems (U=647.00, p=0.01). Similarly, significantly higher scores for the CCVD group were also observed in the sub constructs Withdrawn (U=714.00, p=0.02), Thought Problems (U=438.50, p< 0.001) and Aggressive Behaviour (U=738.00, p=0.04).\n\nCCVD – congenital colour vision deficiency, NCV – normal colour vision\n\nSubsequently, odds ratios were calculated with 95% confidence intervals (CI) as estimates of the relative risk for high Total Behaviour Problem, Externalising Problems, Withdrawn, Thought Problems and Aggressive Behaviour scores among children with CCVD compared with the control group. When a cut-off was applied at the T-score ≥60, which indicates borderline clinical behavioural syndrome, odds for high scores on the Total Problems scale was 2.39(CI 1.0-5.7), Externalising Problems was 2.32 (CI 1.0-5.5), Withdrawn was 2.67(CI 1.1-6.5), Thought Problems 9.64 (CI 3.6-26.1) and Aggressive Behaviour was 10.26 (CI 3.4-31.0) as shown in Table 4.\n\nNCV – normal colour vision\n\n\nDiscussion\n\nThis study compares behavioural and emotional issues among primary school pupils with CCVD and NCV. The results from this study suggest that pupils with CCVD presented more behavioural and emotional problems as compared to NCV pupils. It is found that CCVD pupils might be at a higher risk of developing social and attention problems. Pupils with CCVD having high scores for sub construct of Withdrawn tend to present with behaviours such as preference to be alone, shy, staring blankly and show signs of sadness. Socially withdrawn pupil aged above 7 years, often encounter problems in social interactions with peers and social skills (Fink et al., 2015). Moreover, with increasing age social withdrawal becomes accompanied by feelings of loneliness and depression (Matthews et al., 2015). This may contribute to acting-out behavior in form of social aggression that are associated with behaviours such as arguing, screaming, showing off, attention-seeking, bragging, teasing, being demanding, threatening behaviour and being temperamental. Furthermore, pupils’ class participation and social skills is an important contributor to their academic competence (Rabiner et al., 2016). Additionally, pupils with CCVD having a combination of social withdrawal, aggressive behavior, along with thought problems with characteristics of obsessing on certain thoughts, finding it difficult to concentrate, staring blanking and having strange ideas or behaviours, can diminish the ability to learn, which affects academic performance (Rhoades et al., 2011).\n\nOur finding revealed that pupils with CCVD are experiencing more internalising problems. There is a possibility that in general, people are better at recognising pupil’s externalising problems, such as aggression, internalising problems such as depression and social withdrawal. This is because externalising problems may be more likely to induce a sense of worry in the people surrounding them, while the internalising problems faced by these pupils may be overlooked by their parents and teachers who are unaware of the pupil’s colour vision impairment. This problem can be overcome by an early colour vision screening. For screening, the Ishihara test is widely used as it is quick and easy to administer, inexpensive, and has a high validity (Birch, 1997). Though, in a recent published review article on ‘Is screening for congenital colour vision deficiency in school students worthwhile?’, Ramachandran & colleagues (2014) stated that there’s minimal evidence to support the screening for CCVD in school. However, this article has received disagreement from other researchers. Based on personal experience with congenital colour vision defects, Cole (2015) agree to disagree with the conclusion made by the writer. He believes that pupils with CCVD do need to know about their condition before the end of their schooling. This is also supported by Long & colleagues (2015) as they believe that delaying the diagnosis and awareness of CCVD may create significant emotional and psychological impact. Based on their clinical perspective of working with young adults having CCVD, they see many of them who comes for a comprehensive colour vision examination for job or tertiary education recruitment, often gets shocked which accompanied with grief, disbelief and anger upon discovering their condition. Thus, an early colour vision screening is school would be the best option. This will enable useful early counselling and adaptive strategies to be implemented especially in classrooms as early as possible.\n\nClassroom behaviour is very important in primary school life and pupils who display problematic behaviours also tend to have deficits in social and emotional skills (Wagner & Ruch, 2015). Besides that, behavioural problems present in early childhood may develop into greater problems in later life (Ogundele, 2018). It is therefore important to broaden the age range of the current study in order to take into account mental health problems across the different stages of childhood development such as preschool and secondary school. Research has found generally pupils’ behavioural problems differ according to gender. Previous studies conducted among pupils from various cultures have found that male pupils are more likely to achieve higher scores on Externalising Problems as compared to Internalising Problems (Achenbach et al., 1990; Achenbach et al., 1990).\n\nA limitation of this study was having no data from female pupils with CCVD due to the very low prevalence rate. Thus, in the present study, CBCL/ 4-18 data was only collected from age and class-matched male pupils in both case and control groups for comparison. Therefore, it is not possible to compare scores between male and female pupils. However, this study showed that male pupils with CCVD had significantly higher scores as compared to the male pupils in the control group with NCV. Further understanding of the behavioural problems among female pupils with CCVD is recommended for future studies. Our findings should also be viewed in the context of some methodological limitations. Because this study was conducted in national primary schools in Malaysia, most pupils with CCVD were mainly Malay with only one Chinese and one Indian. Thus, our findings may not generalise to minorities. Future studies are to be conducted in vernacular schools.\n\n\nConclusion\n\nIn conclusion, the analysis of the CBCL/ 4-18 showed that the scores for all problem sub constructs obtained by pupils with CCVD were within the normal range. However, their scores were higher than of their peers with NCV, which suggest that the pupils with CCVD present more behavioural and emotional problems as compared to NCV pupils. These findings provide important new data on the behavioural and emotional problems of Malaysian primary school pupils with CCVD. This study emphasises the importance of additional studies to be conducted to understand this issue in depth which provides insight to assist in the clinical management of the CCVD children. Thus, early school visual screening in Malaysia should also include colour vision so that the child, their family, and school teachers are aware of their condition as early as possible to ensure the well-being of the child.\n\n\nData availability\n\nHarvard Dataverse: Dataset 1. Demographic Data & CBCL/4-18 Scores-Behavioural and Emotional Issues among Primary School Pupils with Congenital Colour Vision Deficiency in Federal Territory of Kuala Lumpur, Malaysia. https://doi.org/10.7910/DVN/DPZHI4 (Thomas et al., 2018)\n\n\nEthics approval\n\nEthical approval for this study was obtained from the Research and Ethics Committee of Universiti Kebangsaan Malaysia (UKMREC Approval Number: UKM PPI/111/8/JEP-2016-326).\n\n\nConsent\n\nWritten informed consent for publication of the participants details were obtained from the participants.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nThe authors would like to thank the Director of Ministry of Education Malaysia and Kuala Lumpur Federal Territory Education Department for granting us the approval to conduct the study among primary school pupil as well as to the Headmasters of the primary schools around the Federal Territory of Kuala Lumpur, Malaysia who have fully cooperated in the data collection process.\n\n\nReferences\n\nAchenbach TM: Integrative guide for the 1991 CBCL/4-18, YSR and TRF Profiles. 1st ed. Burlington, VT: Department of Psychiatry, University of Vermont. 1991. Reference Source\n\nAchenbach TM, Bird HR, Canino G, et al.: Epidemiological comparisons of Puerto Rican and U.S. mainland children: parent, teacher, and self-reports. J Am Acad Child Adolesc Psychiatry. 1990; 29(1): 84–93. PubMed Abstract \| Publisher Full Text\n\nAchenbach TM, Hensley VR, Phares V, et al.: Problems and competencies reported by parents of Australian and American children. J Child Psychol Psychiatry. 1990; 31(2): 265–286. PubMed Abstract \| Publisher Full Text\n\nAchenbach TM, Rescorla LA: Manual for the ASEBA School-Age Forms & Profiles. Burlington, VT: Research Center for Pupil, Youth, & Families, University of Vermont, 2001. Reference Source\n\nBailey J: Color Blindness: Psychological Effects. [Accessed: 14 April 2018], 2013. Reference Source\n\nBhoopal C, Arya S: Significance of Colour Usage in Cognitive Mind Maps to Enhance Academic Achievement. Int J Indian Psychol.. 2016; 3(4): 184–189. Reference Source\n\nBirch J: Efficiency of the Ishihara test for identifying red-green colour deficiency. Ophthalmic Physiol Opt. 1997; 17(5): 403–408. PubMed Abstract\n\nBrown JS, Achenbach TM: Bibliography of published studies using the Child Behaviour Checklist and related materials. 1992nd ed. Burlington, VT: Department of Psychiatry, University of Vermont. 1992. Reference Source\n\nCole BL: Re.: Is screening for congenital colour vision deficiency in school students worthwhile? Clin Exp Optom. 2015; 98(2): 193. PubMed Abstract \| Publisher Full Text\n\nDaniel WW: Determination of sample size for estimating proportions. In: Biostatistics: A Foundation for Analysis in Health. 10th ed. New York: John Wiley & Sons, Inc, 1999; 183.\n\nEspinda SD: Color vision deficiency in third and sixth grade boys in association to academic achievement and descriptive behavioral patterns.Diss Abstr Int.1971; 32. Reference Source\n\nEtikan I: Comparison of convenience sampling and purposive sampling. American Journal of Theoretical and Applied Statistics. 2016; 5(1): 1–4. Publisher Full Text\n\nFink E, Begeer S, Peterson CC, et al.: Friendlessness and theory of mind: a prospective longitudinal study. Br J Dev Psychol. 2015; 33(1): 1–17. PubMed Abstract \| Publisher Full Text\n\nIvanova MY, Achenbach TM, Rescorla LA, et al.: Preschool psychopathology reported by parents in 23 societies: testing the seven-syndrome model of the child behavior checklist for ages 1.5-5. J Am Acad Child Adolesc Psychiatry. 2010; 49(12): 1215–1224. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiljenfeldt J, Pettersson Ö: Distributional justice in Swedish wind power development-An odds ratio analysis of windmill localization and local residents’ socio-economic characteristics. Energy Policy. 2017; 105: 648–657. Publisher Full Text\n\nLong JA, Honson V, Katalinic P, et al.: Re.: Is screening for congenital colour vision deficiency in school students worthwhile? A review. Clin Exp Optom. 2015; 98(2): 192. PubMed Abstract \| Publisher Full Text\n\nMatthews T, Danese A, Wertz J, et al.: Social isolation and mental health at primary and secondary school entry: a longitudinal cohort study. J Am Acad Child Adolesc Psychiatry. 2015; 54(3): 225–232. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNormah CD, Shalisah S: Behavior profiles of specific learning disabled and school children. Malaysian Journal of Psychiatry. 1999; 7(2): 35–44.\n\nOgundele MO: Behavioural and emotional disorders in childhood: A brief overview for paediatricians. World J Clin Pediatr. 2018; 7(1): 9–26. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRabiner DL, Godwin J, Dodge KA: Predicting academic achievement and attainment: the contribution of early academic skills, attention difficulties, and social competence. School Psych Rev. 2016; 45(2): 250–267. Reference Source\n\nRamachandran N, Wilson GA, Wilson N: Is screening for congenital colour vision deficiency in school students worthwhile? A review. Clin Exp Optom. 2014; 97(6): 499–506. PubMed Abstract \| Publisher Full Text\n\nReddy SC, Hassan M: Refractive errors and other eye diseases in primary school children in Petaling Jaya, Malaysia. Asian Journal of Ophthalmology. 2006; 8(5): 195–198. Reference Source\n\nRhoades BL, Warren HK, Domitrovich CE, et al.: Examining the link between preschool social–emotional competence and first grade academic achievement: The role of attention skills. Early Child Res Q. 2011; 26(2): 182–191. Publisher Full Text\n\nTalib MBA, Abdullah R, Mansor M, et al.: Relationship between parenting style and children’s behavior problems. Asian social science. 2011; 7(12): 195–200. Publisher Full Text\n\nTeoh HJ: A survey of urban child and adolescent mental health problems in an urban Malaysian population. Malaysian Journal of Psychiatry. 2010; 19(1): 1–13. Reference Source\n\nThomas BA, Kaur S, Hairol MI, et al.: Dataset 1 Demographic Data & CBCL/4-18 Scores in: Behavioural and Emotional Issues among Primary School Pupils with Congenital Colour Vision Deficiency in Federal Territory of Kuala Lumpur, Malaysia. [Accessed: 7 November 2018]; 2018. http://www.doi.org/10.7910/DVN/DPZHI4\n\nvon Elm E, Altman DG, Egger M, et al.: The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Epidemiology. 2007; 18(6): 800–804. PubMed Abstract \| Publisher Full Text\n\nWagner L, Ruch W: Good character at school: positive classroom behavior mediates the link between character strengths and school achievement. Front Psychol. 2015; 6: 610. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "40971", "date": "07 Jan 2019", "name": "Khairidzan Mohd Kamal", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a good article and research work. It addresses the question of whether congenital colour vision deficiency (CCVD) affects the behavioural and emotional status of school children at the early ages. This is an especially important question to explore given that the current teaching and learning methods in primary schools utilize significant animated and graphically based materials. The researchers have chosen a comprehensive screening method which involved only mainstream education systems (public schools) and adopted standard questionnaires with translation (CBCL/4-18) to be given to parents. The authors have employed appropriate statistical tools for normality and non-parametric analysis. Matching processes between affected children with CCVD with control groups were well performed which allowed the researchers to analyse their background and economic status. For example, education background and household income for each group were compared to show the possibility of these two factors in the children’s behavioural and emotional status. However, since the recruitment was done in schools around metropolitan areas with standardized teaching facilities, it would be difficult to conclude whether this result is applicable to school children with CCVD in the rural area.\n\nThe discussions and conclusions highlighted that children with CCVD were within the normal range but showed significantly higher scores than children without CCVD, Externalising and Total Behavioural Problems. It is more important to note that students in this study came from public school systems with household income that represents borderline urban poverty line. Low awareness rate among parents of both affected and non-affected children is alarming and reasons for this should be highlighted in the article. It would be interesting to find out whether children from affluent backgrounds would have the same results as revealed in this research work.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "43676", "date": "06 Feb 2019", "name": "Stephen J. Dain", "expertise": [ "Reviewer Expertise Colour vision", "clinical examination of colour vision", "occupational consequences of colour vision deficiencies", "colour vision and disease" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an interesting report in a little studied area. It would benefit from a little wider literature review, better characterisation of the CCVD subjects and attention to the multiple tests nature of the analysis.\nIntroduction:\nParagraph 1, Last line: Absorption spectra are not really “defective”. They are altered.\n\nSee Bacon (19711). It would also be useful to show that you have read Steward and Cole (19892) and Cole et al., (20023).\n\nParagraph 2, Line 10: And their parents.\n\nMethods:\nRecruitment, Line 8: Were the pupils also matched for school (as a proxy for socio-economic matching)? This may also be addressed in Procedure paragraph 3 where you say that they were “class-matched”. Is that also implying school matched?\n\nProcedure: Was there an inclusion/exclusion criterion for VA?\n\nProcedure, paragraph 2: Was the pass/fail (4 errors) applied to all plates, screening and diagnostic, numerical and tracing? In reality it should only be the screening plates. Daylight – artificial or natural? If artificial, details of source, correlated colour temperature and general colour rendering index. If natural, from which direction? Not that this matters as much at the latitude of KL. Illuminance level?\n\nResults:\nWas any correction for multiple comparisons made?\n\nHow did you decide if they were deuteranomalous or protanomalous? Strange that there were no dichromats. Or was it that you did not use an anomaloscope and could not actually identify dichromats? In which case you should talk deutans and protans. See also Table 2. Its a pity. Relating the problems to the degree of deficiency would be very informative. Do the dichromats, and strongly anomalous, who generally know, cope better than those who do not or worse because they are more affected?\n\nDiscussion:\nLast paragraph: I don’t see the lack of a female data as a limitation. In colour vision it is a fact of life. Also the female CCVD are predominantly deutan and less affected (manifesting the lesser of the two deficiencies that they carry).\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] }, { "id": "42106", "date": "08 Feb 2019", "name": "Wan-Hazabbah Wan Hitam", "expertise": [ "Reviewer Expertise General Ophthalmology", "Neuro-ophthalmology and Visual Electrophysiology" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nOverall, this is a good study. However, the study only involved mothers of the children to answer the questionnaires. The researchers should also consider interviewing the teachers because nowadays the students spend more of their time in school compared to home. In certain circumstances the mothers may not know what actually happened at school. The teachers may know more of the achievements and problems of the children.\nTitle - Clearly represents the study. Abstract - Satisfactory. Introduction - Good. Able to highlight the importance of the study. Objective - Good. Clearly stated. Methodology - Very good. Clearly elaborate. Results - Well presented. Discussion - Good coverage. Limitation - Well addressed. Conclusion - Good. Answered the objective of the study. References - Good. Up to date. Tables & Figures - Clearly stated results.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1834
https://f1000research.com/articles/7-1576/v1	28 Sep 18	{ "type": "Software Tool Article", "title": "gganatogram: An R package for modular visualisation of anatograms and tissues based on ggplot2", "authors": [ "Jesper L.V. Maag" ], "abstract": "Displaying data onto anatomical structures is a convenient technique to quickly observe tissue related information. However, drawing tissues is a complex task that requires both expertise in anatomy and the arts. While web based applications exist for displaying gene expression on anatograms, other non-genetic disciplines lack similar tools. Moreover, web based tools often lack the modularity associated with packages in programming languages, such as R. Here I present gganatogram, an R package used to plot modular species anatograms based on a combination of the graphical grammar of ggplot2 and the publicly available anatograms from the Expression Atlas. This combination allows for quick and easy, modular, and reproducible generation of anatograms. Using only one command and a data frame with tissue name, group, colour, and value, this tool enables the user to visualise specific human and mouse tissues with desired colours, grouped by a variable, or displaying a desired value, such as gene-expression, pharmacokinetics, or bacterial load across selected tissues. I hope that this tool will be useful by the wider community in biological sciences. Community members are welcome to submit additional anatograms, which can be incorporated into the package. A stable version gganatogram has been deposited to neuroconductor, and a development version can be found on github/jespermaag/gganatogram.", "keywords": [ "Anatograms", "Anatomy", "Tissues", "Organs", "ggplot2", "R", "Expression Atlas" ], "content": "Introduction\n\nEfficiently displaying tissue information in multicellular organisms can be a laborious and time consuming process. Often researchers want to showcase differences in values, such as gene expression or pharmacokinetics between tissues in one organism, or between similar tissues in different groups.\n\nWhereas bar charts and heatmaps provide an informative view of the differences between groups, it can be difficult to immediately observe the biological significance (Figure 1a–b). As compared to an anatogram, where it is easy to quickly spot the differences between tissues or groups, and immediately provide biological context to these observations (Figure 1c). This also has the added benefit that the audience, whether reading a paper or attending a lecture, will have to spend less time and effort to grasp the results.\n\nThe values in the graphs are the same.\n\nSeveral online tools to display gene expression in different tissues already exist1–4. Although these tools provide important information regarding gene expression in various tissues and organisms, other disciplines besides genetics are unable to utilise these applications due to the focus on genes. Moreover, these tools often only include a predefined set of experiments that can be visualised, leading to difficulties in presenting your own data. Other caveats with these tools are that it can be laborious to recreate the plot or automatically create plots from results.\n\nHere I present gganatogram, an open source R package based on ggplot25 utilising the publicly available mouse and human anatograms from the Expression Atlas1,2. With this package it is easy for any R user to quickly visualise anatograms with specified colours, groups, and values. Using the familiar grammar from ggplot25, this program allows for modular anatograms to be generated.\n\n\nMethods\n\ngganatogram is stored on neuroconductor6, an open-source platform for rapid testing and dissemination of reproducible computational imaging software. A development version can be found on github/jespermaag/gganatogram, which allows for the community to post issues with the package, submit requests, or add anatograms by creating coordinate files.\n\n\n\nThe development version can be installed from github:\n\n\n\nBriefly, to generate the main list objects that contain all tissue coordinates, I downloaded SVG files from the Expression Atlas (Available from gganatogram GitHub page2) and processed them using a custom python script (available from GitHub). The script scraped through the SVG files to extract the name, coordinates, and SVG transformations. These were then post-processed in R to create the rda files that make up the tissue coordinates.\n\ngganatogram requires an installation of R≥3.0.0, ggplot25 v.3.0.0 and ggpolypath7 v.0.1.0. The program should be able to run on any computer with the system requirements for R. Plots can be generated using a basic data.frame containing organ name, colour, type, or value, with the specified column names below. Organs are plotted one at a time based on the order of the data.frame. The tissue of each consecutive row will be layered on top of the previous. The gganatogram package provides four such data.frames containing all tissues available to plot, one for each human and mouse, and divided by sex.\n\n\n\nThese data frames have already specified colour, type, and an assigned random number to facilitate the start of plotting.\n\n\n\nThe main function is called gganatogram(). By default, and without any arguments, it plots the outline of a male human with standard ggplot2 parameters. By adding just a few options, it is possible to quickly change to female, fill specified organs by selected colour, or fill the organs based on a value (Figure 2).\n\n(A) Default plot generated by calling gganatogram(), (B) adding female, plotting specified organs by (C) colour, (D) value.\n\n\n\n\nUse cases\n\nThis section provides additional plotting examples.\n\nTo plot all tissues per organism, use the provided key files that exist per organism and sex. This displays all tissues in the order of each data frame. To change the order in which organs are layered on top of each other, reorder the data frame to have those tissues at the bottom (Figure 3).\n\nThe colours are specified in the provided key data frames.\n\n\n\n\n\nTo compare anatograms, e.g. draw one specific anatogram side by side and compare values, a long table has to be created with the type column changed to the variables to compare. The following code recreates (Figure 1c).\n\n\n\nOrgans can also be separated by faceting, as per standard ggplot2 using facet_wrap (Figure 4). This can help to display organs that are nested on top of each other.\n\n\n\nBecause I elected to use ggplot25 for the package, the user can add additional layers from standard plots. This can be useful to show highlight features, such as metastasis, location of tissue biopsies, or gene expression of specific biopsies (Figure 5).\n\nAnother option is to fill both tissues and points by value (top right). Red colour around plot added for emphasis.\n\n\n\n\nSummary\n\nIn summary, I have designed and implemented an R package to easily visualise anatograms based on ggplot25 and the anatograms from Expression Atlas2, which when combined create a powerful tool to plot and display tissue information.\n\nThe one line command to generate these plots should allow for users with even limited R knowledge to create informative anatograms for publications or presentations.\n\n\nData availability\n\nSVG files are available from GitHub: https://github.com/ebi-gene-expression-group/anatomogram/tree/master/src/svg\n\n\nSoftware availability\n\n1. Link to version control repository containing the source code:\n\nhttp://neuroconductor.org/package/gganatogram\n\n2. Link to development version:\n\nhttps://github.com/jespermaag/gganatogram\n\n3. Link to archived source code as at time of publication:\n\nhttp://doi.org/10.5281/zenodo.14342338\n\nSoftware license: GPL-2", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgments\n\nI would like to thank the neuroconductor team: Ciprian Crainiceanu, John Muschelli, Brian Caffo, and Adi Gherman for storing gganatogram on their repository.\n\nPaul Brennan (@brennanpcardiff) for adding additional checks to the package.\n\nI would like to thank Irene Papatheodorou and the Expression Atlas team at EMBL-EBI for making the anatograms available.\n\nI would also like to thank Anna Antoniak for editing the manuscript, and Stephen Rudley for manuscript feedback.\n\n\nReferences\n\nPapatheodorou I, Fonseca NA, Keays M, et al.: Expression Atlas: gene and protein expression across multiple studies and organisms. Nucleic Acids Res. 2018; 46(D1): D246–D251. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPetryszak R, Keays M, Tang YA, et al.: Expression Atlas update--an integrated database of gene and protein expression in humans, animals and plants. Nucleic Acids Res. 2016; 44(D1): D746–D752. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLekschas F, Stachelscheid H, Seltmann S, et al.: Semantic Body Browser: graphical exploration of an organism and spatially resolved expression data visualization. Bioinformatics. 2015; 31(5): 794–796. PubMed Abstract \| Publisher Full Text\n\nPalasca O, Santos A, Stolte C, et al.: TISSUES 2.0: an integrative web resource on mammalian tissue expression. Database (Oxford). 2018; 2018. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWickham H: ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York, 2016; ISBN 978-3-319-24277-4. Publisher Full Text\n\nMuschelli J, Gherman A, Fortin JP, et al.: Neuroconductor: an R platform for medical imaging analysis. Biostatistics. 2018. PubMed Abstract \| Publisher Full Text\n\nSumner MD: ggpolypath: Polygons with Holes for the Grammar of Graphics. Reference Source\n\nMaag J, Muschelli J, Brennan P, et al.: jespermaag/gganatogram: First release (Version V1.0.0). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1434233" }	[ { "id": "39255", "date": "22 Oct 2018", "name": "Helder I. Nakaya", "expertise": [ "Reviewer Expertise Systems biology", "computational tools", "vaccinology" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe author describes an R package that displays discrete and continuous data onto anatomical structures. The structures are based on mouse and human anatograms from the Expression Atlas project and the grammar from ggplot2 R library. The code example was easy to run and the necessary input data was intuitive and simple.\n\nThe author can increase its usage by providing a webtool (even one based on shiny) that takes as input a csv or tsv table with predefined columns. This would allow physicians and scientists with no background in bioinformatics to easily display their data.\n\nAlso, it would be useful to create an anatogram for the human brain and one for the different compartments of a cell (nucleus, mitochondria, etc).\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [] }, { "id": "39943", "date": "30 Oct 2018", "name": "Saskia Freytag", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article describes a new R package, which allows easy plotting of discrete and continuous measurements onto human and mouse anatomy.\nThis is a really valuable R package contribution, as it fills a real void in the current infrastructure. I found the code examples in the manuscript intuitive and easy to run. It is great that the author adopts the popular ggplot2 grammar as well as a tidy data structures.\nMinor comments:\nIt would be useful to know how many tissues (and which) can be plotted using this package.\n\nAn example of changing the order of the data frame to change the layering should be added.\n\nI encountered the following error when trying to install the package through neuroconductor: Error in latest_neuroc_release(release = \"stable\") : unused argument (release = \"stable\")\nThank you for making all code (even for processing) publicly available.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1576
https://f1000research.com/articles/7-1819/v1	20 Nov 18	{ "type": "Research Article", "title": "Inflammatory cell expression of Toll-like receptor-2 (TLR2) within refractory periapical granuloma.", "authors": [ "Eric Chen", "Mahmoud M. Bakr", "Norman Firth", "Robert M. Love", "Eric Chen", "Norman Firth", "Robert M. Love" ], "abstract": "Background: Toll-like receptor-2 (TLR2) is highly important within the immune system. Characterization of the expression of TLR2 within inflammatory cells in periapical lesions could help in diagnosis and management of refractory cases. The aim of the study is identification of Toll-like receptor (TLR2) through immunohistochemical and immunofluroscence expression in inflammatory cells within refractory periapical granuloma cases. Methods: Eight cases of refractory periapical granuloma were selected out of 772 cases. Histological examination and immunohistochemical staining with polyclonal rabbit antihuman TLR2, monoclonal mouse antihuman CD38, CD68 and CD83 primary antibodies, as well as immunofluorescence staining with goat anti-rabbit TLR2, donkey anti-mouse CD38, CD68 and CD83 primary antibodies was conducted. Positive controls, negative controls and experimental sections with no primary antibody were included in the study. Qualitative analysis and double immunofluorescence technique was used to characterize the TLR+ cells. Results: In periapical granuloma, lymphocytes (CD38 cells) expressed the most amount of TLR reactivity followed by macrophages (CD68 cells), and odontogenic epithelial cells. Neutrophils, red blood cells (RBCs) and collagen ground substance were negative to TLR2. Conclusion: TLR2 was highly expressed by lymphocytes and plasma cells indicative of their major role in the inflammatory process and antigen recognition in refractory periapical granuloma. Dendritic cells expressing TLR2 were low in number suggesting a minor role in sustaining these lesions.", "keywords": [ "Antigen presenting cells", "chronic inflammatory cells", "immune response", "periapical granuloma", "toll-like receptors." ], "content": "Introduction\n\nThe goal of conventional root canal treatment is to eliminate root canal microbes, or to substantially reduce microbial load and prevent re-infection by root canal filling and coronal seal. When this is achieved, healing of the peri-radicular tissue is expected. The environment of a well-treated root canal is harsh for most microbes as there are virtually no nutrients for growth. However, microbes associated with persistent intra-radicular infection take on this local ecological change and are viewed as opportunistic pathogens where microbial competitors are eliminated1. However, residual microbes in the apical portion of the root canal system are the major cause of apical periodontitis persisting post-treatment in both poorly and properly treated cases2.\n\nIn refractory periapical lesions, pathogen recognition through innate immunity is achieved by several germline encoded pattern recognition receptors (PRRs). These receptors recognise pathogen-associated molecular patterns (PAMPs) that are conserved, unchangeable and essential for the survival or pathogenicity of the microorganisms3. Some examples of PAMPs include lipopolysaccharides (LPS) of Gram-negative bacteria, Lipoteichoic acids (LTA) of Gram-positive bacteria, unmethylated CpG motifs of bacterial DNA, double-stranded RNA of RNA viruses and mannans of yeast cell walls. PRRs exist in several forms and are found in various locations. They can be expressed on the cell surface as transmembrane proteins (Lectin receptors and Toll-like receptors TLRs), in the intracellular compartments as enzyme (protein kinase PKR) or proteins (NOD-like receptors) or secreted into the bloodstream and tissue fluids as serum proteins (Mannan-binding lectin). The principle functions of PRRs include opsonisation and phagocytosis of pathogens, activation of complements and coagulation cascades, initiation of pro-inflammatory signalling pathways and induction of apoptosis4.\n\nToll-like receptors (TLRs) are type I integral membrane proteins on antigen presenting cells (APCs). Without an effective TLR system, the host is prone to microbial infection and subsequent fatality. It has an extracellular domain, a trans-membrane region and an intracellular domain. In brief, the extracellular domain recognises a variety of antigens during initial microbial challenge. The intracellular domain, with the help of other intracellular proteins, signals the activation of transcription factor NF-κB (nuclear factor kappa-light-chain-enhancer of activated B cells) and ultimately produces inflammatory cytokines that are responsible for mounting a more effective innate response and initiating adaptive immunity5. Activation of NF-κB has been described as the master-switch of the immune system6.\n\nExpression of TLRs is modulated by a variety of factors such as microbial invasion, microbial components and cytokines. Colony-stimulating factor 17, macrophage immigration inhibitory factor8, interferon (INF) -γ9, IL-2, IL-15, IL-1β and TNF-α10 are examples of cytokines that prime phagocyte response to microbial stimulation, up- or down- regulate TLR gene expression and augment inflammatory cytokine production. Zarember and Godowski11 showed that most tissues express at least one TLR, and phagocytes/macrophages in particular show abundant expression of all known TLRs. Moreover, TLR3 is preferentially expressed in T helper cells (Th) and cytotoxic T cells (CTL), whereas B cells show a higher expression of TLR9, TLR 10. Mast cells have been shown to express TLR2, 4, 6 and 8 but not 512.\n\nTLR2 recognises components from a variety of microorganisms. Components that have been demonstrated are peptidoglycan and lipoteichoic acid (LTA) from Gram-positive bacteria13, lipoprotein from Gram-negative bacteria14, glycoinositolphospholipids from protozoa15 and zymosan from fungi16. Furthermore, TLR2 modulates the immune response through mediating the secretion of some antimicrobial peptides in response to bacterial lipoprotein, thereby regulating the direct killing of microbes at different parts of the human body17.\n\nIn the pulp and peri-radicular tissues, TLR2 recognises peptidoglycan and LTAs that are major cell wall constituents of Gram-negative bacteria18. In vivo immunohistochemical analysis has shown that TLR2 is expressed in the pulps of mice19,20 and human teeth21. Interestingly, when stimulated with LTA in vitro, odontoblasts were able to upregulate their expression of TLR2 and initiate an innate immune response, secreting chemokines to recruit immature defensive cells (DCs) while down-regulating dentine matrix synthesis and mineralisation22. The expression of TLR2 on odontoblasts reiterates their role as one of the major surveillance cells in the dentine-pulp complex. Furthermore, immunohistochemistry investigations revealed TLR2 expression in human refractory periapical lesions23 as well as inflamed periodontal ligament around the furcation area24.\n\nThe cells in human periapical granulomas have been characterized and quantitatively analyzed in numerous studies using morphometric, immunoperoxidase and immunofluorescence techniques25–32. All these studies have shown that lymphocytes, plasma cells, macrophages and monocytes were major constituents within the periapical granuloma but only one investigated the presence of mature dendritic cells32. Although the discovery of the TLRs is relatively recent, much progress has been made in our understanding of the mechanisms of innate immune recognition. TLRs recognise microbial components, trigger the innate inflammatory response and instruct the adaptive immune response. However, more research is required to answer the many remaining questions especially with regards to signalling pathways, its role in immune disorders as well as its role in infectious diseases.\n\nThe aim of the present study was to phenotypically characterize TLR2 expressing immune cells within refractory periapical granuloma using immunohistochemical and immunofluroscence expression to demonstrate the specific immune cells with toll-like receptor 2.\n\n\nMethods\n\nThe project received category-A ethical approval from the University of Otago Ethics Committee (07/210) and approval from Ngai Tahu Research Consultation Committee.\n\nPeriapical lesions removed during apical surgery of refractory endodontic cases between 2008 and 2010 were collected from the histopathology-archived records of the University of Otago, School of Dentistry, Oral Pathology Diagnostic Laboratory. Eight cases were selected from the sample pool of 772 for the experiment according to the selection criteria:\n\n1. All lesions were diagnosed as periapical granuloma by registered oral pathologists.\n\n2. All cases were treated at the postgraduate endodontic clinics, School of Dentistry, University of Otago.\n\n3. All periapical granulomas were refractory in nature.\n\n4. All cases had pre-operative signs and symptoms of apical periodontitis.\n\n5. All apical surgeries and periapical biopsies were performed by registered endodontists or endodontic postgraduate students.\n\n6. All samples had patient’s approval to be stored in the histopathology archive and used in experimental studies.\n\n7. Previous root canal treatment met at least one of the following:\n\n• Previous root canal treatment appeared satisfactory on the preoperative periapical radiographs.\n\n• Previous root canal treatment performed by registered endodontists or endodontic postgraduate students.\n\n• Previous retrograde filling appeared satisfactory.\n\n• Satisfactory coronal seal.\n\n• Apical surgery was the only treatment option without jeopardizing the tooth.\n\nEight cases were sufficient for qualitative (descriptive) analysis and allowed a variety of tissues samples to be examined for different morphological variables. Specimens serving as positive and negative tissue controls for the immunohistochemical (IHC) study were also selected from the histopathology archive of the Medlab Dental Oral Pathology Diagnostic Laboratory. Traumatic oral ulcer tissue was used as the positive tissue control for TLR2, mucocele for CD68 (macrophage, monocyte) and lingual tonsil tissue for CD38 (lymphocyte, plasma cell) and CD83 (mature dendritic cell). Periapical scar tissue was selected as the negative tissue control.\n\nEach histology report and the Haemotoxylin and Eosin (H & E) slides of the specimens in the archive were checked with by the oral pathologist (NF) to reconfirm their diagnosis. The details of patient’s age, sex, relevant medical and dental history, signs and symptoms of the presenting condition, pre- and post-operative periapical radiographs, clinical diagnosis and treatment received were recorded from the dental records. This information aided the case selection process. No information that could disclose a patient’s identity was recorded.\n\nAll excised lesions were fixed in 10% formalin, processed and embedded in paraffin before archival storage. 11 4 μm serial sections were cut using a microtome and individually mounted onto a positively charged slide. The first four slides were subjected to IHC staining with TLR2, CD38, CD68 and CD83 antibodies. The fifth, sixth, seventh and eighth slides were subjected to IF staining with TLR2, CD38, CD68 and CD83 antibodies and the last three slides were used for Double Immunofluorescence (DIF) staining of TLR2 expressing cells with CD38, CD68 and CD83 antibodies respectively. In addition, 4 μm section slides were made from the positive and negative tissue controls that were stained simultaneously with the experimenting slides.\n\nDuring routine histopathological diagnosis, each sample had at least one 4 μm section cut from three different levels of the tissue. These sections were picked up onto the same slide and subjected to Haematoxylin & Eosin (H & E) staining. The H & E stained slides were examined under light microscopy (Leica CTR5000, Leica Microsystems, Wetzlar, Germany) at x10, x20 and x40 magnification objectives. The histology report and diagnosis were made by an oral pathologist. Based on the clinical and histological findings, all samples were confirmed to be symptomatic refractory periapical granuloma.\n\nA pilot study was undertaken to establish optimum dilution for each antibody as well as the time required for conditioning the tissue cells in the automated slide stainer (BenchMark XT, Ventana Medical Systems, Inc. AZ, USA). Sections were picked up onto a slide, deparaffinised in xylene, dehydrated in graded alcohol, washed and then heat treated in sodium citrate buffer (pH 7.0), for 10 min at 80°C, to unmask antigens. The sections were cooled and washed in phosphate-buffered saline (PBS, pH 7.2). Endogenous peroxidase activity was quenched by incubating the sections in a solution of 3% H2O2 in methanol for 15 min. Specimens were washed with PBS and incubated in blocking agent (foetal calf serum - Sigma Aldrich – F0392 – USA) 25 μL per 4 mL of normal saline for 15 min. Sections were incubated with rabbit antihuman TLR2 polyclonal primary antibodies (abcam -ab24192- Cambridge, UK) at 1:500 to 1:700 dilution for 30 min at 25°C. The sections were then washed in PBS and incubated with secondary goat anti-rabbit and goat anti-mouse immunoglobulins in phosphate buffered saline (PBS) using the LSAB2-HRP link system (DakoCytomation – K0675 - Carpinteria, CA, USA).\n\nThe manufacturer’s instructions were followed for the sequential incubation and durations for the exposure to the secondary antibodies. After washing with PBS, the sections were incubated with diaminobenzidine substrate kit (K3468 - Dako, Carpinteria, CA, USA) that resulted in a brown-coloured precipitate at the antigen–antibody binding sites. Finally, all the sections were washed in deionized water, dehydrated and mounted on individual slides. Positive (traumatic oral ulcer) and negative controls (periapical scar) as well as experimental sections with no primary antibody were included in all immunohistochemical runs.\n\nThe preparation of the sections for immunohistochemistry staining with antibodies CD38, CD68 and CD83 was similar to the TLR2 antibody technique. The dilutions used for monoclonal mouse antihuman CD38 (abcam -ab49644- Cambridge, UK), CD68 (abcam -ab955- Cambridge, UK), and CD83 (abcam -ab49324- Cambridge, UK) primary antibodies (Dako) were 1:50-1:100, 1:200-1:400, and 1: 20-1:40 respectively. The duration of incubation was 2 hours for both the primary antibodies. The secondary antibody system used by the automatic immunostainer was polymer-based Universal HRP multimer, which was prediluted and ready to be used in dispensers (#760-500 ultra- View™; Ventana Medical systems Inc.). Finally, the sections were incubated with universal alkaline phosphatase red detection chromogen kit (#760-501 ultraView™; Ventana Medical systems Inc.), washed in deionized water, dehydrated and mounted. Positive (lingual tonsils or mucocele) and negative controls (periapical scar) as well as experimental sections with no primary antibody were included in all immunohistochemical runs. Primary antibodies are listed in Table 1 & Table 2.\n\nA pilot study was undertaken to establish optimum dilution for each antibody as well as the time required for conditioning the tissue cells in an automated slide stainer (BenchMark XT, Ventana Medical Systems, Inc. AZ, USA) which was used only for the deparafinisation and cell conditioning standard steps. The primary antibodies were Alexa Fluor® 488 (Invitrogen -A-11008- Carisbad, CA, USA) goat antirabbit for TLR2 green immunofluroscence staining and Alexa Fluor® 594 (Invitrogen -A-21203- Carisbad, CA, USA) donkey antimouse red immunofluroscence staining for CD38, CD68 and CD83. Positive controls (traumatic oral ulcer for TLR2; lingual tonsils and mucocele for CD38, CD68 and CD83), negative controls (periapical scar) as well as as experimental sections with no primary antibody were included in IF protocol.\n\nAll slides were manually mounted with Vectashield® -aqueous hard set mountant (H- 1000 - Maravai LifeSciences) with nucleus counterstain DAPI (4’,6-diamidino-2-phenylindole, stains nucleus in florescent blue) in a fume cupboard. Slides were examined immediately and photographs taken within a 48 hour period under the immunofluorescence microscope (Lecia M205 FCA). Slides were kept in a 4° C light-tight fridge if not viewed immediately. The optimum primary and secondary antibody incubations were investigated, determined in pilot studies and summarized in Table 2 & Table 3. Primary antibody incubation was at 4°C overnight (18–24 hours) and secondary antibody incubation (Alexa Fluor®) at room temperature for one hour in complete darkness.\n\nTo ensure high quality images that reflect the observation, two of the three controls in the camera setting were set with a fixed value: Gain 3.0 (range between 0.6 – 6.0) and Offset -500 (range between -1100 – +2995). The exposure time measured in milliseconds (ms) was the only adjustable camera setting. This was determined for each fluorochrome at each magnification power (Table 4).\n\nThe parameters of the DIF staining procedure were determined by the IF pilot study. However, the primary antibodies were pre-mixed into a cocktail solution (without changing their optimum dilutions) and incubated simultaneously. This pre-mixing and simultaneous incubation also applied to the secondary labeling antibodies.\n\nHistology and immunohistochemistry (IHC). All H & E and IHC stained sections were viewed under a light microscope (Leica DM5000B, Leica Microsystems, Wetzlar, Germany) under magnifications up to x100 objective. A cell was determined as immuno-positive when it demonstrated distinctive brown stain on the cell membrane and/or cytoplasm around a nucleus. Images were taken using a CCD camera (Leica DC500, Leica Microsystems, Wetzlar, Germany), mounted on the microscope, controlled by computer software (Leica FireCam Version 1.5, Leica Microsystem, Heerbrugg, Switzerland).\n\nImmunofluroscence (IF) and double immunofluorescence (DIF). All IF stained sections were viewed under a fluorescence microscope (Olympus AX70, Olympus Corporation, Center Valley, PA, USA) under magnifications up to x100 objectives. Images were taken using the CMOS camera (Go-3, QImaging, Surrey, BC, Canada) mounted on the microscope and controlled by computer software (Macintosh QCapture Suite, 2.98.2 QImaging, Surrey, BC, Canada). A cell was counted as positive when it demonstrated distinctive fluorescence on the cell membrane and/or cytoplasm surrounding the nucleus. Since the fluorescence microscope only observes one wavelength at a time, the separately labeled protein target and the nucleus cannot be observed simultaneously. To overcome this problem Photoshop (CS5 – 12.0 – White Rabbit - Adobe Systems Incorporated, San Jose, CA, USA) software was employed for qualitative analysis. An area of interest was photographed under different wavelength with the slide remaining stationary. Images were superimposed and screened using the Photoshop software to disclose positive cells.\n\nQualitative analysis of the DIF followed the same principles as IF. A cell was identified to co-express two targeted proteins when the superimposed and screened images showed both green and red fluorescence on the cell membrane and/or cytoplasm. The objective of the DIF qualitative analysis was to identify TLR2 expressing cells as lymphocytes/plasma cells (CD38), Macrophages/monocytes (CD68) and/or mature dendritic cells (CD83).\n\n\nResults\n\nThe routine diagnostic H & E stained sections of the selected periapical granuloma lesions were retrieved from the histopathology-archived records. All tissue sections showed characteristics of granulation tissue (Figure 1a, b, c), typically mature fibrous connective tissue with a moderately intense infiltrate of chronic inflammatory cells dominated by lymphocytes. Occasionally, strands of stratified squamous epithelium of odontogenic origin (epithelial rests of Malassez) were found interspersed in the granulation tissue of some lesions. In the periapical scar (negative tissue control) inflammatory cells were absent and the lesion was characteristically acellular, with the exception of fibroblasts associated with collagen, with a dense avascular collagen structure (Figure 1d).\n\n(a) A histopathology section of a selected refractory periapical granuloma showing areas of fibrous connective tissue (F), blood vessels, inflammatory cells (I) and interspersed odontogenic epithelium (Haematoxylin & Eosin staining x50), (b) Proliferating epithelial cells (E) surrounded by chronic inflammatory cells (I) (Haematoxylin & Eosin staining x200), (c) Fibrous connective tissue (F) with moderate chronic inflammatory cell infiltrate (I) (Haematoxylin & Eosin staining x200), (d) Histopathology section of a periapical scar showing the un-inflamed, relatively acellular and avascular dense collagen tissue (Haematoxylin & Eosin staining x200).\n\nCD38. In the lingual tonsil section (positive control), clusters of lymphocytes within the germinal centres were positively stained and appeared as small circular or oval brown cells that were closely packed together (Figure 2a). All the periapical granuloma samples showed CD38+ cells and had the same staining pattern as the CD38+ cells in the lingual tonsil. These CD38+ cells dominated the inflammatory cell infiltrate and were mostly found in large clusters evenly distributed in the granulation tissue with some individual positive cells scattered in between (Figure 2a, b). A closer look of the CD38+ cells under high power magnification (x1000) revealed that the brown stains were mainly located on the cell membrane (Figure 2c, d). However since the surrounding cytoplasm could be quite narrow in width it was not easy to distinguish under low power magnifications (x200 and x400). The CD38+ cells are roughly 7 to 10 μm in diameter. In the negative controls no positive staining was found.\n\n(a) Representative histopathology section of positive control (lingual tonsil) showing CD38+ cells (black arrows) in the germinal centres (CD38 Immunohistochemistry x400), (b) Representative histopathology section from a periapical granuloma showing a cluster of CD38+ cells (red arrows) (CD38 Immunohistochemistry x400), (c, d) Two images of CD38+ cells under high power magnification showing characteristic lymphocyte and plasma cell circular or oval shape with staining mainly on the cell membrane (CD38 Immunohistochemistry x1000).\n\nCD68. In the mucocele (positive control), CD68+ cells appeared as large globular spongy/foamy brown cells, roughly 20 μm in dimension (Figure 3a). They were found adjacent to the cyst lining inside the cyst cavity. Under high power magnification (x1000) small cytoplasmic vesicles and large phagosomes were observed which are also characteristic of macrophages (Figure 3b). The brown stain was mainly located on the cell membrane and cytoplasm.\n\n(a) Representative histopathology section of a mucocele showing positively stained macrophages (red arrows), cystic lining (CL) and cystic cavity (CC) (CD68 Immunohistochemistry x400), (b) High power magnification of CD68+ cells in a mucocele histopathology section. Cytoplasmic vesicles (black arrows) and phagosomes (red arrows) are clearly demonstrated (CD68 Immunohistochemistry x1000), (c) Representative histopathology section of a symptomatic periapical granuloma showing CD68+ cells which stained brown (black arrows) (CD68 Immunohistochemistry x400), (d) High power magnification of the larger CD68+ cells showing spongy appearances and wavy cellular outlines which resembled macrophages. Positive immunostaining is located on the cell membrane as well as the cytoplasm. Unstained intracellular vesicles can be identified (black arrows) (CD68 Immunohistochemistry x1000), (e) High power magnification of the smaller CD68+ cells showing a rounded cellular profile which resembled monocytes (black arrows) (CD68 Immunohistochemistry x1000).\n\nIn the periapical granuloma lesions, five out of eight samples were identified to contain CD68+ cells (Figure 3c). CD68+ cells identified in the five periapical granuloma lesions were abundant but were mainly located in localized areas of the granulation tissue. The remaining three samples did not show distinctive positive staining. Furthermore, two morphologically distinctive cells were observed in the periapical granulomas; large spongy cells that measured up to 20 μm across with irregular wavy outline that appeared to be macrophages (Figure 3d) and small circular cells that were about 10 μm in diameter which resembled monocytes (Figure 3e). Under high power magnification (x1000) small cytoplasmic vesicles were observed within the macrophages. These vesicles were not stained and stood out from the stained cell membrane and cytoplasm and thus contributed to the spongy appearance. No positive staining was found in the negative controls.\n\nCD83. In the lingual tonsil (positive control) sample, CD83+ cells were sparingly scattered throughout the germinal centres (Figure 4a). These cells exhibited variable morphology, being circular, oval or asymmetrical polygonal cells. In addition, some CD83+ cells showed small thorn-like or filamentous extensions under high power magnification (Figure 4b), which is characteristic of dendritic cells. The positive immunohistochemical stain was observed on the cell membrane and the cytoplasm. In the periapical granuloma samples, CD83+ cells were not found in two samples and appeared to be rare in another two samples. In the remaining four samples CD83+ cells were readily detected, however, they constituted a very small population in the inflammatory cell infiltrate (Figure 4c). The size of the CD83+ cells were variable depending on the out stretch of cytoplasmic extensions. The cell body may be 10 μm in diameter and the cytoplasmic extensions up to 5 μm have been observed (Figure 4d, e). All negative controls showed negative immunostaining.\n\n(a) Representative histopathology section of a germinal centre within lingual tonsil demonstrating CD83+ cells (black arrows) (CD83 Immunohistochemistry x400), (b) High power magnification of a CD83+ cell showing a polygonal outline with little thorn-like or filamentous extensions (black arrows) characteristic of a dendritic cell (CD83 Immunohistochemistry x1000), (c) Representative histopathology section from a periapical granuloma showing CD83+ cells (black arrows). Characteristically these cells were found in low numbers within a lesion (CD83 Immunohistochemistry x200), (d, e) Two images of high power magnification of CD83+ cells in a periapical granuloma showing some examples of the variable cellular morphology they exhibited. Little thorn-like extensions can be seen on some cells (black arrows) indicative of dendritic cells (CD83 Immunohistochemistry x100).\n\nTLR2. In the positive control, traumatic oral ulcer tissue, various mononuclear inflammatory cells and the intact oral epithelial cells were stained brown on their cell membrane and cytoplasm, demonstrating TLR2 expression (Figure 5a). Similarly, all periapical granuloma samples showed various mononuclear inflammatory cells displaying the same staining pattern observed in the traumatic oral ulcer (positive control) (Figure 5b, c). The most abundant TLR2+ cells in the periapical granuloma where small circular or oval cells of about 7 to 10 μm in diameter and appeared to be lymphocytes (Figure 5d). In some samples, TLR2+ cells exhibited foamy appearances which resembled macrophages (Figure 5e). These foamy cells are roughly 12 to 15 μm across, almost double in size compared to lymphocytes. Occasionally, interspersed odontogenic epithelial cells also stained brown, suggesting TLR2 positivity. On the other hand, neutrophils, red blood cells (RBC) and collagen ground substance were not stained. In the negative controls no cells or structures showed TLR2 immunostaining.\n\n(a) Representative histopathology section of a traumatic oral ulcer showing an area of intact oral epithelium (OE) which was positively stained for TLR2. Mononuclear inflammatory cells (red arrows) in the connective tissue are also stained brown (TLR2 Immunohistochemistry x400), (b): Representative histopathology section from a periapical granuloma showing numerous mononuclear inflammatory cells stained brown (red arrows), demonstrating TLR2 expression (TLR2 Immunohistochemistry x400), (c) Representative histopathology section from a periapical granuloma showing TLR2+ immunostaining on numerous mononuclear inflammatory cells (red arrows) as well as interspersed odontogenic epithelium (E) (TLR2 Immunohistochemistry x400), (d) Image of TLR2+ cells resembling immune cells of lymphoid lineage under high power magnification (red arrows). Note the cell membrane and cytoplasm are stained brown (TLR2 Immunohistochemistry x1000), (e) High power magnification of a periapical granuloma histopathology section showing large foamy TLR2+ cells resembling macrophages (red arrows). Note that the neutrophils (black arrows), characterised by multi-lobed nuclei, were negative to TLR2 immunostaining (TLR2 Immunohistochemistry x1000).\n\nIn the positive control tissue, various inflammatory cells were observed to fluoresce on the cell surface and in the cytoplasm. The oral epithelium also expressed TLR2 and was identified by its location, distinct morphology and cellular arrangement. Red blood cells (RBC) were identified by their autofluorescent glow, biconcave cellular profile and the absence of a nucleus. All negative controls did not show positive fluorescence.\n\nSome lymphocytes within the germinal centres of the lingual tonsil showed CD38 positivity. These CD38+ cells were small and had a circular or oval profile (Figure 6a). They appeared to be tightly packed to each other, arranged in clusters separated by the CD38- cells and appeared as collections of red fluorescent rings. Under high power magnification, the red fluorescence was mainly located on the cell membrane (Figure 6b). The cytoplasm typically was not stained and appeared as a thin black space surrounding the nucleus. The size of the CD38+ cells was about 10 to 13 μm in diameter. Red blood cells (RBCs) were identified by their smaller size (about 7 μm across), faint purple autofluorescent glow, biconcave cellular profile and the absence of a nucleus. All negative controls showed negative fluorescence (Figure 6c, d).\n\n(a) Representative histopathology section of CD38+ cells in lingual tonsil tissue showing collections of red fluorescing cells (R) separated by areas of CD38- cells that had no red fluorescence (N). RBCs had faint purple autoflorescence (white arrows) (CD38 Immunofluroscence x400). (b) High power magnification of CD38+ cells in lingual tonsil tissue. These cells have a circular or oval profile with a narrow space of cytoplasm surrounding the nucleus, consistent with being lymphocytes. The red fluorescence is located on the cell membrane whereas the cytoplasm was negative. These cells are closely associated to one another. This image also shows a blood vessel containing RBCs which glowed in faint purple (white arrow) (CD38 Immunofluroscence x1000). (c) Representative section of negative antibody control. (d) Representative section of negative tissue control (periapical scar) (x400). Note the red fluorescing cells represent RBCs (white arrows) which exhibit autofluorescence and were scattered in the tissue or within the blood vessels.\n\nMacrophages in the mucocele expressed CD68 and appeared as large red fluorescent spongy cells measured up to 25 μm across (Figure 7a, b). They were found in large numbers in the cystic cavity just beneath the epithelial lining. Under high power magnification the red fluorescence was located on the cell membrane as well as the cytoplasm. Also observed were the intracellular vesicles and phagosomes which contributed to the spongy appearance of CD68+ cells (Figure 7b). The negative controls showed no fluorescence (Figure 7c, d).\n\n(a) Representative histopathology section showing CD68+ cells in a mucocele fluorescing in red. They are found in large number adjacent to the cystic lining (CC) and spread into the cystic cavity (CC). Characteristics of macrophage cells are evident e.g. irregular cell shape, intra-cellular vacuoles (x400). (b) High power magnification of CD68+ cells in a mucocele showing their large spongy globular appearance. The red fluorescence is observed on the cell membrane and cytoplasm but not the intracellular vesicles or phagosomes (x1000). (c) Representative section of negative antibody control. (d) Representative section of negative tissue control (periapical scar) with autofluorescing RBCs (white arrows) (x400).\n\nIn the lingual tonsil tissue, CD83+ cells were a minority group amongst resident cells. Some CD83+ cells were located within germinal centres, connective tissue or in between minor salivary glands. These cells displayed red fluorescence on the cell membrane and cytoplasm. They exhibited several cellular outlines including circular, oval, teardrop or asymmetric polygons (Figure 8). Under high power magnification, little thorn-like projections were sometimes seen extending from the cell surface which is a characteristic of dendritic cells. The cell body of the CD83+ cells are roughly 10 μm in diameter with the cytoplasmic extensions measuring up to 5 μm in length (Figure 8). All negative controls did not show fluorescence.\n\nThe CD83+ cells exhibit several cellular shapes: circular, oval, teardrop and asymmetric polygons. Occasionally, little thorn-like projections could be detected on cell surfaces (white arrows). These characteristics are typical of dendritic cells (x1000).\n\nThe cellular characteristics and distribution of each cell marker in the DIF experiments were similar to IF. In all samples of periapical granulomas, the majority of inflammatory cells and odontogenic epithelium expressed TLR2. CD38+ cells (lymphocytes) remained the most abundant inflammatory cells in all the periapical granuloma lesions followed by CD68+ cells (macrophage). In contrast, CD83+ cells (dendritic cells) were few or rare in the samples.\n\nCD38+ cells expressing TLR2+ (CD38+/TLR2+) were identified in all eight periapical granuloma samples and were abundant throughout the lesion. Under high power magnification CD38 was located on the cell membrane with the cytoplasm left unstained (Figure 9). On the other hand, the TLR2 green fluorescence appeared to have stained both the cell membrane and cytoplasm. The CD38+/TLR2+ cells had a circular profile that resembled lymphocytes in the lingual tonsil and were closely packed together like soap bubbles. These cells displayed a thin orange-red cell membrane which enclosed the narrow green cytoplasm which surrounded the nucleus. The results indicate that lymphocytes express TLR2 in refractory periapical granuloma lesions.\n\nImage A shows CD38+ cells with red fluorescing cell membrane and unstained cytoplasm. Image B shows TLR2+ cells displayed green fluorescence on the cell membrane as well as on the cytoplasm. The superimposed image C reveals CD38+/TLR2+ cells which have orange-red cell membrane and green cytoplasm (white arrows) (CD38/TLR2 Double Immunofluorescence x1000).\n\nThe co-expression of CD68 and TLR2 on cells was identified in five periapical granuloma samples. Under high power magnification both CD68 and TLR2 fluorescence stained the cell membrane and cytoplasm but not the intracellular vesicles. The CD68+/TLR2+ cells fluoresced as a mixture of orange, red and green. The CD68+/TLR2+ cells were large and had spongy globular profiles that resembled macrophages in the mucocele. They were abundant but were found in localized areas of a lesion i.e. they were not dispersed throughout a lesion (Figure 10). Many RBCs were found interspersed in the area. These RBCs autofluoresced in purple and yellow on the red and green images respectively. Although they may appear as orange cells in an overlapped image, they are distinguished by their smaller size, biconcave profile and lack of nucleus (Figure 11).\n\nImage A shows CD68+ cells fluorescing in red and B shows TLR2+ cells fluorescing in green. The superimposed image C reveals the CD68+/TLR2+ cells (white arrows) (CD68/TLR2 Double Immunofluorescence x400).\n\nImage A shows CD68+ cells fluorescing in red. Notice the cell membrane and cytoplasm are stained but not intracellular vesicles. Image B shows TLR2+ cells fluorescing in green. Notice the multi-lobular nuclear cells resembling neutrophils were negative to TLR2 staining (white arrows). The superimposed image C reveals CD68+/TLR2+ cells which have cell membrane and cytoplasm fluorescing in orange or a mixture of red and green speckles (white arrows) (CD68/TLR2 Double Immunofluorescence x1000).\n\nThe co-expression of CD83 and TLR2 were identified in six periapical granuloma samples. Under high power magnification both CD83 and TLR2 fluorescence were observed on the cell membrane and cytoplasm. The CD83+/TLR2+ cells fluoresced as a mixture of orange, red and green lights (Figure 12). These cells exhibited thorn-like extensions and resembled dendritic cells in the lingual tonsil and were sparingly distributed throughout a lesion. The results are consistent with dendritic cell expression of TLR2 in periapical granuloma (Figure 13).\n\nImage A shows CD83+ cells fluorescing in red. Notice the cells have variable outlines and cytoplasmic extension. They contributed as a minor cell population in the periapical granuloma. Image B shows TLR2+ cells fluorescing in green. The superimposed image C reveals the CD83+/TLR2+ cells in orange-red fluorescence (white arrows) (CD83/TLR2 Double Immunofluorescence x400).\n\nImage A shows CD83+ cells fluorescing in red. Notice the cell membrane, cytoplasm and cytoplasmic extension (white arrows) are stained. Image B shows TLR2+ cells fluorescing in green. The superimposed image C reveals CD83+/TLR2+ cells (white arrows) which have cell membrane and cytoplasm fluorescing in orange or a mixture of red and green speckles (CD68/TLR2 Double Immunofluorescence x1000).\n\n\nDiscussion\n\nIn a long-standing apical periodontitis, a periapical granuloma is the most common pathological entity33,34. The lesion itself is a granulomatous tissue infiltrated with inflammatory cells and encapsulated by a fibrous capsule. Lymphocytes are the dominant inflammatory cell in the lesion followed by macrophages, plasma cells, and neutrophils. The capsule mainly consists of collagen fibres and fibroblasts. Blood vessels and sometimes odontogenic epithelial cells are found interspersed throughout the entire lesion35. This is consistent with the histological picture observed in the current study. Neither acute inflammatory cells nor areas of necrosis were evident in the selected cases within the present study as they were all refractory periapical granuloma cases with no signs of acute exacerbation. Lymphocytes being the most numerous cells in periapical granuloma, they showed the most abundant TLR+ reactivity.\n\nResidual microbes in the apical portion of the root canal system is a major cause of apical periodontitis persisting post-treatment in both poorly and well treated cases2. In symptomatic refractory cases, the microbes found are those similar to primary endodontic infections36. In well-treated endodontic cases the number of species that were recovered was typically between 1 to 5 species per canal37 and was mostly Gram-positive facultative or obligatory anaerobes1. A wide diversity of microbes had been isolated in different studies but Enterococci are the most reported species with prevalence between 32–77%, followed by Streptococcus (20–33%) and Actinomyces (3–27%) with Candida, a fungal species, making up a significant proportion (3–9%)20–24,33–41. TLR2 recognizes a number of components on cell walls of a variety of microorganisms including peptidoglycan and lipoteichoic acid (LTA) from Gram-positive bacteria13. This explains the constant expression of TLR2 throughout all cases of refractory periapical granuloma. Lipoteichoic acids (LTA) of Gram-positive bacteria are one of the pathogen-associated molecular patterns (PAMPs) that are conserved, unchangeable and essential for the survival or pathogenicity of the microorganisms3. Therefore, the presence of pattern recognition receptors (PRRs) including TLR2 is an essential mechanism of pathogen recognition in innate immunity4.\n\nThe results showed a positive reaction to TLR2 by odontogenic epithelial cells within the refractory periapical granuloma cases. This is in agreement with a number of other studies where members of the TLR family were identified in a number of epithelial tissues within the body. Examples include epithelial cells on the mucosal surface of the intestinal tract having TLR5 expressed on their basolateral surface42, and TLR4 expression at relatively low levels in intestinal epithelial cells, which may explain why LPS does not elicit a strong inflammatory response in the intestine43. In contrast, intestinal epithelium from patients with inflammatory bowel diseases showed augmented expression of TLR444.\n\nTLR2 ligand recognition, unlike TLR4, involves cooperation with other TLR family members, particularly TLR6 and TLR1. Studies have shown that macrophages in mice that have competent TLR2 but are deficient in TLR6 failed to recognise microbial lipopeptides45,46. The same conclusion was made when a similar experiment was performed on TLR1-deficient mice47. TLR6 and TLR1 are highly homologous and they may, in some cases, compensate for each other when one is deficient48. It is unknown whether TLR2 forms a dimer with other TLRs constitutively or in response to ligand stimulation. This could explain the fact that TLR2 was not expressed in some of the immune cells within refractory periapical granuloma, as it might have been combined with other members of the TLR family within lesions. Furthermore, Expression of TLRs is modulated by a variety of factors such as microbial invasion, microbial components and cytokines. Colony-stimulating factor 17, macrophage immigration inhibitory factor8, interferon (INF) -γ9, IL-2, IL-15, IL-1β and TNF-α10 are examples of cytokines that prime phagocyte response to microbial stimulation, up- or down- regulate TLR gene expression and augment inflammatory cytokine production. Within refractory periapical granuloma, a large number of the above mentioned inflammatory cytokines are produced as a host defensive response and could contribute to the expression of TLR2 within the lesions.\n\nThe consistent expression of TLR2 within refractory periapical granuloma reported in the current study is in agreement with a number of other studies in the literature. Immunohistochemistry investigations on human refractory endodontic lesions only revealed TLR2 but failed to consistently demonstrate TLR4 expression23. The consistent expression of TLR2 in the refractory periapical granuloma supports the observation that Gram-positive bacteria play a dominant role in the root canal infection of these lesions. Other studies investigated the expression of TLR2 in root resorption. Concomitant infection originating from the root canal or the periodontium may initiate an inflammatory response which resorbs unprotected dentine49,50. Lesions isolated from internal, external inflammatory root resorption as well as cervical root resorption have been identified to express TLR2 and TLR451. Although the causative factors in each resorption may be different or somehow obscure, the expression of TLRs strongly suggests that bacteria play the central role of their pathogenesis.\n\nFurthermore, the expression of TLR2 and TLR4 was investigated in the furcal periodontal ligament of the rat molar24. Real-time PCR analysis of the furcal periodontal ligament showed that an unsealed pulpotomy exposed to the oral cavity increased the gene expression of TLR2 and TLR4, as well as several APC markers (MHC-II, CD83 and CD86 also known as T-Lymphocyte activation antigen) at 24 hours. Histochemical analysis also confirmed the presence of TLR2 and TLR4 positive cells in the inflamed periodontal ligament. Finally, TLR2 and TLR4 expressions in the dental pulp has been studied using a murine dental caries model. TLR2 and CD64 positive cells were found in abundance around the odontoblastic layer and at the centre of the pulp, while TLR4 positive cells were very few in both areas. Nonetheless, it was evident that both TLR2 and TLR4 were induced on the macrophages and dendritic-like cells at the early stage of pulpitis. However, it was reported that mRNA concentration for TLR2 was always higher than that of TLR4, reflecting the dominant role of Gram-positive bacteria in the early carious lesions19,20. The consensus from the previous studies provides solid evidence to support the findings from the present study that TLR2 plays a key role in the immune response against long standing bacterial infections, especially gram +ve predominant lesions.\n\nOur results showed that neutrophils, RBCs, the collagen ground substance and the negative control samples from periapical scar tissues showed a negative reaction to immunostaining with TLR2 antibody. This is due to the absence of the stimuli needed for TLR2 activation such as bacterial antigens and inflammatory conditions as hypothesized by a previous study that investigated the expression of TLR2 in periapical lesions52. The same study is in accordance with the current study in terms of TLR2 expression within macrophages with foamy appearance52. The foamy appearance is linked to the production of nitric oxide and reactive oxygen species (ROS) necessary for the pathogenesis of chronic periapical lesions, rather than being a result of necrotic and/or apoptotic TLR2+ cells53. TLR2 might play a role in the production of ROS after activation by bacterial antigen, together with macrophages54. Further studies are necessary to clarify the possible correlation between TLR2, macrophages and ROS in the pathogenesis of periapical lesions.\n\nThe characterization of TLR2+ cells requires the co-expression of TLR2 with an immune cell marker. It is important that the paired antibodies do not cross react during immunostaining and give a false identification. There are several methods to reduce the cross reactivity, all of which are based on the principle of sequential antibody application and the removal or destruction of the binding sites after each antigen-antibody interaction55,56. However, these methods are time consuming and greatly increase the chance of procedural errors. The most effective way is to use primary antibodies raised in different species57,58, allowing them to be pre-mixed in a single cocktail of antibody diluents and react with the antigens in one application. The same principle was also followed in the present study when choosing the labeling antibodies in IF.\n\nIn the present study, various immune cells present in refractory periapical granuloma were identified to express TLR2 which was in agreement with Desai et al. (2011)59. The most common TLR2+ cells in refractory periapical granuloma resembled the morphology of lymphocytes and macrophages. Finally, the odontogenic epithelium found interspersed in the lesion also expressed TLR2 although this was not reported by Desai et al. (2011)59. In contrast to Hayashi et al. (2003)60 but in line with Desai et al. (2011)59, the neutrophils identified in the lesions (characterised by their multi-lobed nucleus) did not reveal the presence of TLR2. The differences may be due to the methodology employed – Hayashi et al. (2003)60 used isolated cells in flow cytometry analysis which is more sensitive in antigen detection than immunofluroscence on tissue sections61.\n\nIn the current study, TLR2 stain was located on the cell surface as well as cytoplasm. Two reasons can be given to explain this observation: biologically, TLR2 is actively synthesized in the cytoplasm and thus it was stained. Technically, the histology section (which is 4 µm in thickness) contains a significant portion of a cell in three dimensions (typical lymphocytes are about 8 to 10 μm in diameter) and thus the stained cell membrane in three dimensions is viewed as stained cytoplasm.\n\nCD38 is a cell surface enzyme involved in transmembrane signaling and cell adhesion62. It is expressed in activated T cells, B cells and plasma cells63. CD38 was selected for this study because it encompasses the entire range of lymphocytes. CD38+ cells representing lymphocytes and plasma cells dominated the infiltrated immune cell population in the refractory periapical granuloma which is consistent with most studies28,31. In areas of secondary abscess where macrophages, monocytes and neutrophils are found, no CD38+ cells could be identified.\n\nCD68 is a transmembrane glycoprotein highly expressed by human monocytes and tissue macrophages64,65. In this study, monocytes and macrophages were identified in 5 samples. They were distinguished according to their morphology, macrophages are large cells (up to 25 µm across) with foamy cytoplasm and monocytes are smaller cells (about 10 µm) with a circular profile. They were found abundantly in areas of secondary abscess which also revealed numerous neutrophils (identified by multi-lobed nucleus). No CD68+ cells were found outside an area of abscess. Although not quantitatively analyzed, the number of macrophages and monocytes were relatively high which was consistent with Stern et al. (1981)25, Johannessen et al. (1984)66, Nilsen et al. (1984)27, Kopp & Schwarting (1989)29, and Piattelli et al. (1991)30 but not Bergenholtz et al. (1983)26 or Kontiainen et al. (1986)28. The proportion of macrophages in the apical lesion has been suggested to correlate positively with clinical signs and symptoms67. All samples selected in the current study were symptomatic which may account for the substantial number of macrophages observed.\n\nThree samples failed to show CD68+ immunofluroscent cells. Two reasons may be given for their absence. Firstly, periapical granuloma are heterogenous in cellular and structural components35. The composition of cells residing in the periapical granuloma can shift in response to the dynamic relationship between the microbial and host factors. In addition, it has been shown that a secondary abscess, where macrophages and neutrophils were found in the present study, can occur anywhere inside the granulation tissue and not necessarily close to the apical foramen35. Therefore, the level where the tissue is sectioned may not include the macrophage-rich areas. Secondly, the tissue samples stored in the archives do not represent the entire lesion because a significant portion had already been removed during the routine diagnostic procedure. Furthermore, the lesions may not always be enucleated completely at the surgery. Some tissue portions may have been left in the bony crypt, lost through the suction tip or were removed in several pieces.\n\nCD83 is a glycoprotein that is expressed on the cell membrane of mature dendritic cells (DC)68,69. In this study, six samples had identified CD83+ cells which exhibited variable body shape (about 10 µm across) and many showed small to large cytoplasmic extensions that are characteristic of dendritic cells32. Although not quantitatively analyzed, the CD83+ cells were infrequently observed and randomly distributed in the lesion. This was consistent with Liapatas et al. (2003)31 and Lukić et al. (2006)32 who quantitatively showed DCs have weak expression (0-10 % and 0.4–1.2% of the total infiltrated cell population respectively) in periapical lesions.\n\nDesai et al. (2011)59 was the first group to characterize TLR2+ cells in periapical lesions using serial section staining by methyl green pyronin (plasma cells) and immunoperoxidase (CD3 and CD19 cells) techniques. Although the expression of TLR2 on T cells (CD3+), B cells (CD19+), plasma cells and macrophages were revealed, the authors acknowledged that their identification was not definitive and more specific cell marking was required to determine cell phenotypes. In this study the DIF technique was employed to overcome the lack of specificity in co-localizing cell markers using serial staining. Both the TLR2 and the respective CD markers were located on the cell membrane and cytoplasm although the intensities and patterns of the fluorescence varied. Overall, the study was in agreement with Desai et al. (2011)59. Furthermore, the current study has successfully marked monocytes (CD68+) and mature dendritic cells (CD83+) and supports the speculation that TLR2+CD3-CD19- MGP cells could be dendritic cell subsets or macrophages59.\n\nIn this study, mature DCs were observed scattered throughout the lymphocyte rich areas and some appeared physically close to them. Since mature DCs are highly mobile units, this may suggest they are actively moving and performing antigen presentation to local lymphocytes but the minute number of mature DC observed remain unexplained. However, the population of DCs in all organs is maintained through a dynamic balance of three parameters: continuous input of pre-DCs from blood, limited DC division in situ, and cell death70. Since mature DCs ultimately migrate to the local lymph nodes, their small presence in periapical granuloma may be due to the difference between their exodus and replenishment from circulating pre-DCs.\n\nToll-like receptor 2 expressed on B cells may play two roles within the periapical granuloma. Firstly, it may work with the surface immunoglobulin in taking up the antigen thus facilitating B cells as Antigen Presenting Cells (APC). Secondly, the ligation of LTA to TLR2 on memory B cells may reactivate B cell immune responses without signals from T cells71,72. In contrast to memory B cells, the expression of TLRs on plasma cells remains unknown in the current literature. Nonetheless, the possibility that plasma cells express TLR2 has been suggested59,72 and appeared likely in this study. It has been hypothesized that direct activation of plasma cells by TLR ligands could contribute to plasma cell-derived IgG secretion72.\n\nIn this study, macrophages were identified to express TLR2. Unlike dendritic cells, macrophages were abundant but only located in the areas of secondary abscesses and they were not found scattered throughout the lymphocyte-rich regions. This may suggest the main role of macrophages in the refractory periapical granuloma is to recognize microbes through TLR2 and to perform phagocytosis. Antigen presentation to lymphocytes may be a lesser role for macrophages and only occurs to nearby lymphocytes. In this study, macrophages expressed TLR2 both on the cell membrane and cytoplasm. Since the size of a macrophage is typically 20 µm, the detection of the TLR2 in the cytoplasm is best explained by the role they play in periapical granuloma. To sustain effective microbe recognition and their removal, the cell membrane of the macrophage requires constant resupply of TLR2 and thus TLR2 protein synthesis is up-regulated in the cytoplasm. Furthermore, as a scavenger, the positive staining may represent the phagocytosed cellular debris that had expressed TLR259.\n\nFurther investigations to identify lymphocytes and their sub-populations have yet to be disclosed. The identification of these lymphocytes and their expression of TLR2 may provide an insight into the effect these cells have on the pathogenesis of refractory periapical granuloma. Furthermore, a quantitative analysis based on flow cytometry could be supplemented to obtain additional information on the proportion and intensity of TLR2 expression amongst immune cells. Based on this research model, periapical lesions of endodontic origin with varied characteristics (such as primary or secondary, symptomatic or asymptomatic, granuloma or cyst) could be explored with different classes of TLR. Toll-like receptor pathways act as a bridge between the innate and adaptive immunity. Detailed investigations on the pattern of TLRs expression and cellular interaction may contribute new discoveries regarding the dynamic relationship between bacteria and host, the development and growth of the lesion and future treatment strategies.\n\n\nConclusion\n\nThe protocol for immunohistochemical procedures of TLR2, CD38, CD68 and CD83 have been standardized and the expression of each cell marker (TLR2, CD38, CD68 and CD83) within refractory periapical granuloma was identified. Refractory periapical granuloma consistently expressed TLR2 through lymphocytes and plasma cells (CD38+), macrophages and monocytes (CD68+) and mature dendritic cells (CD83+). Lymphocytes and plasma cells appeared to be the dominant inflammatory cells expressing TLR2. Although CD68+ and CD83+ cells may not always be present in a lesion, they did express TLR2 whenever they were identified. TLR2+ dendritic cells play a minor role in periapical granuloma with regards to antigen recognition. The current study confirms previous findings that highlight the significant role of Gram-positive bacteria in clinical cases of refractory periapical granuloma. This was demonstrated through the expression of TLR2 in symptomatic cases.\n\n\nData availability\n\nAll microscopic images are available in the following dataset including microscopic images for positive control and negative control cases. The zip file contains a key for the microscope image files.\n\nF1000Research: Dataset 1. Microscopic images for immunohistochemistry, immunofluorescence and double immunofluorescence TLR2 expression in different inflammatory tissues, https://dx.doi.org/10.5256/f1000research.16678.d22414673", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nSundqvist G, Figdor D: Life as an endodontic pathogen. Endodontic Topics. 2003; 6(1): 3–28. Publisher Full Text\n\nNair PN: On the causes of persistent apical periodontitis: a review. Int Endod J. 2006; 39(4): 249–81. PubMed Abstract \| Publisher Full Text\n\nMedzhitov R, Janeway CA Jr: Innate immunity: impact on the adaptive immune response. Curr Opin Immunol. 1997; 9(1): 4–9. PubMed Abstract \| Publisher Full Text\n\nJaneway CA Jr, Medzhitov R: Innate immune recognition. Annu Rev Immunol. 2002; 20: 197–216. PubMed Abstract \| Publisher Full Text\n\nKumar H, Kawai T, Akira S: Toll-like receptors and innate immunity. Biochem Biophys Res Commun. 2009; 388(4): 621–5. PubMed Abstract \| Publisher Full Text\n\nMale D: Mechanisms of innate immunity. In: David Male JB, David B Roth, Ivan Roitt, ed. Immunology, Seventh edn; 2006; 129: Mosby Elsevier.\n\nSweet MJ, Campbell CC, Sester DP, et al.: Colony-stimulating factor-1 suppresses responses to CpG DNA and expression of toll-like receptor 9 but enhances responses to lipopolysaccharide in murine macrophages. J Immunol. 2002; 168(1): 392–9. PubMed Abstract \| Publisher Full Text\n\nRoger T, David J, Glauser MP, et al.: MIF regulates innate immune responses through modulation of Toll-like receptor 4. Nature. 2001; 414(6866): 920–4. PubMed Abstract \| Publisher Full Text\n\nBosisio D, Polentarutti N, Sironi M, et al.: Stimulation of toll-like receptor 4 expression in human mononuclear phagocytes by interferon-gamma: a molecular basis for priming and synergism with bacterial lipopolysaccharide. Blood. 2002; 99(9): 3427–31. PubMed Abstract \| Publisher Full Text\n\nMatsuguchi T, Musikacharoen T, Ogawa T, et al.: Gene expressions of Toll-like receptor 2, but not Toll-like receptor 4, is induced by LPS and inflammatory cytokines in mouse macrophages. J Immunol. 2000; 165(10): 5767–72. PubMed Abstract \| Publisher Full Text\n\nZarember KA, Godowski PJ: Tissue expression of human Toll-like receptors and differential regulation of Toll-like receptor mRNAs in leukocytes in response to microbes, their products, and cytokines. J Immunol. 2002; 168(2): 554–61. PubMed Abstract \| Publisher Full Text\n\nSupajatura V, Ushio H, Nakao A, et al.: Protective roles of mast cells against enterobacterial infection are mediated by Toll-like receptor 4. J Immunol. 2001; 167(4): 2250–6. PubMed Abstract \| Publisher Full Text\n\nSchwandner R, Dziarski R, Wesche H, et al.: Peptidoglycan- and lipoteichoic acid-induced cell activation is mediated by toll-like receptor 2. J Biol Chem. 1999; 274(25): 17406–9. PubMed Abstract \| Publisher Full Text\n\nAliprantis AO, Yang RB, Mark MR, et al.: Cell activation and apoptosis by bacterial lipoproteins through toll-like receptor-2. Science. 1999; 285(5428): 736–9. PubMed Abstract \| Publisher Full Text\n\nCampos MA, Almeida IC, Takeuchi O, et al.: Activation of Toll-like receptor-2 by glycosylphosphatidylinositol anchors from a protozoan parasite. J Immunol. 2001; 167(1): 416–23. PubMed Abstract \| Publisher Full Text\n\nUnderhill DM, Ozinsky A, Hajjar AM, et al.: The Toll-like receptor 2 is recruited to macrophage phagosomes and discriminates between pathogens. Nature. 1999; 401(6755): 811–5. PubMed Abstract \| Publisher Full Text\n\nBirchler T, Seibl R, Büchner K, et al.: Human Toll-like receptor 2 mediates induction of the antimicrobial peptide human beta-defensin 2 in response to bacterial lipoprotein. Eur J Immunol. 2001; 31(11): 3131–7. PubMed Abstract \| Publisher Full Text\n\nTakeuchi O, Hoshino K, Kawai T, et al.: Differential roles of TLR2 and TLR4 in recognition of gram-negative and gram-positive bacterial cell wall components. Immunity. 1999; 11(4): 443–51. PubMed Abstract \| Publisher Full Text\n\nMutoh N, Tani-Ishii N, Tsukinoki K, et al.: Expression of toll-like receptor 2 and 4 in dental pulp. J Endod. 2007; 33(10): 1183–6. PubMed Abstract \| Publisher Full Text\n\nMutoh N, Watabe H, Chieda K, et al.: Expression of Toll-like receptor 2 and 4 in inflamed pulp in severe combined immunodeficiency mice. J Endod. 2009; 35(7): 975–80. PubMed Abstract \| Publisher Full Text\n\nJiang HW, Zhang W, Ren BP, et al.: Expression of toll like receptor 4 in normal human odontoblasts and dental pulp tissue. J Endod. 2006; 32(8): 747–51. PubMed Abstract \| Publisher Full Text\n\nDurand SH, Flacher V, Roméas A, et al.: Lipoteichoic acid increases TLR and functional chemokine expression while reducing dentin formation in in vitro differentiated human odontoblasts. J Immunol. 2006; 176(5): 2880–7. PubMed Abstract \| Publisher Full Text\n\nDesai S: Expression of toll-like receptors-2 in periapical lesions of endodontic origin (DclinDent thesis). Dunedin: University of Otago, 2009.\n\nChokechanachaisakul U, Kaneko T, Okiji T, et al.: Increased gene expression of Toll-like receptors and antigen-presenting cell-related molecules in the onset of experimentally induced furcation lesions of endodontic origin in rat molars. J Endod. 2010; 36(2): 251–5. PubMed Abstract \| Publisher Full Text\n\nStern MH, Dreizen S, Mackler BF, et al.: Quantitative analysis of cellular composition of human periapical granuloma. J Endod. 1981; 7(3): 117–22. PubMed Abstract \| Publisher Full Text\n\nBergenholtz G, Lekholm U, Liljenberg B, et al.: Morphometric analysis of chronic inflammatory periapical lesions in root-filled teeth. Oral Surg Oral Med Oral Pathol. 1983; 55(3): 295–301. PubMed Abstract \| Publisher Full Text\n\nNilsen R, Johannessen AC, Skaug N, et al.: In situ characterization of mononuclear cells in human dental periapical inflammatory lesions using monoclonal antibodies. Oral Surg Oral Med Oral Pathol. 1984; 58(2): 160–5. PubMed Abstract \| Publisher Full Text\n\nKontiainen S, Ranta H, Lautenschlager I: Cells infiltrating human periapical inflammatory lesions. J Oral Pathol. 1986; 15(10): 544–6. PubMed Abstract \| Publisher Full Text\n\nKopp W, Schwarting R: Differentiation of T lymphocyte subpopulations, macrophages, and HLA-DR-restricted cells of apical granulation tissue. J Endod. 1989; 15(2): 72–5. PubMed Abstract \| Publisher Full Text\n\nPiattelli A, Artese L, Rosini S, et al.: Immune cells in periapical granuloma: Morphological and immunohistochemical characterization. J Endod. 1991; 17(1): 26–9. PubMed Abstract \| Publisher Full Text\n\nLiapatas S, Nakou M, Rontogianni D: Inflammatory infiltrate of chronic periradicular lesions: an immunohistochemical study. Int Endod J. 2003; 36(7): 464–71. PubMed Abstract \| Publisher Full Text\n\nLukić A, Vasilijić S, Majstorović I, et al.: Characterization of antigen-presenting cells in human apical periodontitis lesions by flow cytometry and immunocytochemistry. Int Endod J. 2006; 39(8): 626–36. PubMed Abstract \| Publisher Full Text\n\nLove RM, Firth N: Histopathological profile of surgically removed persistent periapical radiolucent lesions of endodontic origin. Int Endod J. 2009; 42(3): 198–202. PubMed Abstract \| Publisher Full Text\n\nBecconsall-Ryan K, Tong D, Love RM: Radiolucent inflammatory jaw lesions: a twenty-year analysis. Int Endod J. 2010; 43(10): 859–65. PubMed Abstract \| Publisher Full Text\n\nNair PN: Apical periodontitis: a dynamic encounter between root canal infection and host response. Periodontol 2000. 1997; 13(1): 121–48. PubMed Abstract \| Publisher Full Text\n\nPinheiro ET, Gomes BP, Ferraz CC, et al.: Microorganisms from canals of root-filled teeth with periapical lesions. Int Endod J. 2003; 36(1): 1–11. PubMed Abstract \| Publisher Full Text\n\nSiqueira JF Jr, Rôças IN: Diversity of endodontic microbiota revisited. J Dent Res. 2009; 88(11): 969–81. PubMed Abstract \| Publisher Full Text\n\nMolander A, Reit C, Dahlén G, et al.: Microbiological status of root-filled teeth with apical periodontitis. Int Endod J. 1998; 31(1): 1–7. PubMed Abstract \| Publisher Full Text\n\nSundqvist G, Figdor D, Persson S, et al.: Microbiologic analysis of teeth with failed endodontic treatment and the outcome of conservative re-treatment. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 1998; 85(1): 86–93. PubMed Abstract \| Publisher Full Text\n\nHancock HH 3rd, Sigurdsson A, Trope M, et al.: Bacteria isolated after unsuccessful endodontic treatment in a North American population. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2001; 91(5): 579–86. PubMed Abstract \| Publisher Full Text\n\nSiqueira JF Jr, Rôças IN: Polymerase chain reaction-based analysis of microorganisms associated with failed endodontic treatment. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2004; 97(1): 85–94. PubMed Abstract \| Publisher Full Text\n\nGewirtz AT, Navas TA, Lyons S, et al.: Cutting edge: bacterial flagellin activates basolaterally expressed TLR5 to induce epithelial proinflammatory gene expression. J Immunol. 2001; 167(4): 1882–5. PubMed Abstract \| Publisher Full Text\n\nAbreu MT, Vora P, Faure E, et al.: Decreased expression of Toll-like receptor-4 and MD-2 correlates with intestinal epithelial cell protection against dysregulated proinflammatory gene expression in response to bacterial lipopolysaccharide. J Immunol. 2001; 167(3): 1609–16. PubMed Abstract \| Publisher Full Text\n\nCario E, Podolsky DK: Differential alteration in intestinal epithelial cell expression of toll-like receptor 3 (TLR3) and TLR4 in inflammatory bowel disease. Infect Immun. 2000; 68(12): 7010–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOzinsky A, Underhill DM, Fontenot JD, et al.: The repertoire for pattern recognition of pathogens by the innate immune system is defined by cooperation between toll-like receptors. Proc Natl Acad Sci U S A. 2000; 97(25): 13766–71. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTakeuchi O, Kawai T, Mühlradt PF, et al.: Discrimination of bacterial lipoproteins by Toll-like receptor 6. Int Immunol. 2001; 13(7): 933–40. PubMed Abstract \| Publisher Full Text\n\nTakeuchi O, Sato S, Horiuchi T, et al.: Cutting edge: role of Toll-like receptor 1 in mediating immune response to microbial lipoproteins. J Immunol. 2002; 169(1): 10–4. PubMed Abstract \| Publisher Full Text\n\nTakeda K, Kaisho T, Akira S, et al.: Toll-like receptors. Annu Rev Immunol. 2003; 21: 335–76. PubMed Abstract \| Publisher Full Text\n\nTrope M: Root Resorption due to Dental Trauma. Endod Topics. 2002; 1(1): 79–100. Publisher Full Text\n\nHeithersay GS: Invasive cervical resorption. Endod Topics. 2004; 7(1): 73–92. Publisher Full Text\n\nLin YP, Love RM, Friedlander LT, et al.: Expression of Toll-like receptors 2 and 4 and the OPG-RANKL-RANK system in inflammatory external root resorption and external cervical resorption. Int Endod J. 2013; 46(10): 971–981. PubMed Abstract \| Publisher Full Text\n\nDesai SV, Love RM, Rich AM, et al.: Toll-like receptor 2 expression in refractory periapical lesions. Int Endod J. 2011; 44(10): 907–916. PubMed Abstract \| Publisher Full Text\n\nLin SK, Kok SH, Lin LD, et al.: Nitric oxide promotes the progression of periapical lesion via inducing macrophage and osteoblast apoptosis. Oral Microbiol Immunol. 2007; 22(1): 24–9. PubMed Abstract \| Publisher Full Text\n\nMarcato LG, Ferlini AP, Bonfim RC, et al.: The role of Toll-like receptors 2 and 4 on reactive oxygen species and nitric oxide production by macrophage cells stimulated with root canal pathogens. Oral Microbiol Immunol. 2008; 23(5): 353–9. PubMed Abstract \| Publisher Full Text\n\nNakane PK: Simultaneous localization of multiple tissue antigens using the peroxidase-labeled antibody method: a study on pituitary glands of the rat. J Histochem Cytochem. 1968; 16(9): 557–60. PubMed Abstract \| Publisher Full Text\n\nWang BL, Larsson LI: Simultaneous demonstration of multiple antigens by indirect immunofluorescence or immunogold staining. Novel light and electron microscopical double and triple staining method employing primary antibodies from the same species. Histochemistry. 1985; 83(1): 47–56. PubMed Abstract \| Publisher Full Text\n\nMason DY, Micklem K, Jones M: Double immunofluorescence labelling of routinely processed paraffin sections. J Pathol. 2000; 191(4): 452–61. PubMed Abstract \| Publisher Full Text\n\nMiller K: Immunocytochemical techniques. In: Bancroft JD, Gamble M, eds. Theory and Practice of Histological Techniques, 5th edn; 2002; 421–64. London: Churchill Livingstone.\n\nDesai SV, Love RM, Rich AM, et al.: Toll-like receptor 2 expression in refractory periapical lesions. Int Endod J. 2011; 44(10): 907–16. PubMed Abstract \| Publisher Full Text\n\nHayashi F, Means TK, Luster AD: Toll-like receptors stimulate human neutrophil function. Blood. 2003; 102(7): 2660–9. PubMed Abstract \| Publisher Full Text\n\nMuirhead KA, Horan PK, Poste G: Flow Cytometry: Present and Future. Nat Biotech. 1985; 3: 337–56. Publisher Full Text\n\nDeaglio S, Aydin S, Vaisitti T, et al.: CD38 at the junction between prognostic marker and therapeutic target. Trends Mol Med. 2008; 14(5): 210–8. PubMed Abstract \| Publisher Full Text\n\nMalavasi F, Deaglio S, Funaro A, et al.: Evolution and Function of the ADP Ribosyl Cyclase/CD38 Gene Family in Physiology and Pathology. Physiol Rev. 2008; 88(3): 841–86. PubMed Abstract \| Publisher Full Text\n\nParwaresch MR, Radzun HJ, Kreipe H, et al.: Monocyte/macrophage-reactive monoclonal antibody Ki-M6 recognizes an intracytoplasmic antigen. Am J Pathol. 1986; 125(1): 141–51. PubMed Abstract \| Free Full Text\n\nMicklem K, Rigney E, Cordell J, et al.: A human macrophage-associated antigen (CD68) detected by six different monoclonal antibodies. Br J Haematol. 1989; 73(1): 6–11. PubMed Abstract \| Publisher Full Text\n\nJohannessen AC, Nilsen R, Skaug N: Enzyme histochemical characterization of mononuclear cells in human dental periapical chronic inflammatory lesions. Scand J Dent Res. 1984; 92(4): 325–33. PubMed Abstract \| Publisher Full Text\n\nMatsuo T, Ebisu S, Shimabukuro Y, et al.: Quantitative analysis of immunocompetent cells in human periapical lesions: Correlations with clinical findings of the involved teeth. J Endod. 1992; 18(10): 497–500. PubMed Abstract \| Publisher Full Text\n\nLechmann M, Berchtold S, Steinkasserer A, et al.: CD83 on dendritic cells: more than just a marker for maturation. Trends Immunol. 2002; 23(6): 273–5. PubMed Abstract \| Publisher Full Text\n\nCao W, Lee SH, Lu J: CD83 is preformed inside monocytes, macrophages and dendritic cells, but it is only stably expressed on activated dendritic cells. Biochem J. 2005; 385(Pt 1): 85–93. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiu K, Nussenzweig MC: Origin and development of dendritic cells. Immunol Rev. 2010; 234(1): 45–54. PubMed Abstract \| Publisher Full Text\n\nBekeredjian-Ding I, Inamura S, Giese T, et al.: Staphylococcus aureus protein A triggers T cell-independent B cell proliferation by sensitizing B cells for TLR2 ligands. J Immunol. 2007; 178(5): 2803–12. PubMed Abstract \| Publisher Full Text\n\nChiron D, Bekeredjian-Ding I, Pellat-Deceunynck C, et al.: Toll-like receptors: lessons to learn from normal and malignant human B cells. Blood. 2008; 112(6): 2205–13. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChen E, Bakr MM, Firth N, et al.: Dataset 1 in: Inflammatory cell expression of Toll-like receptor-2 (TLR2) within refractory periapical granuloma. F1000Research. 2018. https://dx.doi.org/10.5256/f1000research.16678.d224146" }	[ { "id": "40928", "date": "26 Nov 2018", "name": "Paul Vincent Abbott", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper has addressed a particular clinical problem that is often found in specialist Endodontic practices - i.e. refractory cases that do not heal and show persistent radiolucency despite all the best clinical endodontic procedures and various anti-bacterial strategies being used. The authors have investigated the role of Toll-like receptor-2 (TLR-2) in these refractory periapical lesions. They showed that lymphocytes and plasma cells from these lesions have high expression levels of TLR-2 which indicates their role in these cases. This information has the potential to lead to better diagnostic and management strategies to identify these refractory cases earlier.\nThis is a well-conducted study that has utilized samples of refractory cases from the university’s collection of cases. The methodology involved numerous tests and analysis of the biopsy material along with positive and negative controls to support the methodology used. The paper has comprehensive descriptions and histological material to support the discussion and conclusions. The results support previous studies.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "40929", "date": "30 Nov 2018", "name": "Lakshmi Thangavelu", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this manuscript Bakr et al. gave promising evidence about the protocol employed in immunohistochemical analysis of various immunomarkers. This research provides brief information about the role of Toll-like receptors and their applications in clinical practice for diagnosis. I was curious to know about the sample size selected in the study - when I went through the criteria for selecting the sample size was found to be satisfactory. The introduction part of the article provides lots of information about the Toll-like receptors’ role in the field of endodontics. The authors highlighted the immunohistochemical analysis well in the manuscript.\n\nMinor comments:\n\nThe introduction is too elaborate. I would appreciate it if the key points are focused on.\n\nEven though the sample size utilised is 8 for qualitative analysis out of 772 samples, this needs further explanation.\n\nThe statistical analysis is not incorporated in the study to show the significance.\n\nOverall the authors have done excellent work. The authors have provided a brief insight about the Toll-like receptors and refractory periapical lesions.\n\nIf the points raised above will be addressed, indexing is appropriate for this manuscript.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1819
https://f1000research.com/articles/7-1212/v1	07 Aug 18	{ "type": "Research Article", "title": "Effectiveness of implementation of “mental health nursing students’ clinical competency model” on academic performance of nursing students", "authors": [ "Foroozan Atashzadeh-Shoorideh", "Jamileh Mohtashami", "Seyed Amir Hosein Pishgooie", "Tayebeh Jamshidi", "Sara Sedghi", "Foroozan Atashzadeh-Shoorideh", "Seyed Amir Hosein Pishgooie", "Tayebeh Jamshidi", "Sara Sedghi" ], "abstract": "Background: Clinical nursing competence in mental health is one of the most important topics in theoretical and practical nursing training with many factors affecting it. The purpose of this study is to determine the impact of the implementation of the “mental health nursing students’ clinical competence model” on nursing students’ academic performance. Methods: This study is a semi experimental following one group of student nurses. “mental health nursing students’ clinical competence model” for undergraduate nursing student’s education was applied. The study population included 50 nursing students, who were studying from fifth semester to seventh semester and selected through census sampling. During the seventh semester after the completion of theoretical and practical courses in mental health nursing, re-evaluation was conducted and the scores before and after the implementation of the clinical competence model were compared. Results: Rate of clinical competency before the intervention, was estimated at the level of non-mastered; and after intervention was at the level of mastered, demonstrating a significant difference (p<0.001). Areas of clinical competency scores before and after the intervention were compared which showed significant difference in all the areas except the mental competency areas (p<0.05). Conclusions: The implementation of the “mental health nursing students’ clinical competence model” and appropriate planning for achievement of mental health nursing specialized competency can ensure the achievement of clinical competency by nursing students.", "keywords": [ "Clinical competency", "Nursing", "pattern", "academic performance" ], "content": "Introduction\n\nThe purpose of nursing education is to transfer knowledge and help students, gain insights and skills necessary for nursing care1. One of the important aims of universities and higher education centers of medical science is building capacity and skills in students, as well as to prepare them for health services and to provide health care to all members of the community2–4. A survey of one study of nursing education courses showed that abilities acquired by students were far from optimal, and they have not gained the skills and abilities necessary at the end of their training5. It seems, education plays an important role in the development of professional nursing skills and it provides opportunities to gain a wide range of knowledge, problem solving abilities and critical thinking6. Previous studies on novice nurses also suggest that in the transition from student to professional role, students have experienced a lack of preparation. Stressful experiences in novice nurses during this period are often associated with a lack of necessary skills in nursing practices and the lack of comparability between undergraduate education in universities and the situation in the workplace7. However, Manoochehri et al.8,9, concluded that student employment after graduation can help.\n\nNurses' knowledge and competence are based on their knowledge and the curriculum that is taught in universities. The training program is very important in gaining nursing values and achieving educational goals10. This knowledge and the skills acquired are valuable, and have a direct effect on the student’s future career, and immediately after graduation when they start work11,12.\n\nNursing education programs with an emphasis on skill development provide an opportunity to increase student competence and adaptability to meet the clinical needs of novice nurses entering the workplace13. Some studies have shown that there is a direct relationship between the level of clinical competence and the ability to apply their skills. In other words, a nurse who has higher competency, can take advantage of their skills in clinical practice14. Studies have shown that novice nurses are not adequately prepared to deal with challenges in the work place15–18, nursing undergraduate curriculum cannot prepare students adequately for independent performance19, and therefore they may have limited technological skills20. One study found that nearly 49 percent of newly graduated nurses were involved in errors, indicating the gap between meeting minimum standards of practice and professional competence21.\n\nCompetence is a complex and ambiguous concept and one of the more controversial issues in the nursing standards in various fields including education, clinical and management22. Competencies comprise of different aspects of learning including knowledge, skills and attitudes. According to this definition, people should be able to fulfill their role or set of tasks to an appropriate level5. The Australian Nursing and Midwifery Council (2005) defines competence as a combination of skills, knowledge, attitudes, values and abilities that underlies effective and high professional performance in a professional career area23,24.\n\nTherefore, nurse educators are trying to design competency-based training programs. Competencies are built on scientific knowledge, but their development requires several activities, the most significant of which is the application of theoretical content to real life situations25. In the curriculum and teaching-learning processes where a competency-based approach is used the program is based on a set of skills that every student should master26,27.\n\nIn reviews of competency in nursing, there are challenges which reflect the difficulties and complexity of the profession. Mental health nursing for various reasons, including lack of clarity about roles within mental health nursing and lack of standards for mental health care, is faced with many challenges. The Nurses Association of America defines psychiatric nursing (Mental Health Nursing) as the diagnosis and treatment of human responses to actual and potential mental health problems. The psychiatric mental health nursing identifies with aspects of care such as communication28. Mental health nurses are nurse practitioners who show competencies, knowledge, skills and special abilities to care for people with mental health problems and mental disorders29. The nature of mental health nursing, like all areas of care, is undergoing profound changes. These changes include dealing with an aging population, increase in variation in cultures, case management, changes in care situation (such as from hospital to community), competition in care patterns, maintaining employment opportunities, providing the necessary training for career development, new science and information technology, and finally the work of mental health nursing practice30,31.\n\nIn relation to the lack of adequate mental health content in nursing curricula prior to employment, there are concerns which led it being alleged that new graduate nurses were not adequately prepared to care for patients with mental disorders. These concerns could have significant effects on the standards of care provided32,33. In a study that was conducted by Melnyck and colleagues at Yale University, officials believed that nearly half of new graduates were impaired when it came to providing comprehensive advice, especially to families at risk34. There are also growing concerns about mental health nurses, in that they have not been adequately trained in medication management. This problem is even present in countries such as Britain, which since 2003 has given the right to prescribe medication to mental health nurses35.\n\nIn Iran, the necessity of clinical competence has become more of an important issue in recent years, as communities now expect a higher quality of service, forcing health care system to increase the effectiveness of their staff22. Despite the extraordinary importance of the topic, understanding of nursing clinical competency and skill level is limited and little research has been done in this area14. Clinical competence requires a framework that includes access to all essential competencies during students' education. This model means that all the requirements of clinical competence of nursing students needs be considered. Planning a curriculum with respect to all issues of clinical competence may be possible. A model for “mental health nursing students' clinical competence” is a guide which provides a framework for achieving educational goals, performing evaluations and also allowing students to develop appropriate professional experience. Such a model helps establish and provide clinical competency-based planning. The aim of this study was to determine the impact of the implementation of ‘mental health nursing students’ clinical competence model’ on nursing students’ academic performance.\n\n\nMethods\n\nThis study is a semi experimental study conducted from January 2014 to January 2017. The sample was one group of undergraduate nursing students followed from fifth till their seventh semester. During three semester (18 months) students were taught based on the ‘mental health nursing students’ clinical competence model’36 in theoretical and clinical programs. The study population included 50 nursing students of Shahid Beheshti University of Medical Sciences, who were been selected through census sampling.\n\nFirst, using the \"mental health nursing students' clinical competencies” checklist (Supplementary File 1), designed in 2014 by Mohtashami et al.37 pre-tests were conducted among 5th semester nursing students who were taking the mental health nursing course. The checklist includes 73 statements which assess clinical competencies in two areas of general competency (emotional competency, 7 statements; moral competency, 11 statement; general nursing skill competency, 7 statements) and specific competencies (therapeutic communication competency, 12 statements and caring for mentally ill patient skill competency, 36 statements) using a Likert scale of 5 from “always” to “never”. Clinical competency were divided into 4 categories of weak (scores from 73 to 146), Average (scores from 146 to 219), good (scores from 219 to 292) and Excellent (scores from 292 to 365). Regarding to determination of mastery, we considered a score of 300 as the cutoff point, with scores below 300 indicating lack of mastery and scores above 300 showing competencies having been mastered. Reliability coefficient was conducted through internal consistency reliability (Cronbach's alpha 0.93).\n\nAfter the pre-test, the ‘mental health nursing students’ clinical competence model’36 was conducted. The model has 4 related phases which started from a mental health nursing course (fifth semester) to apprentice in clinical field (seventh semester). In each phrase, every student is required to gain the knowledge, skills and special abilities of that phase. When passing through each phase, competency-based assessments were conducted. In the fifth semester where all students are studying mental health nursing, implementation of these models began. Later in the 6th semester, psychiatric disorders and mental health nursing course were completed. Internships were also undertaken in this semester. Finally the intervention was completed in the 7th semester after the apprenticeship in clinical field. The aim of this model is to achieve clinical competency through a systematic and scientific process and continuously be developed these competencies. Model dimensions were considered in four domains of orientation and preparation, confrontation, involvement and achieving clinical competency (Figure 1). A breakdown of teaching aims for each domain is provided in Supplementary File 2.\n\nIn all domains of the model, cognitive/emotional abilities and specific skills of instructors and students will have a certain impact on clinical competency achievement. It should be noted that while transitioning from one stage to the next one, students must be evaluated to ensure achievement of required competencies of that stage.\n\nUpon completion of the theoretical and clinical mental health nursing courses in the 7th semester, re-evaluation was performed; and scores of before and after the implementation of the clinical competency model were compared.\n\nThe study was approved by the Ethics Committee of the School of Nursing in July 2015 with approval number SBMU2.REC.2015.6. All the ethical considerations were taken into account. All the participants signed a consent form before participating in the study.\n\nIn this study, data analysis was conducted through the use of SPSS software-version 21. First, to describe the demographic characteristics of subjects, review frequency distribution, mean and standard deviation using descriptive statistical tests, were used. In the following, Paired t-test analysis and RM-ANOVA, the nonparametric equivalents of it depending on the type and distribution of data, were used. The significance level for all tests was set at 0.05.\n\n\nResults\n\n50 nursing students were enrolled in this study. Most (90%), were aged 21 to 25 years old (N=45) 62 percent (N=31) were women and 38 percent (N=19) were male. Statistical analysis of the relationship between age and gender and the clinical competencies did not find any significant relationship. Average scores in different domains of clinical competency before and after the implementation of competence model are shown in Table 1. Rate of clinical competency prior to the implementation was estimated in the level of the non-mastered and after the implementation was in the level of mastered, a statistically significant difference (Table 1). Clinical competency areas using paired t-test scores were compared before and after the intervention. All scores before and after treatment in all areas except the area of mental competencies, showed a significant difference (p <0.05).\n\n\nDiscussion\n\nClinical competence requires a framework that includes all aspects of access to essential competencies during students' education. This model means that all the requirements of clinical competence of nursing students need to be considered, and appropriate curriculum with respect to all issues related to clinical competence is useful. Existence of a model for mental health nursing students' clinical competence, in addition to determine the achievement of educational goals, provides the possibility to evaluate and gain feedback; it also provides the opportunity for reformation of professional practice for nursing students. Such a model helps establish and provide clinical competency-based planning. The aim of this study was to evaluate the effects of the ‘mental health nursing students’ clinical competence model’ on nursing students’ academic performance.\n\nThis research findings showed that rate of clinical competency before the intervention, was estimated at the level of non-mastered; and after intervention was at the level of mastered, this difference was statistically significant. Wangensteen et al.38 in relation to newly graduated nurses' perception of competence and possible predictors also found the same results. In line with the current study, Safadi et al.39 conducted a cross-sectional study to reviews competencies of nursing graduates in universities in Jordan. The results were rated as satisfactory clinical competence in line with this research Mohtashami et al.40 in the article appearing with the aim of determining the relationship between professional competence and professional identity, concluded that there is a positive correlation between professional competencies and professional identity.\n\nBased on these findings and the researcher’s opinions, the development of the critical thinking abilities of nursing students during nursing school is crucial to strengthen novice’s clinical competence. Accordingly, the authors acknowledged that student-centered learning models such as problem-based learning are significantly associated with the critical thinking. Therefore, we identify a need for collaboration between all of the nursing educators for clinical training of students and using the student-centered models. But the results of a cross-sectional study by Salonen et al.41 in relation to the clinical competency of Finnish novice nurse’s and factors affecting it showed, fair to good level of clinical competence which was different from the findings of the current study. Reasons for these differences may include using a different instrument, the study samples and that the nurse’s had a higher level of experiences. Phillips et al.42 in this regard reported that an increase in level of clinical experience and more hours in school and placement in the clinical setting, especially in nursing education programs, promotes decision-making skills, more professional performance in individuals and gaining teamwork experience in the real working environment.\n\nIn a study by Sabancıogullari & Dogan43 the effects of the professional identity development program on the professional identity, job satisfaction and burnout levels of registered nurses was conducted on 63 nurses working in a university hospital. The program of 10 sessions (once a week, and follow-up 6 months later) improved the professional identity of nurses in the intervention group compared to the control group which the difference was statistically significant. During the study period burnout among nurses in the intervention group decreased, but increased in the control group. But there was no statistically significant difference between the intervention group and the control group in terms of job satisfaction.\n\nMohtashami et al.44 conducted a qualitative study aimed to clarify the concept and how to achieve clinical competencies in mental health nursing students, concluding that nursing students in undergraduate education in order to gain mental health nursing competencies, must pass special stages before being be able to work. During each stage there are indicators that can help students to acquire the competencies. Therefore changes in the curriculum and students training methods should be considered. In support of this claim, Mohtashami et al.45 wrote a competency-based curriculum to facilitate the teaching-learning process is one of the first steps. Revision of curriculum can be used to reduce gap between theory and practice so competencies can be acquired effectively.\n\nIn a literature review, the researchers could not find similar studies, therefore it can be argued this study aiming to influence the implementation of clinical competence of nursing students' mental health and academic performance is in itself an innovation. On the other hand no similar study was found to allow more comprehensive discussion possibility and that can be considered a limitation of this study.\n\n\nImplications for practice\n\nThe implementation of the ‘mental health nursing students’ clinical competence model’ and appropriate planning for achievement of mental health nursing specialized competency can ensure the achievement of clinical competency by nursing students. Such a model helps establish and provide clinical competency-based planning.\n\n\nData availability\n\nDataset 1: Test scores pre and post intervention with demographic information 10.5256/f1000research.14284.d20707546", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis project has received funding from Shahid Beheshti University of Medical Sciences with project number 5561.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgement\n\nThe authors gratitude extends to the authorities of the university and the School of Nursing and Midwifery, nursing teachers, nurses and psychiatric nurses and students, this study was not possible without their cooperation. Also, our appreciation goes to Mrs. Mahdieh Ghalenoee for her valuable suggestions.\n\n\nSupplementary material\n\nSupplementary File 1: \"mental health nursing students' clinical competencies” checklist.\n\nClick here to access the data.\n\nSupplementary File 2: File contain expansion of study design including detailed explanation of teaching aims for each domain.\n\nClick here to access the data.\n\n\nReferences\n\nSabeti F, Akbari-nassaji N, Haghighy-zadeh MH: Nursing students' self-assessment regarding clinical skills achievement in Ahvaz Jundishapur University of Medical Sciences (2009). Iranian Journal of Medical Education. 2011; 11(5): 506–515. Reference Source\n\nAmini A, Hasanzadeh Salmasi S, Shaghaghi A, et al.: Effect of clinical skills training necessary in labour on medicine student’s clinical competency of Tabriz Medical University. Iranian Journal of Education in Medical Sciences. 2005; 5(1): 8–12.\n\nGidman J, McIntosh A, Melling K, et al.: Student perceptions of support in practice. Nurse Educ Pract. 2011; 11(6): 351–5. PubMed Abstract \| Publisher Full Text\n\nMohammadi F, Hosseini MA: Rehabilitation Sciences Students' Perception from Clinical Self-Efficacy Compared to Evaluation by Clinical Teachers. Iranian Journal of Medical Education. 2010; 10(2): 155–63. Reference Source\n\nParsa YZ, Ramezani BF, Khatuni AR: Nursing students' viewpoints about their clinical competencies and its achievement level. Iranian Journal of Nursing Research. 2007; 1(3): 7–14. Reference Source\n\nDias JM, Ajani K, Mithani Y: Conceptualization and operationalization of a baccalaureate nursing curriculum in Pakistan: Challenges; hurdles and lessons learnt. Procedia-Social and Behavioral Sciences. 2010; 2(2): 2335–7. Publisher Full Text\n\nHeshmati Nabavi F, Vanaki Z: Effective clinical instructor: A qualitative study. Iranian Journal of Nursing Research. 2009; 4(12–13): 39–53. Reference Source\n\nManoochehri H, Imani E, Atashzadeh-Shoorideh F, et al.: Competence of novice nurses: role of clinical work during studying. J Med Life. 2015; 8(Spec Iss 4): 32–38. PubMed Abstract \| Free Full Text\n\nManouchehri H, Imani E, Atashzadeh-Shoorideh F, et al.: Experience of Work While Studying: Novice Nurses Entering the Clinical Arena. Research Journal of Medical Sciences. 2016; 10(5): 465–74. Reference Source\n\nNajafi Kolyani M, Sharif F, Jamshidi N, et al.: Students' perceptions of effective teaching in nursing education: a qualitative study. Iranian Journal of Nursing Research. 2011. 5(19): 6–15. Reference Source\n\nShayan S: Using Patient Management Problem (EPMP) in Assessment of Clinical Competency. Iranian Journal of Medical Education. 2011; 10(5): 1087–1092. Reference Source\n\nMojab F, Zaefarian R, Azizi AHD: Applying competency based approach for entrepreneurship education. Procedia-Social and Behavioral Sciences. 2011; 12: 436–47. Publisher Full Text\n\nShin KR, Jung D, Kim MW, et al.: Clinical supervisors' satisfaction with the clinical competence of newly employed nurses in Korea. Nurs Outlook. 2010; 58(3): 129–34. PubMed Abstract \| Publisher Full Text\n\nMahreini M, Moatary M, Akaberian S, et al.: Determining nurses' clinical competence in hospitals of Bushehr University of Medical Sciences by self assessment method. Iran South Med J. 2008; 11(1): 69–75. Reference Source\n\nMcCarthy B: Translating person-centred care: a case study of preceptor nurses and their teaching practices in acute care areas. J Clin Nurs. 2006; 15(5): 629–38. PubMed Abstract \| Publisher Full Text\n\nMooney M: Newly qualified Irish nurses’ interpretation of their preparation and experiences of registration. J Clin Nurs. 2007; 16(9): 1610–7. PubMed Abstract \| Publisher Full Text\n\nMooney M: Facing registration: the expectations and the unexpected. Nurse Educ Today. 2007; 27(8): 840–7. PubMed Abstract \| Publisher Full Text\n\nO'shea M, Kelly B: The lived experiences of newly qualified nurses on clinical placement during the first six months following registration in the Republic of Ireland. J Clin Nurs. 2007; 16(8): 1534–42. PubMed Abstract \| Publisher Full Text\n\nNewton JM, McKenna L: The transitional journey through the graduate year: a focus group study. Int J Nurs Stud. 2007; 44(7): 1231–7. PubMed Abstract \| Publisher Full Text\n\nOermann MH, Garvin MF: Stresses and challenges for new graduates in hospitals. Nurse Educ Today. 2002; 22(3): 225–30. PubMed Abstract \| Publisher Full Text\n\nKlein CJ, Fowles ER: An investigation of nursing competence and the competency outcomes performance assessment curricular approach: senior students' self-reported perceptions. J Prof Nurs. 2009; 25(2): 109–21. PubMed Abstract \| Publisher Full Text\n\nNesami M, Rafiee F, Parvizi S, et al.: Concept analysis of competency in nursing: Qualitative research. J Mazandaran Univ Med Sci. 2008; 18(67): 35–42. Reference Source\n\nHanley E, Higgins A: Asssessment of practice in intensive care: students’ perceptions of a clinical competence assessment tool. Intensive Crit Care Nurs. 2005; 21(5): 276–83. PubMed Abstract \| Publisher Full Text\n\nLevett-Jones T, Gersbach J, Arthur C, et al.: Implementing a clinical competency assessment model that promotes critical reflection and ensures nursing graduates’ readiness for professional practice. Nurse Educ Pract. 2011; 11(1): 64–9. PubMed Abstract \| Publisher Full Text\n\nDelaney KR, Carlson-Sabelli L, Shephard R, et al.: Competency-based training to create the 21st century mental health workforce: strides, stumbles, and solutions. Arch Psychiatr Nurs. 2011; 25(4): 225–34. PubMed Abstract \| Publisher Full Text\n\nApplin H, Williams B, Day R, et al.: A comparison of competencies between problem-based learning and non-problem-based graduate nurses. Nurse Educ Today. 2011; 31(2): 129–34. PubMed Abstract \| Publisher Full Text\n\nLeCuyer E, DeSocio J, Brody M, et al.: From objectives to competencies: operationalizing the NONPF PMHNP competencies for use in a graduate curriculum. Arch Psychiatr Nurs. 2009; 23(3): 185–99. PubMed Abstract \| Publisher Full Text\n\nMohtashami J, Noughani F: Psychiatric nursing. Teimoorzadeh Co, Tehran. 2011.\n\nBoyd MA: Psychiatric nursing, contemporary practice. Edition 5, editor. Philadelphia.: Wolters Kluwer/Lippincott Williams & Wilkins; 2012. Reference Source\n\nGass J, McKie A, Smith I, et al.: An examination of the scope and purpose of education in mental health nursing. Nurse Educ Today. 2007; 27(6): 588–96. PubMed Abstract \| Publisher Full Text\n\nVarcarolis EM, Carson VB, Shoemaker NC: Foundations of Psychiatric Mental Health Nursing: A clinical approach. Saunders Elsevier, St. Louis, MO. 2006. Reference Source\n\nHappell B: Moving in circles: a brief history of reports and inquiries relating to mental health content in undergraduate nursing curricula. Nurse Educ Today. 2010; 30(7): 643–8. PubMed Abstract \| Publisher Full Text\n\nMcCann TV, Moxham L, Farrell G, et al.: Mental health content of Australian pre-registration nursing curricula: summary report and critical commentary. Nurse Educ Today. 2010; 30(5): 393–7. PubMed Abstract \| Publisher Full Text\n\nMelnyk BM, Hawkins-Walsh E, Beauchesne M, et al.: Strengthening PNP curricula in mental/behavioral health and evidence-based practice. J Pediatr Health Care. 2010; 24(2): 81–94. PubMed Abstract \| Publisher Full Text\n\nSnowden A: Integrating medicines management into mental health nursing in UK. Arch Psychiatr Nurs. 2010; 24(3): 178–88. PubMed Abstract \| Publisher Full Text\n\nMohtashami J, Pazargadi M, Salsali M, et al.: Developing mental health nursing students’ clinical competency model. International Journal of Physical and Social Sciences. 2014; 4(11): 7–22. Reference Source\n\nMohtashami J, Salsali M, Pazargadi M, et al.: Developing and Psychometric Properties Check List of Clinical Competency in Mental Health Nursing Students. Iranian Journal of Psychiatric Nursing. 2014; 2(3): 46–57. Reference Source\n\nWangensteen S, Johansson IS, Björkström ME, et al.: Newly graduated nurses' perception of competence and possible predictors: a cross-sectional survey. J Prof Nurs. 2012; 28(3): 170–81. PubMed Abstract \| Publisher Full Text\n\nSafadi R, Jaradeh M, Bandak A, et al.: Competence assessment of nursing graduates of Jordanian universities. Nurs Health Sci. 2010; 12(2): 147–54. PubMed Abstract \| Publisher Full Text\n\nMohtashami J, Rahnama H, Farzinfard F, et al.: A Survey of Correlation between Professional Identity and Clinical Competency of Psychiatric Nurses. Open J Nurs. 2015; 5(9): 765–72. Publisher Full Text\n\nSalonen AH, Kaunonen M, Meretoja R, et al.: Competence profiles of recently registered nurses working in intensive and emergency settings. J Nurs Manag. 2007; 15(8): 792–800. PubMed Abstract \| Publisher Full Text\n\nPhillips C, Kenny A, Smith C, et al.: Pre-registration paid employment choice: the views of newly qualified nurses. Nurse Educ Today. 2012; 32(1): 10–4. PubMed Abstract \| Publisher Full Text\n\nSabancıogullari S, Dogan S: Effects of the professional identity development programme on the professional identity, job satisfaction and burnout levels of nurses: A pilot study. Int J Nurs Pract. 2015; 21(6): 847–57. PubMed Abstract \| Publisher Full Text\n\nMohtashami J, Salsali M, Pazargadi M, et al.: Clinical Competency in Psychiatric Nursing Students: A Qualitative Study. J Qual Res Health Sci. 2013; 2(3): 261–76. Reference Source\n\nMohtashami J, Salsali M, Pazargadi M, et al.: Competency-based curriculum education in mental health nursing. Open J Nurs. 2013; 3(8): 545–551, 40835. Publisher Full Text\n\nAtashzadeh-Shoorideh F, Mohtashami J, Pishgooie SAH, et al.: Dataset 1 in: Effectiveness of implementation of “mental health nursing students’ clinical competency model” on academic performance of nursing students. F1000Research. 2018. Data Source" }	[ { "id": "36903", "date": "12 Sep 2018", "name": "Fataneh Ghadirian", "expertise": [ "Reviewer Expertise Psychiatry", "Nursing" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe study’s subject is one of the most important issues for mental health nursing in the world and especially in Iran and their authors clearly and accurately presented its background and rationales.\nThe work is technically sound with sufficient details. It was better to have a brief history of used checklist due to varied competency indicators all over the world but this is not so serious.\nThere are sufficient details of statistical analysis and the datasets are clearly described. Although there is a sufficient discussion included the necessary details but the study has no final conclusion statement.\n\nFinally, the study is valuable and approved.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "38804", "date": "08 Oct 2018", "name": "Julie Bradshaw", "expertise": [], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThere is a fundamental flaw in the research question - clinical competence improving academic performance. Clinical competence was improved, not necessarily academic performance. No other intervention was tested. It is not known whether this specific intervention improved competence specifically. We are left wondering, isn't any education on mental health nursing likely to improve competence? There seemed to be a number of assumptions made in the introduction. It seemed that the authors were mixing up clinical competency and critical thinking. This is obviously an international study. I would have liked to see an argument built around the deficits in nursing education in this country and why this tool was necessary.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [ { "c_id": "4177", "date": "17 Nov 2018", "name": "Jamileh Mohtashami", "role": "Author Response", "response": "Thanks for your patience. I would like to say we corrected some parts of article according to reviewer Julie Bradshaw: In introduction we explained about competency in my country. We corrected some parts in methodology and statistical analysis. Moreover we deleted one paragraph and added some articles for adequately support in conclusion. Finally, the references are corrected." } ] } ]	1	https://f1000research.com/articles/7-1212
https://f1000research.com/articles/7-1799/v1	15 Nov 18	{ "type": "Research Article", "title": "Antiasthmatic effect of Curcuma aeruginosa extract on isolated organ of the trachea", "authors": [ "Swandari Paramita", "Emil Bachtiar Moerad", "Sjarif Ismail", "Eva Marliana", "Emil Bachtiar Moerad", "Sjarif Ismail", "Eva Marliana" ], "abstract": "Background: Asthma is a major health problem worldwide. Antiasthma drugs have side effects and can be expensive. It is important to develop antiasthma drugs from medicinal plants that have fewer side effects and are cheaper. One of the medicinal plants used for antiasthma treatment comes from Curcuma aeruginosa (Zingiberaceae family). The aim of the research is to examine spasmolytic activity of ethanol extract of C. aeruginosa on isolated guinea pig tracheas to determine the antiasthma effects. Methods: The spasmolytic activity of C. aeruginosa extracts was tested in separated organs of guinea pig trachea. Guinea pig was sacrificed and its trachea rings were suspended in L-shaped wire loops in organ baths containing the Krebs solution aerated with carbogen. Isometric contractions of tracheal rings were measured by the transducer coupled to the amplifier. The trachea rings were exposed to DMSO as negative control, aminophylline as positive control and C. aeruginosa extracts. The single concentration-relaxation curve was obtained in every preparation. Results: The result showed that the decrease of the spasmolytic activity in the guinea pig tracheal tone due to C. aeruginosa extract was significantly better (p=0.022) when compared to the negative control. Meanwhile, the EC50 value of aminophylline (0.019 ± 0.05) was not significantly different (p=0.454) with C. aeruginosa (0.024 ± 0.05). Conclusion: It could be concluded that C. aeruginosa extracts have the potency to be further developed as a new natural source of the antiasthma agents.", "keywords": [ "antiasthma", "Curcuma aeruginosa", "spasmolytic" ], "content": "Introduction\n\nAsthma is an inflammatory airway disease characterized by the occurrence of an respiratory airway hyper response and reversible narrowing of the airway1. Asthma is one of the major non-communicable diseases in the world. About 235 million people worldwide suffer from asthma, particularly children. The strongest risk factors for developing asthma are a combination of genetic susceptibility to certain inhalable allergens and environmental exposure to them2. Asthma medications are given to manage asthma sufferers3. Herbal preparations are one of the most popular complementary treatments used by asthmatic patients. Many important asthma drugs such as B2-agonists, anticholinergics, methylxanthines, and cromones have herbal origins4. Some medicinal plants have the effect of reducing smooth muscle stiffness, similar to the mechanism of asthma drugs, especially the anticholinergic drugs5. Research has also shown that some medicinal plants have the anti-inflammatory effects, following the same mechanism of corticosteroid drug used in asthma treatment6.\n\nThe genus Curcuma (family Zingiberaceae) consisting of more than 100 species is used widely as food a and in traditional medicines7. Indonesia is home to many species of Curcuma. The various species of Curcuma often used are C. longa (turmeric), C. xanthorrhiza, C. heyneana, C. aeruginosa, C. mangga, and C. zedoaria. Turmeric is the most frequently used plant for traditional medicine in Indonesia. C. aeruginosa considered as indigenous Curcuma species in Indonesia are currently not extensively studied, yet8.\n\nImportant medicinal plants from the genus Curcuma with anti-asthmatic potential include C. longa9. Other rhizomes of Curcuma species are traditionally used in the treatment of asthma, i.e. C. aeruginosa10, C. mangga11, C. caesia12, and C. zedoaria13. The antiasthma effects of C. aeruginosa are currently known, therefore, the objective of this study was to establish the tracheospasmolytic activity of C. aeruginosa applied on isolated tracheas of guinea pigs.\n\n\nMethod\n\nThe sampling of medicinal plants was conducted in Kutai Kartanegara District, East Kalimantan (0°59’51.1”S 116°58’33.1”E). Plants were then identified in the Faculty of Mathematics and Natural Sciences, Mulawarman University by comparing to the university herbarium collection.\n\nThe rhizomes of C. aeruginosa were sliced and dried at room temperature for 3 days, crushed and transferred into a glass container. Approximately 1 kg of crushed rhizomes was soaked in 1 L of absolute ethanol (9401-03 Alcohol, Anhydrous, Reagent, J.T. Baker) for 5 days. The mixture was shaken occasionally with a shaker (3525 Incubator Orbital Shaker, Lab-Line, US). After 5 days, the materials were filtered (Whatman Filter Paper 11µm, Sigma-Aldrich) and evaporated using a rotary evaporator (RV06-ML Rotary Evaporator, IKA, Germany). The dried extracts were obtained and stored at 4°C in a dark bottle until use.\n\nOne male guinea pig (Cavia porcellus) (6 months old, 485 g) was obtained from Animal House Faculty of Medicine (Mulawarman University). They were treated in a controlled room temperature of 25°C, with a 12-hour light/dark cycle, and access to food pellets and filtered water ad libitum. The guinea pig was anesthetized intraperitoneally with a ketamine injection (Hameln Pharmaceuticals, Germany) at a dose of 60 mg/kg before the trachea was taken. After anesthetized, animals were euthanized by cervical dislocation. The trachea was quickly dissected by adhering fat and connective tissue of guinea pig.\n\nThe trachea rings were suspended in L-shaped wire loops in 10 ml organ baths (PL3508B6 Panlab Organ Bath System, ADInstruments), containing the Krebs solution (K3753 Krebs-Henseleit Buffer, Sigma-Aldrich) aerated with carbogen by maintaining the temperature at 37°C. Isometric contractions of tracheal rings were measured by the transducer (7004 Isometric Force Transducers, Ugo Basile) coupled to the amplifier (FE 221 BridgeAmp, ADInstruments) connected to PC running LabChart V5 software. An equilibration period of 90 minutes was done in Krebs solution. At the end of the equilibration period, the tracheal rings were stimulated with histamine in order to establish viability. After equilibration, the tracheal rings were exposed to DMSO (W387520 Methyl sulfoxide, Sigma-Aldrich) as the negative control, aminophylline (A1755 Aminophylline, Sigma-Aldrich) as positive control drugs and extract of C. aeruginosa according to the experimental protocol by Janbaz et al.14. The dosage for DMSO, aminophylline and plant extract were 0.0001, 0.0003, 0.001, 0.003, 0.01, and 0.03 mg/ml given 700, 750, 800, 850, 900, and 950 seconds after equilibration on the organ baths. The dose-response curve for trachea relaxation activity was obtained in every preparation.\n\nTrachea relaxation activity is tabulated in the mean ± SD curve of the dose-response curve. The value of EC50 was calculated with Microsoft Excel 2016 as shown in Dataset 1. Data were analyzed using the Mann-Whitney because not normally distributed. All statistical analysis was performed using SPSS version 16.0 for Windows. A p-value of ≤ 0.05 was considered to be significant.\n\nAll protocols used in this experiment received approval from the Ethical Animal Care from the Medical and Health Research Ethics Commission, Faculty of Medicine, Mulawarman University No. 72/KEPK-FK/V/2018. All efforts were made to ameliorate any suffering of animals used in this research.\n\n\nResults\n\nThe results of trachea relaxation between negative control, aminophylline, and C. aeruginosa extract presented in Figure 1. The result showed that the decrease of spasmolytic activity of C. aeruginosa extract was significantly better (p=0.000) than that in negative control. Meanwhile, the EC50 value of aminophylline (0.019 ± 0.05) was not significantly different (p=0.454) with C. aeruginosa (0.024 ± 0.05), as shown in Figure 2.\n\n\nDiscussion\n\nC. aeruginosa (Supplementary File 1) is known in Indonesia as temu ireng or “pink and blue ginger” in English15. The color of fresh the rhizome can be yellows or greenish blue in color and mildly aromatic with a ginger-like aroma16. C. aeruginosa has been used as a traditional medicine in South and Southeast Asia17. The rhizomes have been used for gastrointestinal and uterine disorders, as well as parasitic and fungal infection18.\n\nOther pharmacological activities of C. aeruginosa that have been reported include inhibition of HIV, anti-cancer activity, hepatoprotective, antiandrogenic, estrogenic properties, antimicrobial, antioxidant, antiplatelet-activating factor-like, antipyretic, antinociceptive, and anti-inflammatory19. Germacrone, zedoarone, dehydrocurdione, curcumenol, zedoarondiol, and isocurcumenol were chemical constituents from sesquiterpenes isolated from rhizomes of C. aeruginosa20. In this study, the examination of antiasthma effects of C. aeruginosa has been reported. Further research is needed to identify the chemical compounds from C. aeruginosa that could convey antiasthma activity.\n\n\nConclusion\n\nThe results of this study indicate that ethanol extract of C. aeruginosa has an antiasthma effect based on the tracheospasmolytic activity. Therefore, C. aeruginosa can be developed as a possible source of new antiasthma drugs.\n\n\nData availability\n\nF1000Research: Dataset 1. Trachea relaxation between C. aeruginosa (CA), aminophylline (A) and negative control (N) and EC50 result on trachea relaxation between C. aeruginosa (CA) and aminophylline (A)., 10.5256/f1000research.16416.d22169021", "appendix": "Grant information\n\nThis study was supported with funding from the Directorate General for Research and Development, Ministry of Research and Technology Higher Education of the Republic of Indonesia, for financing this research, as part of the implementation of Basic Research of Flagship University Year of 2018 No. 0045/E3/LL/2018.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary material\n\nSupplementary File 1: Picture of rhizome of Curcuma aeruginosa Roxb.\n\nClick here to access the data\n\n\nReferences\n\nBarnes PJ: Asthma. In Harrison’s Principles of Internal Medicine. Nineteenth DL Kasper, SL Hauser, JL Jameson, AS Fauci, DL Longo and Loscalzo J, Eds. New York: McGraw-Hill Education, 2015; 1669–1681. Reference Source\n\nWorld Health Organization: Factsheets: Asthma. World Health Organization. 2018. Reference Source\n\nChestnutt MS, Prendergast TJ: Pulmonary Disorders. In Current Medical Diagnosis & Treatment. Fifty-Sixt, MA Papadakis and SJ McPhee, Eds. New York: McGraw-Hill Education, 2017; 241–321. Reference Source\n\nKaraman M, Firinci F, Cilaker S, et al.: Anti-inflammatory effects of curcumin in a murine model of chronic asthma. Allergol Immunopathol (Madr). 2012; 40(4): 210–214. PubMed Abstract \| Publisher Full Text\n\nKumar A: Medicinal properties of Acorus calamus. J Drug Deliv Ther. 2013; 3(3): 143–144. Reference Source\n\nParamita S, Kosala K, Dzulkifli D, et al.: Anti-inflammatory activities of ethnomedicinal plants from dayak abai in North Kalimantan, Indonesia. Biodiversitas. 2017; 18(4): 1556–1561. Reference Source\n\nNararat A, Sirilun S, Julsrigival J, et al.: Chemical profiling and antimicrobial activity of essential oil from Curcuma aeruginosa Roxb., Curcuma glans K. Larsen & J. Mood and Curcuma cf. xanthorrhiza Roxb. collected in Thailand. Asian Pac J Trop Biomed. 2017; 7(10): 881–885. Publisher Full Text\n\nRahayu M, Sangat H, Lestari V: The Ethnobotanical Aspects of Curcuma spp. In Javanese Traditional Medicine. In Proceedings of 2nd International Symposium on Temulawak and 40th Meeting of National Working Group on Indonesian Medicinal Plant. 2011. Reference Source\n\nShakeri F, Soukhtanloo M, Boskabady MH: The effect of hydro-ethanolic extract of Curcuma longa rhizome and curcumin on total and differential WBC and serum oxidant, antioxidant biomarkers in rat model of asthma. Iran J Basic Med Sci. 2017; 20(2): 155–165. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRajamma A, Bai V, Nambisan B: Antioxidant and antibacterial activities of oleoresins isolated from nine Curcuma species. Phytopharma. 2012; 2(2): 312–317. Reference Source\n\nDewangan MK, Dwivedi C, Sivna PL, et al.: Medicinal Value of Curcuma cassia Roxb: An Overview. Crit Rev Pharm Sci. 2014; 3(4): 1–9. Reference Source\n\nPakkirisamy M, Kalakandan S, Ravichandran K: Phytochemical Screening, GC-MS, FT-IR Analysis of Methanolic Extract of Curcuma caesia Roxb (Black Turmeric). Pharmacogn J. 2017; 9(6): 952–956. Publisher Full Text\n\nPathan A, Alshahrani A, Al-Marshad F: Pharmacological Effect of Curcuma zedoaria rhizomes on milk induced eosinophilia in the management of Asthma. Int J Ethnobiol Ethnomed. 2015; 1(1): 1–5. Reference Source\n\nJanbaz KH, Hamid I, Mahmood MH, et al.: Bronchodilator, cardiotonic and spasmolytic activities of the stem barks of Terminalia arjuna. Can J Appl Sci. 2011; 1(3): 104–120. Reference Source\n\nNurcholis W, Priosoeryanto B, Purwakusumah E, et al.: Antioxidant, Cytotoxic Activities and Total Phenolic Content of Four Indonesian Medicinal Plants. Valensi. 2012; 2(4): 501–510. Reference Source\n\nSrivilai J, Khorana N, Waranuch N, et al.: Anti-androgenic activity of furanodiene isolated from Curcuma aeruginosa Roxb. Extract. Naresuan Univ J. 2011; Special Is: 33–37. Reference Source\n\nHossain CF, Al-Amin M, Sayem AS, et al.: Antinociceptive principle from Curcuma aeruginosa. BMC Complement Altern Med. 2015; 15(1): 191. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJose S, Thomas TD: Comparative phytochemical and anti-bacterial studies of two indigenous medicinal plants Curcuma caesia Roxb. and Curcuma aeruginosa Roxb. Int J Green Pharm. 2014; 8(1): 65–71. Reference Source\n\nTheanphong O, Mingvanish W, Kirdmanee C: Chemical Constituents and Biological Activities of Essential Oil from Curcuma aeruginosa Roxb. Rhizome. Bull Heal Sci Technol. 2015; 13: 6–16. Reference Source\n\nSuphrom N, Pumthong G, Khorana N, et al.: Anti-androgenic effect of sesquiterpenes isolated from the rhizomes of Curcuma aeruginosa Roxb. Fitoterapia. 2012; 83(5): 864–871. PubMed Abstract \| Publisher Full Text\n\nParamita S, Moerad EB, Ismail S, et al.: Dataset 1 in: Antiasthmatic effect of Curcuma aeruginosa extract on isolated organ of the trachea. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16416.d221690" }	[ { "id": "53554", "date": "12 Sep 2019", "name": "Ian Sinha", "expertise": [ "Reviewer Expertise Paediatric asthma." ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an interesting paper. Cumin has been shown to have potential anti-inflammatory effects in other papers, and I think you should do a proper literature review around this - you will need to incorporate this in both your background and discussion which are both focussed on cumin but not asthma.\nI am not sure why you chose aminophylline rather than a better bronchodilator such as salbutamol, and you will need to explain the limitation of looking at the trachea in an animal model, rather than being able to evaluate the small airways.\nNonetheless, this is an interesting paper.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "53981", "date": "03 Oct 2019", "name": "Waras Nurcholis", "expertise": [ "Reviewer Expertise Biochemistry" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis manuscript mainly describes the antiasthmatic activity from ethanol extract of Curcuma aeruginosa rhizome. The work purpose is quite interesting but the organization is weak. So, the major decision was required for the following reasons.\nIn title, please give information about your extract (ethanolic extract) and part used in this study (rhizome). I recommend changing the title to “Antiasthmatic activity from ethanol extract of Curcuma aeruginosa Roxb. rhizome”.\n\nPlease check and re-write some sentences in Abstract.\n\nIntroduction\nParagraph 2. “The genus Curcuma … and in traditional medicines”. Please write the sentence with an appropriate meaning. Please add a reference in “Turmeric is the most…in Indonesia”. Please review in your introduction for several pharmacological activities from C. aeruginosa rhizome such as cytotoxicity1, antioxidant2 etc. Please create important novelty in your work in this article.\n\nMethod\nIn spasmolytic activity:\nPlease give information, how to calculate for trachea relaxation? Please give a calculation formula in your method. Please give information for negative control? Are you sure DMSO? Please explain more. Your data and in method must be in line.\n\nIn plant extraction: Please give information for extract yield? Please give information for certificate analysis in your ethanolic extract or standardize your extract?\n\nResults\nIn Figure 1 and Figure 2, please show the value of significantly? Also, note in your figure. Analysis data with the results must be appropriate.\n\nDiscussion is poor of information. Please discuss more about your data. Please study more about antiasthmatic activity from medicinal plant extract and metabolite?\n\nConclusion: Please give information value for potency your sample in antiasthma activity?\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] }, { "id": "53982", "date": "08 Nov 2019", "name": "Sven-Erik Dahlén", "expertise": [ "Reviewer Expertise Airway pharmacology", "Translational asthma research" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors attempt to evaluate if an ethanolic extract of the plant Curcuma Aeruginosa possesses bronchodilatory actions in the guinea pig trachea. Figure 1 shows that the influence of the extract is no different from a presumed negative control (DMSO). The data therefore do not support the author’s conclusion that the extract has anti-asthmatic effects. The title of the manuscript is therefore wrong.\nSpecific major concerns:\nAlthough figure 1 displays data without indications of the variability, the curves for the extract and the negative control are almost superimposed, whereas the curve for aminophylline shows the expected relaxation.\n\nThe calculations for Figure 2 are impossible to understand, and as far as I can see incorrect. EC 50 values are often calculated as 50% of the maximal response to e.g. a supramaximal dose of aminophylline. Here, it seems that the value causing 50% of the response to the highest dose tested (3 mg/mL) has been used for aminophylline, but this dose causes only 35% relaxation of the preparation. It is even more difficult to understand how an EC 50 value can be calculated for the extract because figure 1 indicates it causes no relaxation. In addition, the weight of an extract containing many undefined ingredients cannot be compared for potency with a pure pharmacologic agonist. The only comparison that would be relevant would be the qualitative documentation that the extract has bioactivity. The study has failed to show that.\n\nThe method description is unclear. Was the extract and the controls tested in a histamine-precontracted preparation, or only in a preparation kept at baseline tension?\nMinor comments:\nWhy was DMSO the negative control? Was really the lyophilized extract dissolved in DMSO for the tests?\n\nThe dosage of DMSO is given in mg/mL and seems to be quite high.\n\nThe statement in the introduction suggesting that herbal preparations are popular for treatment of asthma is new to me. This needs to be referenced.\n\nThe description of the experimental protocol is incomplete, eg. dose of histamine not stated, the isometric load not defined, etc.\n\nIn the discussion, the authors report a number of pharmacologic activities of the extract. This would seem to preclude the use of the extract as many of the effects eg on endocrine regulation would be expected to be associated with adverse reactions.\n\nThe language of the manuscript needs language review. The wording is often imprecise and incorrect.\n\nIs the work clearly and accurately presented and does it cite the current literature? No\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [] } ]	1	https://f1000research.com/articles/7-1799
https://f1000research.com/articles/7-1797/v1	15 Nov 18	{ "type": "Study Protocol", "title": "Disulfiram repurposing combined with nutritional copper supplement as add-on to chemotherapy in recurrent glioblastoma (DIRECT): Study protocol for a randomized controlled trial", "authors": [ "Asgeir Store Jakola", "Katja Werlenius", "Munila Mudaisi", "Sofia Hylin", "Sara Kinhult", "Jiri Bartek Jr.", "Øyvind Salvesen", "Sven Magnus Carlsen", "Michael Strandéus", "Magnus Lindskog", "David Löfgren", "Bertil Rydenhag", "Louise Carstam", "Sasha Gulati", "Ole Solheim", "Jiri Bartek", "Tora Solheim", "Katja Werlenius", "Munila Mudaisi", "Sofia Hylin", "Sara Kinhult", "Jiri Bartek Jr.", "Øyvind Salvesen", "Sven Magnus Carlsen", "Michael Strandéus", "Magnus Lindskog", "David Löfgren", "Bertil Rydenhag", "Louise Carstam", "Sasha Gulati", "Ole Solheim", "Jiri Bartek", "Tora Solheim" ], "abstract": "Background: Disulfiram (DSF) is a well-tolerated, inexpensive, generic drug that has been in use to treat alcoholism since the 1950s. There is now independent preclinical data that supports DSF as an anticancer agent, and experimental data suggest that copper may increase its anti-neoplastic properties. There is also some clinical evidence that DSF is a promising anticancer agent in extracranial cancers. In glioblastoma, DSF induced O6-methylguanine methyltransferase (MGMT) inhibition may increase response to alkylating chemotherapy. A recent phase I study demonstrated the safety of DSF in glioblastoma patients when DSF was administered at doses below 500 mg/day together with chemotherapy. We plan to assess the effects of DSF combined with nutritional copper supplement (DSF-Cu) as an adjuvant to alkylating chemotherapy in glioblastoma treatment. Methods: In an academic, industry independent, multicenter, open label randomized controlled phase II/III trial with parallel group design (1:1) we will assess the efficacy and safety of DSF-Cu in glioblastoma treatment. The study will include 142 patients at the time of first recurrence of glioblastoma where salvage therapy with alkylating chemotherapy is planned. Patients will be randomized to treatment with or without DSF-Cu. Primary end-point is survival at 6 months. Secondary end-points are overall survival, progression free survival, quality of life, contrast enhancing tumor volume and safety. Discussion: There is a need to improve the treatment of recurrent glioblastoma. Results from this randomized controlled trial with DSF-Cu in glioblastoma will serve as preliminary evidence of the future role of DSF-Cu in glioblastoma treatment and a basis for design and power estimations of future studies. In this publication we provide rationale for our choices and discuss methodological issues. Trial registration: The study underwent registration in EudraCT 2016-000167-16 (Date: 30.03.2016,) and Clinicaltrials.gov NCT02678975 (Date: 31.01.2016) before initiating the study.", "keywords": [ "Randomized controlled trial", "glioma", "glioblastoma", "disulfiram", "alkylating agents", "brain tumor" ], "content": "Introduction\n\nGlioblastoma is a highly malignant brain tumor associated with limited treatment responses and population based series report median survival of only 10 months1. Standard initial treatment is extensive surgical resections followed by postoperative fractionated radiotherapy, and concomitant and adjuvant temozolomide treatment. Unfortunately, initial treatment prolongs survival only modestly and in principle all glioblastomas eventually recur2. Initial responses to chemotherapy are more often seen in patients with O6-methylguanine methyltransferase (MGMT) hypermethylation, and consequently this is an important predictive biomarker3. At time of recurrence there is no international treatment consensus4,5. A common approach is to reintroduce temozolomide as long as the progression occurs more than 3–6 months after end of the initial treatment. In patients with tumor progression early after primary treatment or seen on ongoing temozolomide therapy, lomustine (also known as CCNU)6 or the combination regimen procarbazine, CCNU and vincristine (PCV) are often the treatment of choice. However, effect of second line treatment is often limited and new improved treatments are therefore imperative.\n\n‘Repurposing’ of drugs for new indications has gained increased interest7. Disulfiram (DSF) is an inexpensive, generic drug that has been used since 1947 to treat alcohol dependency8. When alcohol is avoided, the drug is well tolerated in doses below 500 mg daily even with concomitant administration of temozolomide9.\n\nDSF has demonstrated a wide range of antineoplastic effects, potentially also in gliomas10–16. The interaction of DSF with copper (DSF-Cu) is likely responsible for the anti-cancer effect where DSF-Cu inhibit so-called glioma initiating cells at very low concentrations17. DSF-Cu may further inhibit the proteasome system10,17,18 and potentiate alkylating DNA damage through impaired DNA repair in glioma cells, an effect that may be mediated through inhibition of MGMT17,19. Finally, a recent paper demonstrated benefit of cancer patients using disulfiram and suggested ditiocarb–copper complex to be the metabolite of disulfiram with anti-cancer effects through effects on protein turnover20. As a consequence, it is now suggested that DSF-Cu should be implemented in clinical trials17,21.\n\nWe seek to study the efficacy and safety of adding DSF combined with nutritional copper supplement (DSF-Cu) to patients with recurrent glioblastomas and concomitant with alkylating chemotherapy. The study is an academic, multicenter, open labeled randomized controlled phase II/III superiority trial with parallel group design. The aim of this paper is to present the intervention, design, endpoint, and power calculations of this study.\n\n\nMethods and design\n\nThe study is a multi-center, open label randomized controlled phase II/III trial with parallel group design. All patients receive alkylating chemotherapy, and in the experimental arm treatment with DSF-Cu is added. Patients will be in the study at a maximum of 24 months or until death.\n\nStudy setting. The patients will be recruited from and treated at University Hospitals in Norway and Sweden. The study involves eight centers (Gothenburg, Stockholm, Linköping, Lund, Jönköping, Uppsala, Örebro and Trondheim). The study was opened for inclusion January 2017 and per October 14 2018 we have randomized 36 patients.\n\nEligibility criteria. Adult patients (≥ 18 years) with first recurrence of glioblastoma will be screened for inclusion from the regular clinical practice at regional oncological departments (Figure 1). The full eligibility criteria are presented in Table 1. In short, adult patients intended for treatment with alkylating chemotherapy for recurrent glioblastoma are eligible if they have a Karnofsky Performance status 60–100 (ECOG Grade 0-2)22, are willing to refrain from alcohol, and have no major contraindications towards DSF-Cu.\n\nInterventions. There is no consensus on choice of treatment at first glioblastoma recurrence5. In the Nordic countries, temozolomide, lomustine or PCV are commonly applied. A pragmatic comparator is thus needed to reflect clinical practice. Preclinical data suggest that DSF may enhance the effect of alkylating agents through inhibition of DNA repair17. We therefore use DSF-Cu combination as add-on to “any alkylating chemotherapy” in the intervention group while the comparator group receives “any alkylating chemotherapy” alone.\n\nThe administration of DSF and nutritional copper supplement is scheduled to start concomitant with alkylating chemotherapeutic treatment. Patients will take DSF (Antabus®) as a once daily oral dose of 400 mg, based on the known toxicity profile9. In case of intolerance, dose reduction to 200 mg per day is allowed. Copper supplementation will be administered separately from DSF with DSF administration in the evening. Copper nutritional supplement will correspond to 2.5 mg of elementary copper, which is not considered a safety concern23. Adherence to treatment will be assessed at each control with a reminder, self-reporting of drug intake and drug tablet count of remaining DSF. The only surveillance of nutritional copper intake is by self-reporting.\n\nBefore study closure no crossover from the control group to DSF-Cu is allowed.\n\nThe intervention should continue also after change of chemotherapy, or during temporary chemotherapy withdrawal due to reached maximal cumulative dose or side effects from chemotherapy.\n\nPatients are to discontinue permanently the treatment with DSF-Cu for any of the following reasons:\n\nUnacceptable adverse effect(s) as judged by the treating doctor.\n\nNo longer indication for tumor targeted therapy.\n\nThe patient wishes to withdraw from treatment.\n\nIf patients develop conditions that contribute to excess risk as judged by the treating physician.\n\nOther experimental therapy, including compassionate use programs, are not allowed in either treatment group as long as the patient is in the trial and receives the assigned treatment.\n\nOutcomes. The primary end-point of the study is survival assessed at 6 months post-randomization.\n\nSecondary end-points are:\n\nOverall survival from date of randomization.\n\nMedian progression free survival.\n\n6 and 12 month progression free survival.\n\nHealth-related generic quality of life24.\n\nVolumetric growth rate assessed from baseline MRI to first follow-up MRI scan25.\n\nSafety assessment (toxicities).\n\nParticipant timeline. To reduce patient burden, timing of data collection after randomization is scheduled according to choice of chemotherapy regimen, where patients that receive temozolomide usually are assessed every 4th week while patients treated with PCV/lomustine are assessed every 6th week. Consequently, this may lead to four additional observations in the temozolomide group compared with the lomustine/PCV group. The more detailed clinical control and MRI assessment every 3rd month is independent of chemotherapy regimen. The primary endpoint will not be affected by the differences in timing of assessments during the trial.\n\nFinally, hematological status and liver function will be monitored as long as patients are on active treatment every 4–6 weeks, once again depending on the chemotherapeutic protocol (Table 2).\n\nAE denotes adverse event, SAE; serious adverse event, SAR; serious adverse reaction SUSAR; suspected unexpected serious adverse reaction, MRI; magnetic resonance imaging, KPS; Karnofsky Performance Status\n\nThe interval for patients treated with lomustine/PCV is 6 weeks, while 4 weeks is clinical routine in the case of temozolomide\n\nContinued alcohol screen only relevant if randomized to the disulfiram group.\n\n*Blood samples as indicated here relevant as long as on active treatment and includes: hemoglobin, thrombocytes, leukocytes with differential count, ASAT, ALAT, G-GT, ALP, bilirubin, INR).\n\nSample size and study power. The sample size calculation is based on data from previous trials reporting that 50% of patients treated with lomustine deceased within 8 months6. We use a phase III end-point in a phase II setting because this study will be the basis for power calculation of a later more rigid phase III study, and because progression is less relevant in patients with already demonstrated progression/recurrence and with a short expected survival. Due to the phase II/III setting we are relaxing the alpha value to 0.10 (by convention 0.05) since phase II studies are “preliminary assessment of a new intervention before embarking on a larger and expensive randomized controlled trial”26. Thus, we argue that it is good science to slightly increase the chance of false positive results (due to chance) for a relevant end-point instead of using less relevant and difficult to interpret surrogate end-points (such as progression).\n\nBased on the assumption that the treatment group will have an improvement in proportion achieving 6-month survival from 60% to 80% we need a final sample size of 128 in total (64 in each group; alpha 0.1, power 80% and two sided test). Considering 10% attrition, the actual included sample need to be 142 with 71 patients included in each group.\n\nAssignment of interventions. A web-based randomization system (WebCRF 3.0) developed and administered by the Unit of Applied Clinical Research, Faculty of Medicine and Health Sciences, Norwegian University of Science and Technology (NTNU, Trondheim, Norway), will be used. Randomizations are computer generated in a 1:1 ratio with stratification by center. Blocks of varying sizes will be used to make prediction of allocation impossible. Randomization will be performed only after the initial chemotherapy treatment has been selected. Patient enrollment is performed by a study member with appropriate delegation at the local site.\n\nData collection. Patients will be in the trial for a maximum of 24 months or until death (see study schedule in Table 2). Generic quality of life is followed every 3rd month using EQ-5D 3L. Quality of life will be monitored until progression, death or 24 months post-inclusion24,27.\n\nMRI scanning will be performed at a minimum of every 3rd month. MRI assessment may continue up to 24 months post-inclusion and this should be implemented also after documented progression in the DSF-Cu group, as long as the patient continues on only DSF-Cu, or in combination with another established rescue therapy such as surgery, radiation therapy or other alkylating chemotherapy. Consequently, MRI controls are indicated and should be continued as long as tumor targeted therapy is administered (both in control and intervention group). The RANO criteria is used in MRI assessment and for evaluation of progression28.\n\nAfter withdrawal of tumor target therapy no additional follow-up (including MRI) is planned within the context of this study, in order to minimize patient burden in the end of life setting. Safety assessment are according to the Common Terminology Criteria for Adverse Events (CTCAE) version 4.029. Survival is followed until death or 24 months post-inclusion.\n\nData management. Data will be recorded in the web-based eCRFs (WebCRF 3.0) at The Unit for Applied Clinical Research at the NTNU. All data will be entered in the WebCRF 3.0 at the local study site. All electronic patient information is handled using a study participation number that is not logically connected to any personal data. It is possible to manually “decode” back to personal data using a name list stored securely at the respective study sites.\n\nAll data entered into the WebCRF are stored at two separate servers at NTNU. These servers are protected by the general firewall at the university and then by another firewall dedicated to protect clinical research data. The server that stores the data is not the same server as the one in contact with the Internet. There are several further safety measures not to be exposed. There are backups taken every 24-hour throughout the years. The responsibility for this system lies with the Unit for Applied Clinical Research at NTNU.\n\nStatistical methods. All outcomes will be analyzed according to the intention to treat principle unless otherwise specified in post-hoc analyses. This is an open label trial, however the statistician performing analyses will be blinded for treatment assignment. Also, an investigator blinded for treatment allocation will perform the assessment of radiological end-points.\n\nThe 6-month survival (i.e primary end-point) will be assessed with 2x2 tables and analyzed with chi-square test. This analysis will be performed 6-months after last patient inclusion. Concerning the primary end-point we do not expect any missing data.\n\nConcerning the secondary outcomes they will be evaluated as follows:\n\nActual survival data at 24 months post-inclusion will be analyzed similar to the primary end-point.\n\nOverall survival analyses will be analyzed using Kaplan-Meier plot and tested with log-rank test. This analysis will be performed at 6 months after the last patient inclusion together with the assessment of primary outcome. Also, this will be analyzed in a longer-term follow-up study (24 months post-inclusion).\n\nOverall progression free survival will be analyzed with Kaplan-Meier plot and tested with log-rank test. This analysis will also be performed at 6 months and at 24 months after the last patient inclusion. The assessment of progression will be done in accordance with the latest RANO criteria by the local investigator.\n\nActual progression free survival analyses at 6 and 24 months after the last patient inclusion. The assessment of progression will be done in accordance with the latest RANO criteria by the local investigator and analyzed similar to the primary end-point.\n\nComparison of general health related quality of life at 6 and 24 months after last included patients. We will use an area under the curve approach (i.e. cumulative quality of life)27.\n\nSemiautomatic segmentation of contrast-enhancing tumor volume assessed at baseline and first MRI control to assess growth dynamics. The data will be analyzed with independent samples t-test or Mann-Whitney U test, based on data distribution.\n\nSafety assessment with comparing proportion of a composite of grade 3, 4 and 5 toxicities according to the common terminology Criteria for Adverse Events (CTCAE).\n\nWe stratify only for study center, however important baseline variables like repeated surgery, time since primary treatment, age, functional status, lomustine versus temozolomide treatment, and biomarker status from initial surgery if available (IDH and MGMT) will be registered and adjusted for in post-hoc exploratory analyses. Also, we will explore outcomes in post-hoc as treated analyses if non-compliance to treatment protocol becomes apparent.\n\nAll source data will be accessible for auditing and monitoring. The external, independent monitor will maintain patient confidentiality. Authorities have the right to perform inspections as well as the monitor.\n\nAn initiation visit from study monitors takes place at each site before the enrolment of study subjects at the specific site can start. The first monitoring visit after patient recruitment has started will be performed after inclusion of 5 participants for the specific site, however a minimum of one contact (by e-mail or telephone) and one visit per year to each site is required. After the last study subject has completed the last visit and all CRFs are completed a close-out monitoring visit will take place at each site.\n\nHarms. Adverse events and toxicities is assessed at each study visit and graded according to the NCI Common Toxicity Criteria for adverse events (NCI-CTCAE) Version 4.0 (http://ctep.cancer.gov/reporting/ctc.html).\n\nSerious Adverse Event (SAE), Serious Adverse Reaction (SAR) and Suspected Unexpected Serious Adverse Reactions (SUSAR). A serious adverse reaction (SAR) is a serious adverse event (SAE) that may be related to trial treatment. Suspected unexpected serious adverse reactions (SUSAR) are reactions where DSF-Cu may be suspected as the reason for an unexpected SAR. The assessment of relatedness is the responsibility of Department of Clinical Pharmacology at Sahlgrenska University Hospital (Gothenburg, Sweden) and the local investigator responsible for the patient.\n\nAll reportable cases of adverse events are reported to the Department of Clinical Pharmacology, Sahlgrenska University Hospital. They are responsible to notify sponsor, authorities and local investigators when indicated.\n\nInterim analysis and criteria for aborting the trial. At 50% inclusion (71 included patients) an interim analysis on primary end-point and safety will be performed. An independent data monitoring scientific committee (DMSC) will review the results of this interim analysis. This committee consists of professor Anja Smits (Gothenburg), associate professor Petter Förander (Stockholm) and Dr. Annika Malmström (Linköping). A recommendation signed by all members of the monitoring committee will be handed to the sponsor after this meeting and the final decision is with the sponsor.\n\nFurther, the sponsor may terminate the study prematurely for any reasonable cause. The Ethics committees and competent authorities will then be informed. Conditions that may warrant termination include, but are not limited to:\n\nThe discovery of an unexpected, significant, or unacceptable risk to the patients enrolled in the study.\n\nIf competent authorities obtain information that raises doubts about the safety or scientific validity of the study, the competent authorities can suspend or prohibit the study.\n\nThe reporting of this protocol is consistent with the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) recommendation30. Reporting of the trial will be consistent with the consolidated standards of reporting trials (CONSORT) 2010 statement31.\n\nThis study will be conducted in compliance with the protocol and according to Good Clinical Practice guidelines and applicable regulatory standards. Further, the study will be conducted in accordance with the latest Declaration of Helsinki32. The regional committee in Gothenburg approved the study for all centers in Sweden in a multicenter approval (Dnr 332-16). Ethical approval for the Norwegian site was acquired separately from the ethical committee in region Central Norway (2016/283). Further, the study is approved by the Norwegian medicines agency (16/04133) and the Swedish medical product agency (5.1-2016-25235). Protocol modifications will be reported to the local ethical committees and significant modifications will in addition be communicated to the Norwegian medicines agency and the Swedish medical product agency.\n\nPatients will be included only after the informed consent form is signed (Supplementary File 1, Swedish version). The informed consent needs to be explicit and written and may at any time be withdrawn by the participant. Also, the right of the participant to refuse to participate without giving reasons must be respected.\n\nPatients without possibility to provide written, informed consent (e.g. severe cognitive deficit and aphasia) are not to be included via proxies.\n\nPatient confidentiality will be maintained through de-identified and secure storage of data and a secure, local storage of the name lists at the respective study sites.\n\nThe principal investigator (ASJ) and the study statistician (ØS) will have access to the final dataset. A cleaned dataset from specific sites may be given back in anonymous form to sites upon request, but only after the final scientific publication (after 24 months follow-up). No other data sharing is planned.\n\nThe trial results will be communicated through publication of a scientific report in a peer-reviewed journal. There are no restrictions on publication (e.g. a negative trial will still be published).\n\nAt the time of submission the recruitment has not been completed (started January 2017).\n\nThis paper is based on protocol version 2.0, dated December 12th 2017.\n\n\nDiscussion\n\nThis study investigates the role of DSF-Cu in recurrent glioblastoma. To our knowledge this is the first planned, randomized controlled trial with DSF-Cu in glioblastoma. The results from this trial will consequently serve as preliminary evidence of the future role of DSF-Cu in glioblastoma treatment as well as to serve as a basis for power estimations in future studies on the same topic.\n\nThe study has several pragmatic features, including an open design and a patient centered primary end-point33. Due to the nature of our study end-point, we need to recruit more patients than if we had chosen a surrogate marker which is more common in the phase II setting. Nevertheless, we believe that surrogate markers are less relevant in the setting of glioblastoma recurrence since survival is short and image interpretation and response assessment is difficult6,28. Consequently, the trial is best described as a phase II/III trial. Further, the trial is to a high degree integrated within the frame of regular clinical practice, reducing the need of extra resources and thereby cost33. The study will thus reflect clinical practice and the little deviation from everyday practice also reduce patient burden.\n\nIt may be a personal sacrifice for some patients to abstain from alcohol intake completely. Nevertheless, DSF has a well-known risk profile and the major risks of treatment with 400 mg DSF daily is mainly associated with lack of compliance with respect to alcohol intake and infrequently serious idiosyncratic hepatitis may be seen. In addition, there is a risk of neuropathy that is reversible upon drug discontinuation9. In sum, the risks of DSF-Cu, also combined with chemotherapy, are presumably limited9. The potential benefit lies in improved response to treatment.\n\nThe main strengths of this study are the multi-center setting, a hard primary end-point, secondary endpoints relevant to patients, a randomized controlled design and clinically relevant control group representing present best clinical practice. Further, the study being independent from the pharmaceutical industry is another strength. Limitations include the open design making both patients and investigators aware of treatment assignment. Thus, it is difficult to completely guarantee against unscheduled use of DSF-Cu in the control group. To minimize the risk of bias the statistical analyses both for primary and secondary endpoints will be performed by an investigator blinded for the actual treatment. Endpoints will be analyzed before analyzes of adverse events that may reveal actual treatment in the two groups.\n\nIn conclusion, we present a protocol for investigating the efficacy of disulfiram and nutritional copper supplement (DSF-Cu) as a supplement to alkylating chemotherapy in the setting of recurrent glioblastoma. We have provided rationale for our choices and discussed methodological issues of importance for this open labeled, randomized controlled trial.\n\n\nData availability\n\nNo data is associated with this article.", "appendix": "Grant information\n\nThe sponsor, Department of Neurosurgery, Sahlgrenska University Hospital, has no role in the research planning or execution of the trial. The data belongs to the investigator and is guaranteed for by the principal investigator. Further, the sponsor or principal investigator has no connection with the pharmaceutical industry.\n\nThe study is financed by grants from:\n\n• The Swedish state under the agreement between the Swedish government and the county councils concerning , the ALF-agreement (ALFGBG-716671)\n\n• AG Foundation.\n\n• Sahlgrenska Foundation.\n\n• Nordic Cancer Union.\n\n• The Swedish Society of Medicine (SLS-776901)\n\nThe funding bodies have no role in the research planning or execution of the trial.\n\n\nSupplementary material\n\nSupplementary File 1: Consent form, in Swedish.\n\nClick here to access the data\n\n\nReferences\n\nRønning PA, Helseth E, Meling TR, et al.: A population-based study on the effect of temozolomide in the treatment of glioblastoma multiforme. Neuro Oncol. 2012; 14(9): 1178–1184. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStupp R, Mason WP, van den Bent MJ, et al.: Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med. 2005; 352(10): 987–996. PubMed Abstract \| Publisher Full Text\n\nHegi ME, Liu L, Herman JG, et al.: Correlation of O6-methylguanine methyltransferase (MGMT) promoter methylation with clinical outcomes in glioblastoma and clinical strategies to modulate MGMT activity. J Clin Oncol. 2008; 26(25): 4189–4199. PubMed Abstract \| Publisher Full Text\n\nSeystahl K, Wick W, Weller M: Therapeutic options in recurrent glioblastoma--An update. Crit Rev Oncol Hematol. 2016; 99: 389–408. PubMed Abstract \| Publisher Full Text\n\nWeller M, van den Bent M, Hopkins K, et al.: EANO guideline for the diagnosis and treatment of anaplastic gliomas and glioblastoma. Lancet Oncol. 2014; 15(9): e395–403. PubMed Abstract \| Publisher Full Text\n\nTaal W, Oosterkamp HM, Walenkamp AM, et al.: Single-agent bevacizumab or lomustine versus a combination of bevacizumab plus lomustine in patients with recurrent glioblastoma (BELOB trial): a randomised controlled phase 2 trial. Lancet Oncol. 2014; 15(9): 943–953. PubMed Abstract \| Publisher Full Text\n\nCollins FS: Mining for therapeutic gold. Nat Rev Drug Discov. 2011; 10(6): 397. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChick J: Safety issues concerning the use of disulfiram in treating alcohol dependence. Drug Saf. 1999; 20(5): 427–435. PubMed Abstract \| Publisher Full Text\n\nHuang J, Campian JL, Gujar AD, et al.: A phase I study to repurpose disulfiram in combination with temozolomide to treat newly diagnosed glioblastoma after chemoradiotherapy. J Neurooncol. 2016; 128(2): 259–66. PubMed Abstract \| Publisher Full Text\n\nChen D, Cui QC, Yang H, et al.: Disulfiram, a clinically used anti-alcoholism drug and copper-binding agent, induces apoptotic cell death in breast cancer cultures and xenografts via inhibition of the proteasome activity. Cancer Res. 2006; 66(21): 10425–10433. PubMed Abstract \| Publisher Full Text\n\nKast RE, Boockvar JA, Bruning A, et al.: A conceptually new treatment approach for relapsed glioblastoma: coordinated undermining of survival paths with nine repurposed drugs (CUSP9) by the International Initiative for Accelerated Improvement of Glioblastoma Care. Oncotarget. 2013; 4(4): 502–530. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiu P, Brown S, Goktug T, et al.: Cytotoxic effect of disulfiram/copper on human glioblastoma cell lines and ALDH-positive cancer-stem-like cells. Br J Cancer. 2012; 107(9): 1488–1497. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLoo TW, Bartlett MC, Clarke DM: Disulfiram metabolites permanently inactivate the human multidrug resistance P-glycoprotein. Mol Pharm. 2004; 1(6): 426–433. PubMed Abstract \| Publisher Full Text\n\nRae C, Tesson M, Babich JW, et al.: The role of copper in disulfiram-induced toxicity and radiosensitization of cancer cells. J Nucl Med. 2013; 54(6): 953–960. PubMed Abstract \| Publisher Full Text\n\nTriscott J, Lee C, Hu K, et al.: Disulfiram, a drug widely used to control alcoholism, suppresses the self-renewal of glioblastoma and over-rides resistance to temozolomide. Oncotarget. 2012; 3(10): 1112–1123. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWesthoff MA, Zhou S, Nonnenmacher L, et al.: Inhibition of NF-κB signaling ablates the invasive phenotype of glioblastoma. Mol Cancer Res. 2013; 11(12): 1611–1623. PubMed Abstract \| Publisher Full Text\n\nLun X, Wells JC, Grinshtein N, et al.: Disulfiram when Combined with Copper Enhances the Therapeutic Effects of Temozolomide for the Treatment of Glioblastoma. Clin Cancer Res. 2016; 22(15): 3860–75. PubMed Abstract \| Publisher Full Text\n\nHothi P, Martins TJ, Chen L, et al.: High-throughput chemical screens identify disulfiram as an inhibitor of human glioblastoma stem cells. Oncotarget. 2012; 3(10): 1124–1136. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nParanjpe A, Zhang R, Ali-Osman F, et al.: Disulfiram is a direct and potent inhibitor of human O6-methylguanine-DNA methyltransferase (MGMT) in brain tumor cells and mouse brain and markedly increases the alkylating DNA damage. Carcinogenesis. 2014; 35(3): 692–702. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSkrott Z, Mistrik M, Andersen KK, et al.: Alcohol-abuse drug disulfiram targets cancer via p97 segregase adaptor NPL4. Nature. 2017; 552(7684): 194–199. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKast RE, Karpel-Massler G, Halatsch ME: CUSP9 treatment protocol for recurrent glioblastoma: aprepitant, artesunate, auranofin, captopril, celecoxib, disulfiram, itraconazole, ritonavir, sertraline augmenting continuous low dose temozolomide. Oncotarget. 2014; 5(18): 8052–8082. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKarnofsky DA, Burchenal JH: The clinical evaluation of chemotherapeutic agents in cancer. New York: Columbia University Press. 1949; 191–205. Reference Source\n\nEuropean Food Safety Authority: Tolerable upper intake levels for vitamins and minerals. Reference Source\n\nEuroQol Group: EuroQol--a new facility for the measurement of health-related quality of life. Health Policy. 1990; 16(3): 199–208. PubMed Abstract \| Publisher Full Text\n\nStensjøen AL, Solheim O, Kvistad KA, et al.: Growth dynamics of untreated glioblastomas in vivo. Neuro Oncol. 2015; 17(10): 1402–11. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKhan I, Sarker SJ, Hackshaw A: Smaller sample sizes for phase II trials based on exact tests with actual error rates by trading-off their nominal levels of significance and power. Br J Cancer. 2012; 107(11): 1801–1809. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSagberg LM, Solheim O, Jakola AS: Quality of survival the 1st year with glioblastoma: a longitudinal study of patient-reported quality of life. J Neurosurg. 2015; 124(4): 989–97. PubMed Abstract \| Publisher Full Text\n\nWen PY, Macdonald DR, Reardon DA, et al.: Updated response assessment criteria for high-grade gliomas: response assessment in neuro-oncology working group. J Clin Oncol. 2010; 28(11): 1963–1972. PubMed Abstract \| Publisher Full Text\n\nNational Cancer Institute: Common Terminology Criteria for Adverse Events (CTCAE), version 4.0. Bethesda, USA. 2010. Reference Source\n\nChan AW, Tetzlaff JM, Gøtzsche PC, et al.: SPIRIT 2013 explanation and elaboration: guidance for protocols of clinical trials. BMJ. 2013; 346: e7586. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchulz KF, Altman DG, Moher D, et al.: CONSORT 2010 Statement: updated guidelines for reporting parallel group randomised trials. Trials. 2010; 11: 32. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWorld Medical Association: World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects JAMA. 2013; 310(20): 2191–2194. PubMed Abstract \| Publisher Full Text\n\nFord I, Norrie J: Pragmatic Trials. N Engl J Med. 2016; 375(5): 454–463. PubMed Abstract \| Publisher Full Text" }	[ { "id": "40689", "date": "23 Nov 2018", "name": "Kalkunte S. Srivenugopal", "expertise": [ "Reviewer Expertise Brain tumor chemotherapy", "redox-stress signaling and anticancer drug resistance" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe alcohol aversion and metal-chelating drug disulfiram (DSF) exhibits strong affinity for cysteines present in various regulatory proteins. The irreversible mixed disulfides thus formed inactivate the biological activities of several signaling, DNA repair, and anticancer targets, through the same mechanism by which the drug inhibits ALDH and prevents binge drinking. The strong anticancer effects of DSF particularly in combination with copper against gliomas, its long-term use, safety and tolerability, penetration through the blood brain barrier have made it attractive to repurpose the drug for brain cancers. DSF’s ability to Inhibit MGMT, the single most important determinant of glioma resistance to alkylating agents is an additional feature that can increase the efficacy of temozolomide in the trial.\nThe current protocol to evaluate the use Cu-DSF in a non-alcoholic setting for recurrent glioblastoma is well designed and well justified. First, there is no consensus on treatment for GBM recurrence and the high-risk patient selection is appropriate. The therapeutic outcome between the two arms (Cu-DSF with or without alkylators) should be well-comparable. Whether the Cu-DSF by itself will have anti-glioma effects in a clinical setting is not known and difficult to predict. This is because, the copper chelated with DSF [BisN, N-diethyldithiocarmato) copper II], essentially appears to be an inactivated molecule and lacks the ability to form S-conjugates and fails to inhibit proteasome and other targets. The fate of DDC-Cu and its cytotoxic mechanisms in vivo are still unclear. Other comments and suggestions are listed below.\nWhether the protocol uses copper gluconate or another chelated form is not mentioned. There is one study (Proc Amer Assoc Cancer Res. 58, Abstract 4053, 2017) which describes that the efficacy of DSF-Cu is highly dependent on the time (at least 6 h) and schedule (DSF prior to copper) in brain tumor cell cultures.\n\nThere is a good chance of myelosuppression, particularly, in patients taking DSF-Cu and temozolomide or other alkylating agents. Therefore, blood cell counts more frequently than 4-6 weeks is suggested.\n\nCu-DSF is likely to sensitize GBMs to radiation as well. If there are a subset of patients getting radiation alone or in combination with DSF-Cu, it is recommended that the outcomes be assessed as well.\n\nIs the rationale for, and objectives of, the study clearly described? Yes\n\nIs the study design appropriate for the research question? Yes\n\nAre sufficient details of the methods provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Yes", "responses": [] }, { "id": "42298", "date": "03 Jan 2019", "name": "Frantz Rom Poulsen", "expertise": [ "Reviewer Expertise Neurosurgery" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe study protocol describes a randomized controlled trial investigating if addition of disulfiram and copper to the standard Stupp regimen for treatment of glioblastoma will increase survival at 6 months, overall survival, progression free survival etc.The study is limited to patients with first recurrence.\nThe study is well designed and well planned to answer the research questions. It is, however, more and more common practice to perform surgical resection of the recurrence (if possible) prior to additional chemo therapy. I cannot from the study protocol read if this is also the case here and how the authors will deal with this.\nThe RANO criteria will be used in the assessment of progression. Although these criteria are well recognized they may not always be that easy to use and conclude from. Therefore it was very nice, the authors also use volumetric measurements to evaluate growth rate. The authors might, however, consider using other measures than contrast enhancement to assess volumetric growth rate (flair, T2 etc).\n\nIs the rationale for, and objectives of, the study clearly described? Yes\n\nIs the study design appropriate for the research question? Yes\n\nAre sufficient details of the methods provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Not applicable", "responses": [] } ]	1	https://f1000research.com/articles/7-1797
https://f1000research.com/articles/7-704/v1	06 Jun 18	{ "type": "Data Note", "title": "Head models of healthy and depressed adults for simulating the effects of non-invasive brain stimulation", "authors": [ "Nya Mehnwolo Boayue", "Gábor Csifcsák", "Oula Puonti", "Axel Thielscher", "Matthias Mittner", "Nya Mehnwolo Boayue", "Gábor Csifcsák", "Oula Puonti", "Axel Thielscher" ], "abstract": "During the past decade, it became clear that the effects of non-invasive brain stimulation (NIBS) techniques such as transcranial direct current stimulation (tDCS) and transcranial magnetic stimulation (TMS) are substantially influenced by variations in individual head and brain anatomy. In addition to structural variations in the healthy, several psychiatric disorders are characterized by anatomical alterations that are likely to further constrain the intracerebral effects of NIBS. Here, we present high-resolution realistic head models derived from structural magnetic resonance imaging data of 19 healthy adults and 19 patients diagnosed with major depressive disorder (MDD). By using a freely available software package for modelling the effects of different NIBS protocols, we show that our head models are well-suited for assessing inter-individual and between-group variability in the magnitude and focality of tDCS-induced electric fields for two protocols targeting the left dorsolateral prefrontal cortex.", "keywords": [ "transcranial direct current stimulation (tDCS)", "transcranial magnetic stimulation (TMS)", "non-invasive brain stimulation (NIBS)", "major depressive disorder (MDD)", "Head models", "computational modelling", "magnetic resonance imaging (MRI)" ], "content": "Introduction\n\nNon-invasive brain stimulation (NIBS) techniques such as transcranial direct current stimulation (tDCS) and transcranial magnetic stimulation (TMS) have been used to investigate the relationship between activity in different cortical regions and cognitive processes. A key advantage of NIBS is that it allows direct manipulation of neural excitability. Therefore when used carefully, it allows causal interpretation of how specific brain regions might be involved in mental phenomena such as perception, working memory, attention, decision-making or emotional regulation. In addition, NIBS has been used as a clinical tool to obtain symptom reduction in several neurological and psychiatric conditions1–4. Importantly, the same stimulation protocol can result in different neural effects across individuals because the distribution of stimulation-induced electric fields (E-fields) in the brain is strongly contingent on anatomical variability5,6. This can manifest in strong variability in the effects of NIBS on cognitive performance in healthy individuals (e.g., working memory7,8 and on treatment outcomes in patients (e.g., depression9–13).\n\nGiven the heterogeneity in the efficacy of NIBS protocols in modulating behavior and clinical symptoms, there has been a move towards computational modelling of the spatial distribution of E-fields in the brain to understand how stimulation parameters such as electrode placement, rotation or type affects current flow in the neural tissue. After an initial phase during which simplistic spherical head models were used14, the focus has shifted to very detailed, realistic head models created from individual structural magnetic resonance (MR) images using freely available tools (e.g. SimNIBS15 or ROAST16). Creating individual head models remains challenging, however, due to (1) the requirement of high-quality structural MR images for each study participant and (2) the need to manually improve the automatic segmentation of the MR images into the different tissue types (i.e., bone, cerebrospinal fluid, white and grey matter, air, etc). As a consequence, individual head models are rarely used in practice. Instead, researchers usually rely on E-fields induced in a reference individual which is generalized to all participants. This approach, of course, neglects the importance of individual brain anatomy which can have strong influences on E-field distribution5.\n\nTo circumvent this issue, the New York (NY) Head model17 was created from the ICBM152 (v6 template18,19 and v2009b template20), which is a non-linear average of 152 individual MRIs and is extended to the full head and neck region by fusing it with an average of an additional 26 brains. Therefore, this head model represents an unbiased population average and should be a better representation of the individual participants than any randomly picked reference individual. However, calculation of the electric field on a single head model does not allow to quantify variation across individuals that can be substantial both in cortical regions directly below the electrodes and those that are farther away from the stimulation sites. Also, given that the NY Head was created using healthy individuals, possible systematic differences between patient groups may not be noticed.\n\nTo improve this situation, we present 38 head models created from MR images of 19 healthy participants and 19 patients with major depressive disorder (MDD). These head models were manually checked and edited to improve their accuracy in giving a faithful representation of cortical regions. Our head models can aid future research in at least three ways: First, calculating electric fields across a large sample allows to get a sense of the variability of the induced E-field due to anatomical differences. Second, our head models also enable comparison of the neural effects on NIBS protocols in healthy vs. depressed brains. Given the different protocols used for left dorsolateral prefrontal cortex (lDLPFC) targeting with NIBS21,22, computational modelling using these head models will enable the development of protocols for more selective lDLPFC stimulation in this disorder. Third, EEG source localization also relies on head models to calculate the spatial location of possible current sources in the brain. Usually, boundary element models (BEM) are used which have the limitation of less anatomical detail. Therefore, finite element models (FEM) derived from high resolution MR images have become more widespread because they are able to incorporate more tissue types, increasing the precision of EEG source localization. Our head models can be used for source localization using open-source software23. Thus, our head models can help researchers to optimize NIBS protocols and EEG source localization methods, and to test them on a larger sample (including both healthy and patient data).\n\n\nMethods\n\nHigh-resolution head models were created from T1-weighted anatomical images that were collected in a separate functional magnetic resonance imaging (fMRI) study24. The data was obtained from the OpenfMRI database (accession number:ds000171). Structural scans of 19 healthy adult participants with no history of depression or other psychiatric disorders (11 females; mean ± SD age: 28.79 ± 10.86; range: 18–59) and 19 individuals diagnosed with MDD and experiencing a depressive episode at the time of the scanning (11 females; mean ± SD age: 33.52 ± 13.35; range: 18–56) were used. Data of one control participant (’sub-control20’) was excluded due to technical problems with head model creation. Patients did not suffer from current or past manic episodes, current comorbid anxiety disorders or current alcohol dependence or abuse. At the time of data collection, all patients were unmedicated, but 6 received antidepressant pharmacotherapy in the past. For full details regarding demographic data, we refer to the supporting information of the original paper24.\n\nAs the very first step, we inspected scans of all participants, and manually removed signals corresponding to the MRI marker placed on the forehead of each subject using FreeSurfer 5.3.025. Automated tissue segmentation was performed in SPM1226 for skin, skull, eyeballs and CSF, and in FreeSurfer for gray and white matter. Subsequently, segmented images of each participant were visually inspected and manually corrected with FreeSurfer (done by investigator G.Cs., verified by O.P.). Manual corrections were primarily restricted to the skull-CSF boundary, but in some cases also involved the skin-skull interface. In addition, during manual corrections we verified that the segmentation of the cortical gray matter corresponded to the anatomical scans except for medial temporal lobe structures (i.e., the parahippocampal gyrus and the hippocampus proper). Moreover, the resulting masks were not corrected for inconsistencies relating to subcortical nuclei and thus, these head models are not suitable for estimating stimulation-related E-fields in structures such as the thalamus, basal ganglia, amygdala, or brainstem nuclei. Additionally, because the original dataset was de-faced and did not include the neck/shoulder region, our head models do not include these regions. This limits their usability regarding the simulation of tDCS montages with extracephalic return electrodes. Nevertheless, they can be used for all tES protocols using scalp electrodes and most TMS protocols.\n\nHead models were created with an extended, pre-release version of SimNIBS 2.115, a freely available software package for simulating the effects of NIBS techniques. The final head mesh of each participant consisted of a total number of approximately 3,200,000 tetrahedral elements, assigned to six tissue types (Figure 1). For comparability with other datasets, we also report measurements for head size: The distance between the nasion and inion (mean ± SD: 19.2 cm ±1.01 cm; range: 16.7 cm – 21.3 cm) and the distance between the right and left pre-auricular points (mean ± SD: 14.8 cm ±0.65 cm; range: 13.6 cm – 16.3 cm). The total volume of the brain was 1.22 dm3 ±0.11 dm3 (range: 1.02 dm3 – 1.49 dm3). Individual measurements are found at our data repository27\n\nTissue conductivities were set as follows: 0.465 S/m (skin), 0.01 S/m (skull), 0.5 S/m (eyeballs), 1.654 S/m (cerebrospinal fluid), 0.275 S/m (gray matter), 0.126 S/m (white matter). Although our head models do not account for white matter anisotropy, this property has been shown to primarily influence current density in deeper structures, leaving superficial gray matter relatively unaffected28. The accuracy of tissue segmentation and the good correspondence between anatomical scans and the resulting head models for 8 individuals are shown in Figure 2.\n\n\nDataset validation\n\nExcept for two manual steps (removal of MRI markers from the forehead, manual correction of tissue segmentations), the process of head model creation was automated using a pre-release version of SimNIBS 2.1 that employed FreeSurfer 5.3.0 for brain segmentation (as described in 29) and SPM12 for segmentation of the remaining tissues (similarly to 30). This new SimNIBS pipeline provides more accurate tissue segmentation relative to other protocols. The scripts used for automated simulation of tDCS effects for all head models (as shown in Figure 3) are available for download at our data repository27.\n\nPlease note the large degree of variability in E-field magnitude (both protocols) and in the lateralization of effects (bipolar protocol).\n\nFor validating the reliability of our head models, we compared the effects of two tDCS protocols targeting the lDLPFC (one conventional bipolar montage and one multi-electrode 4×1 setup) against the effects observed in the NY Head17. The mesh for the NY Head (abaqus format; https://www.parralab.org/nyhead/) has been reformatted to be compatible with SimNIBS and is also available for download in our data repository27. For each head model, tDCS electrodes of appropriate size (bipolar montage: 5 × 5 cm, circular connectors (diameter: 0.5 cm) at the middle of the electrode pads; 4×1 montage: diameter of 1.2 cm) and thickness (1 mm for all electrodes + 2.5 mm sponge pocket/gel layer for the bipolar and 4×1 montages, respectively) were placed at scalp locations corresponding to electrode positions of the International 10/10 system (bipolar montage: anode - F3, cathode - F4; 4×1 montage: anode: F3, cathodes: C3, FT7, Fp1, Fz). Stimulation intensity for the anode was set to 2 mA, with equal distribution of return currents for the 4 cathodes (0.5 mA for each) in the 4×1 protocol. Results of the simulations were visualized using Gmsh31 (Figure 3).\n\nIn our previous study6, we reported stronger stimulation-induced E-fields in the lDLPFC for the bipolar montage used by Brunoni and colleagues (2013)9 than for the 4×1 protocol, albeit the bipolar montage was also associated with reduced focality (i.e., more intensive stimulation of other cortical areas). Therefore, we extracted three measures for both tDCS protocols for our 38 head models: stimulation strength (the norm of the electric field vector, ’normfield’) in the lDLPFC was assessed by extracting individual mean (calculated across all nodes in this region) and maximum (peak) E-field values, whereas spatial focality of the stimulation was analyzed by calculating the focality-index (FI), with the target region as reference. FI was quantified as the proportion of highest-intensity nodes (nodes within the upper 1% percentile of all E-field values) in the lDLPFC relative to the whole cortex. Mean and peak E-field values corresponding to both tDCS montages were extracted by reconstructing the two-dimensional cortical surface (more precisely, the middle of the cortical sheet) of each individual along with the corresponding E-field cortical map in FreeSurfer, and applying automated atlas-based parcellation of the frontal lobe to delineate the lDLPFC region in each brain32.\n\nAs a result, we show that (1) both protocols induce strong E-fields in the DLPFC (with symmetrical effects for the bipolar montage and unilateral E-field distribution for the 4×1 protocol), (2) effects are similar for our head models and for the NY Head, (3) all E-field measures (peak and mean strength, FI) show great degree of variability, and (4) montage-specific effects are consistent with previous results reported in the literature regarding both the spatial distribution and the magnitude of E-fields33–37 (Figure 3, Figure 4).\n\nBoth peak (left panel) and mean (middle panel) E-field magnitudes in the left dorsolateral prefrontal cortex (lDLPFC) are stronger for the bipolar montage, whereas the 4×1 protocol yields more focal stimulation of the target region (right panel). Group means (red: healthy individuals, green: major depressive disorder patients) do not differ substantially, but large degree of inter-individual variability can be observed in both groups. The triangles show data for the New York Head. Horizontal lines within boxes represent median values, whereas lower and upper box hinges correspond to the first and third quartiles (25th and 75th percentiles). Lengths of upper/lower whiskers extend to the largest/smallest values that do not exceed 1.5* the inter-quartile range; dots represent individual data.\n\nFor detailed discussion of the results presented in Figure 4 we refer the reader to our accompanying paper6.\n\n\nData availability\n\nOpen Science Framework (OSF): Dataset 1. Head models of healthy and depressed adults for simulating the effects of non-invasive brain stimulation, http://doi.org/10.17605/OSF.IO/EXBD527\n\nLicense: CC0 1.0 Universal\n\nAt our data repository27, the following data are available for download for all subjects (healthy adults: ’sub-control’, patients: ’sub-mdd’): T1-weighted anatomical scans registered to FreeSurfer conform space (’nii.gz’ files), the corresponding segmentation masks for the 6 tissue types (’nii.gz’ files), the final head models (’msh’ files), the files containing electrode coordinates (International 10/10 system) for all participants (’txt’ files can be used for performing new script-based simulations, whereas the ’geo’ files can be used to plot electrode positions directly onto the mesh files in Gmsh31), and files organized in 2 folders (’fs_.tar.gz’ and ’m2m_.tar.gz’) that enable creating simulation outputs in average FreeSurfer space (’fsaverage’). We also included a README file with a detailed description of the data and scripts.\n\nOur head models are compatible with SimNIBS 2.0 (http://simnibs.de/) for simulating the effects of tDCS and TMS protocols. This software package has an easy-to-use graphical user interface (GUI) for setting all stimulation parameters (scalp location, intensity, etc.) for both NIBS techniques. By using our custom-written script27, it is also possible to run tDCS simulations for any given montage for all participants at once. Although the current version of SimNIBS does not allow exporting the resulting E-field maps into FreeSurfer, this option will be available in an upcoming version. This will enable registering data to an average surface (’fsaverage’) and creating group averaged data, as we have shown previously6. In addition, using the upcoming new version of SimNIBS, researchers will have the opportunity to extract E-field components that are either radial (normal) or tangential relative to the cortical surface, and have been associated with different cellular effects38. At our data repository27 we also provide the manually corrected segmentations for the different tissue types for those who would like to create high-quality meshes of their own using open-source software such as iso2mesh39. Finally, our meshes can be used for improving the anatomic precision of EEG source localization, using open-source tools23.", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by the Northern Norway Regional Health Authority (grant no. PFP1237-15) for G.Cs. and M.M. and by the Lundbeck foundation (grant no. R118-A11308) and the Novonordisk foundation (grant no. NNF14OC0011413) for O.P. and A.T. The publication charges for this article have been funded by a grant from the publication fund of UiT The Arctic University of Norway.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nRidding MC, Rothwell JC: Is there a future for therapeutic use of transcranial magnetic stimulation? Nat Rev Neurosci. 2007; 8(7): 559–67. PubMed Abstract \| Publisher Full Text\n\nKuo MF, Paulus W, Nitsche MA: Therapeutic effects of non-invasive brain stimulation with direct currents (tDCS) in neuropsychiatric diseases. NeuroImage. 2014; 85 Pt 3: 948–960. PubMed Abstract \| Publisher Full Text\n\nFlöel A: tDCS-enhanced motor and cognitive function in neurological diseases. NeuroImage. 2014; 85 Pt 3: 934–947. PubMed Abstract \| Publisher Full Text\n\nDi Lazzaro V, Profice P, Pilato F, et al.: Motor cortex plasticity predicts recovery in acute stroke. Cereb Cortex. 2009; 20(7): 1523–1528. PubMed Abstract \| Publisher Full Text\n\nOpitz A, Paulus W, Will S, et al.: Determinants of the electric field during transcranial direct current stimulation. NeuroImage. 2015; 109: 140–150. PubMed Abstract \| Publisher Full Text\n\nCsifcsák G, Boayue NM, Puonti O, et al.: Effects of transcranial direct current stimulation for treating depression: A modeling study. J Affect Disord. 2018; 234: 164–173. PubMed Abstract \| Publisher Full Text\n\nBrunoni AR, Vanderhasselt MA: Working memory improvement with non-invasive brain stimulation of the dorsolateral prefrontal cortex: a systematic review and meta-analysis. Brain Cogn. 2014; 86: 1–9. PubMed Abstract \| Publisher Full Text\n\nHill AT, Fitzgerald PB, Hoy KE: Effects of anodal transcranial direct current stimulation on working memory: a systematic review and meta-analysis of findings from healthy and neuropsychiatric populations. Brain Stimul. 2016; 9(2): 197–208. PubMed Abstract \| Publisher Full Text\n\nBrunoni AR, Valiengo L, Baccaro A, et al.: The sertraline vs. electrical current therapy for treating depression clinical study: results from a factorial, randomized, controlled trial. JAMA psychiatry. 2013; 70(4): 383–391. PubMed Abstract \| Publisher Full Text\n\nBlumberger D, Tran L, Fitzgerald P, et al.: A randomized double-blind sham- controlled study of transcranial direct current stimulation for treatment-resistant major depression. Front Psychiatry. 2012; 3: 74. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBennabi D, Nicolier M, Monnin J, et al.: Pilot study of feasibility of the effect of treatment with tDCS in patients suffering from treatment-resistant depression treated with escitalopram. Clin Neurophysiol. 2015; 126(6): 1185–1189. PubMed Abstract \| Publisher Full Text\n\nLoo CK, Sachdev P, Martin D, et al.: A double-blind, sham-controlled trial of transcranial direct current stimulation for the treatment of depression. Int J Neuropsychopharmacol. 2010; 13(1): 61–69. PubMed Abstract \| Publisher Full Text\n\nLoo CK, Alonzo A, Martin D, et al.: Transcranial direct current stimulation for depression: 3-week, randomised, sham-controlled trial. Br J Psychiatry. 2012; 200(1): 52–59. PubMed Abstract \| Publisher Full Text\n\nMiranda PC, Lomarev M, Hallett M: Modeling the current distribution during transcranial direct current stimulation. Clin Neurophysiol. 2006; 117(7): 1623–1629. PubMed Abstract \| Publisher Full Text\n\nThielscher A, Antunes A, Saturnino GB: Field modeling for transcranial magnetic stimulation: A useful tool to understand the physiological effects of TMS? Conf Proc IEEE Eng Med Biol Soc. 2015; 2015: 222–225. PubMed Abstract \| Publisher Full Text\n\nHuang Y, Datta A, Bikson M, et al.: Realistic volumetric-approach to simulate transcranial electric stimulation–roast–a fully automated open-source pipeline. bioRxiv. 2017; 217331. Publisher Full Text\n\nHuang Y, Parra LC, Haufe S: The New York Head-A precise standardized volume conductor model for EEG source localization and tES targeting. NeuroImage. 2016; 140: 150–162. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFonov VS, Evans AC, McKinstry RC, et al.: Unbiased nonlinear average age-appropriate brain templates from birth to adulthood. NeuroImage. 2009; 47(Suppl 1): S102. Publisher Full Text\n\nFonov V, Evans AC, Botteron K, et al.: Unbiased average age-appropriate atlases for pediatric studies. NeuroImage. 2011; 54(1): 313–327. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGrabner G, Janke AL, Budge MM, et al.: Symmetric atlasing and model based segmentation: an application to the hippocampus in older adults. Med Image Comput Comput Assist Interv. Springer, 2006; 9(Pt 2): 58–66. PubMed Abstract \| Publisher Full Text\n\nFitzgerald PB, Maller JJ, Hoy KE, et al.: Exploring the optimal site for the localization of dorsolateral prefrontal cortex in brain stimulation experiments. Brain Stimul. 2009; 2(4): 234–237. PubMed Abstract \| Publisher Full Text\n\nDe Witte S, Klooster D, Dedoncker J, et al.: Left prefrontal neuronavigated electrode localization in tDCS: 10-20 EEG system versus MRI-guided neuronavigation. Psychiatry Res Neuroimaging. 2018; 274: 1–6. PubMed Abstract \| Publisher Full Text\n\nZiegler E, Chellappa SL, Gaggioni G, et al.: A finite-element reciprocity solution for EEG forward modeling with realistic individual head models. NeuroImage. 2014; 103: 542–551. PubMed Abstract \| Publisher Full Text\n\nLepping RJ, Atchley RA, Chrysikou E, et al.: Neural Processing of Emotional Musical and Nonmusical Stimuli in Depression. PLoS One. 2016; 11(6): e0156859. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFischl B, Sereno MI, Dale AM: Cortical surface-based analysis. II: Inflation, flattening, and a surface-based coordinate system. NeuroImage. 1999; 9(2): 195–207. PubMed Abstract \| Publisher Full Text\n\nFriston KJ, Holmes AP, Worsley KJ, et al.: Statistical parametric maps in functional imaging: a general linear approach. Hum Brain Mapp. 1994; 2(4): 189–210. Publisher Full Text\n\nBoayue NM, Csifcsák G, Puonti O, et al.: Dataset 1. Head models of healthy and depressed adults for simulating the effects of non-invasive brain stimulation. 2018. Data Source\n\nWagner S, Rampersad SM, Aydin Ü, et al.: Investigation of tDCS volume conduction effects in a highly realistic head model. J Neural Eng. 2014; 11(1): 016002. PubMed Abstract \| Publisher Full Text\n\nWindhoff M, Opitz A, Thielscher A: Electric field calculations in brain stimulation based on finite elements: an optimized processing pipeline for the generation and usage of accurate individual head models. Hum Brain Mapp. 2013; 34(4): 923–935. PubMed Abstract \| Publisher Full Text\n\nNielsen JD, Madsen KH, Puonti O, et al.: Automatic skull segmentation from MR images for realistic volume conductor models of the head: Assessment of the state-of-the-art. NeuroImage. 2018; 174: 587–598. PubMed Abstract \| Publisher Full Text\n\nGeuzaine C, Remacle JF: Gmsh: A 3-d finite element mesh generator with built-in pre- and post- processing facilities. Int J Numer Meth Eng. 2009; 79(11): 1309–1331. Publisher Full Text\n\nRanta ME, Chen M, Crocetti D, et al.: Automated MRI parcellation of the frontal lobe. Hum Brain Mapp. 2014; 35(5): 2009–2026. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBai S, Dokos S, Ho KA, et al.: A computational modelling study of transcranial direct current stimulation montages used in depression. NeuroImage. 2014; 87: 332–344. PubMed Abstract \| Publisher Full Text\n\nSeibt O, Brunoni AR, Huang Y, et al.: The Pursuit of DLPFC: Non-neuronavigated Methods to Target the Left Dorsolateral Pre-frontal Cortex With Symmetric Bicephalic Transcranial Direct Current Stimulation (tDCS). Brain Stimul. 2015; 8(3): 590–602. PubMed Abstract \| Publisher Full Text\n\nLaakso I, Tanaka S, Mikkonen M, et al.: Electric fields of motor and frontal tDCS in a standard brain space: A computer simulation study. NeuroImage. 2016; 137: 140–151. PubMed Abstract \| Publisher Full Text\n\nOpitz A, Falchier A, Yan CG, et al.: Spatiotemporal structure of intracranial electric fields induced by transcranial electric stimulation in humans and nonhuman primates. Sci Rep. 2016; 6: 31236. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHuang Y, Liu AA, Lafon B, et al.: Measurements and models of electric fields in the in vivo human brain during transcranial electric stimulation. eLife. 2017; 6: pii: e18834. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRahman A, Reato D, Arlotti M, et al.: Cellular effects of acute direct current stimulation: somatic and synaptic terminal effects. J Physiol. 2013; 591(10): 2563–2578. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFang Q, Boas DA: Tetrahedral mesh generation from volumetric binary and grayscale images. In Biomedical Imaging: From Nano to Macro, 2009. ISBI’09. IEEE International Symposium on. Ieee, 2009; 1142–1145. Publisher Full Text" }	[ { "id": "34789", "date": "18 Jun 2018", "name": "Ilkka Laakso", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nNon-invasive brain stimulation techniques are based on generating, either magnetically or electrically, electric fields that can alter brain neuronal activity. However, the generated electric fields can vary greatly depending on the individual anatomy of the scalp, skull and brain. This data note presents a collection of 38 individual head models (both healthy and patients) that can be used for characterisation of variability in the electric fields.\nHead models were constructed from T1-weighted MRI using freely available software: FreeSurfer, SPM12 and SimNIBS. Manual verification and corrections were applied when necessary. The approach is state of art.\nThe datasets are provided in the Gmsh format, which can be used directly in modelling software or converted to multiple other formats using open-source software. Furthermore, raw data, including FreeSurfer subject data, are included, allowing further processing and adaptation of the data.\n\nMinor comments:\n1. In the text and Figures 1 and 2, it is written that the head models are segmented to six tissue types. Actually, it seems that there are more than six tissue compartments, as cerebellar GM and cerebellar WM are separate from the cerebral GM and WM. Also, at least some air cavities are segmented (missing from the caption of figure 2).\n2. Brainstem is segmented as cerebellar white/grey matter. It may be helpful to list this as a limitation, for instance, in \"Creation of head models\".\n3. In the abstract, sentence \"... effects of non-invasive brain stimulation ...\", and in \"Dataset validation\", sentence \"The scripts used for automated simulation of tDCS effects for all head models ...\". Electric fields are not \"effects\".\n\nIs the rationale for creating the dataset(s) clearly described? Yes\n\nAre the protocols appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and materials provided to allow replication by others? Yes\n\nAre the datasets clearly presented in a useable and accessible format? Yes", "responses": [ { "c_id": "4090", "date": "15 Nov 2018", "name": "Nya Boayue", "role": "Author Response", "response": "We would like to thank Dr. Laakso for reviewing our manuscript and for the helpful comments. These comments have been addressed in the new version of the manuscript. We have also updated the scripts in our data repository such that they are compatible with the currently released SimNIBS (version 2.1.1). Thanks for pointing out this issue. We have now added air cavities to the caption of Figure 2. Concerning the discrepancy between the segmentations and the final head models, we have now added the following clarification to the creation of head models section “The initial segmentation included more than 6 tissue compartments (e.g., separate tissue types for cerebellar gray and white matter; available in the m2m_sub-* folders) but they were later combined into one of 6 tissue types: skin, skull, CSF, GM, GM and eyeballs in the final head models for simulation purposes. In addition, air cavities were modeled by not adding tetrahedra to these locations, similar to the air surrounding the head.” This is an important point. We have now added the following sentence to the creation of head models section “Additionally, the segmentation of the brainstem is not accurate because it arbitrarily assigned brain tissue to white and grey matter. We concur with this point. The abstract sentence now reads “… the electric field elicited by non-invasive brain stimulation ...” The \"Dataset validation\" sentence which was changed a bit now reads “ We provide scripts compatible with SimNIBS 2.1.1 for automated simulation of tDCS-induced electric fields for all head models ...”" } ] }, { "id": "37428", "date": "17 Sep 2018", "name": "Yu Huang", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a modeling effort trying to predict how electric field distributes under trascranial electric stimulation for up to 19 normal and 19 pathological subjects. It is well written and a good contribution to the literature. The authors are advised to address the following issues before indexing:\n1) Only transcranial electric stimulation (TES) is done in this work. There is no simulation on magnetic stimulation (TMS). So the title of this manuscript should be “non-invasive electric brain stimulation”\n2) The authors only discussed there are inter-individual variabilities in E-field distributions, but did not say anything about how the E-field differs (or is similar) between the healthy subjects and depression subjects, and why.\n3) As shown in Figure 1, the head models are all cut off without the lower part of the head. This will give significant difference in the E-field distributions compared to a head model that covers the entire head (Huang and Liu, et al, 20171). Please discuss this as one limitation of this work.\n4) There are results in the Section Methods. So the section name should be “Methods and Results”.\n5) It is unclear that how the lDLPFC region was separated. You only mentioned “...applying automated atlas-based parcellation of the frontal lobe to delineate the lDLPFC region in each brain”. Please briefly describe how this was done. Was some atlas registered to each individual brain to extract the region?\n6) You said SimNIBS 2.1 was used, but did not mention which function was used for the segmentation. In SimNIBS 2.1, “headreco” calls SPM12 to segment all the head tissues, and “mri2mesh” uses FreeSurfer to segment gray and white matter, and FSL to segment non-brain tissues. You said “...SimNIBS 2.1 that employed FreeSurfer 5.3.0 for brain segmentation (as described in 29) and SPM12 for segmentation of the remaining tissues (similarly to 30).” This is not clear to me. Did you combine “headreco” and “mri2mesh”? Also if you used SPM12, there should be a tissue type “air cavities”, but from Figure 1, this “air cavities” is not there.\n7) The thickness of electrodes in your models “1 mm for all electrodes + 2.5 mm sponge pocket/gel layer for the bipolar and 4×1 montages, respectively” is not clear. Please clarify this sentence.\n8) Minor:\nIn “Introduction” section, the first 3 sentences in the first paragraph are general statements but lack references. Please add. In “Introduction” section, 2nd paragraph, what do you mean by “rotation or type”?\n\nIs the rationale for creating the dataset(s) clearly described? Yes\n\nAre the protocols appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and materials provided to allow replication by others? Partly\n\nAre the datasets clearly presented in a useable and accessible format? Yes", "responses": [ { "c_id": "4091", "date": "15 Nov 2018", "name": "Nya Boayue", "role": "Author Response", "response": "We are grateful to Dr. Huang for reviewing our manuscript. Your comments have helped give the revised version of our manuscript more clarity. We have addressed all the points raised in the new version of the MS. We have also updated the scripts at our data repository such that they are compatible with the currently released SimNIBS (version 2.1.1). The title of the new manuscript has been changed based on your suggestion. We chose to keep the focus here on the data since this is a data note. However, we have added the following to the Dataset validation section. “Accessing group differences between MDD and healthy subjects was not the primary aim of the current data note, however, analysis of the spatial distribution of tDCS-induced E-fields in the bilateral DLPFC and medial prefrontal cortex showed subtle group differences between the healthy and MDD groups. For detailed discussion of these results and that of Figure 4 we refer the reader to our accompanying paper https://www.sciencedirect.com/science/article/pii/S0165032717324746.” Many thanks for this very important point and pointing us to the article. This has now been addressed in our manuscript as one of the limitations of using our head models in the creation of head models section . “... an extended head model with field of view covering the entire head would further increase the predictive accuracy of the head models https://elifesciences.org/articles/35178” The Methods section has been renamed to “Methods and Results”. Through personal communication with the authors of https://onlinelibrary.wiley.com/doi/full/10.1002/hbm.22309, we received a script which was used for the automatic parcellation of the frontal lobe. We have now clarified this aspect: “an automated atlas-based parcellation of the frontal lobe https://onlinelibrary.wiley.com/doi/full/10.1002/hbm.22309 was applied to each individual brain to delineate the lDLPFC.” This is a good point. The reason for the lack of clarity is that we used a custom version of the software which, at the time, resulted in better results than using the latest released version. In the meantime, a newer SimNIBS version has been released. We have added the following detail to the MS to avoid any misunderstandings. “… the process of head model creation was automated using a custom version of SimNIBS 2.1 that employed FreeSurfer 5.3.0 for brain segmentation (as described in https://onlinelibrary.wiley.com/doi/full/10.1002/hbm.21479 and implemented in mri2mesh) and SPM12 for segmentation of the remaining tissues (similarly to https://www.sciencedirect.com/science/article/pii/S1053811918301800 and implemented in headreco). This pipeline provides more accurate tissue segmentation relative to other protocols. It was a custom pipeline developed before the official release of SimNIBS 2.1. However, using headreco combined with the CAT12 toolbox (included with SimNIBS 2.1.1 ) for cortical reconstruction, the same accuracy can be achieved.”As for the lack of “air cavities” in the Figure 1, we have added the following clarification in the creation of head models section “… air cavities were modeled by not adding tetrahedra to these locations, similar to the air surrounding the head.” This sentence has now been clarified. It reads “ 1 mm for all electrodes + a sponge pocket of 2.5 mm thickness for the bipolar montages and a gel layer of 2.5 mm thickness for the 4x1 montages.” We have now added the required references. We meant “electrode rotation or electrode type”. this has been corrected in the MS." } ] } ]	1	https://f1000research.com/articles/7-704
https://f1000research.com/articles/7-1795/v1	15 Nov 18	{ "type": "Research Article", "title": "Postoperative pain and antibacterial effect of 980 nm diode laser versus conventional endodontic treatment in necrotic teeth with chronic periapical lesions: A randomized control trial", "authors": [ "Dina A. Morsy", "Maged Negm", "Alaa Diab", "Geraldine Ahmed", "Maged Negm", "Alaa Diab", "Geraldine Ahmed" ], "abstract": "Background: Many challenges encounter the endodontist, especially when dealing with necrotic teeth with chronic periapical lesions. Postoperative pain may be induced following conventional endodontic therapy and total eradication of the bacteria is almost unachievable even with recently available techniques. In recent years, diode laser usage in the endodontic field has gained acceptance. Thus, this study aimed to investigate the ability of the diode laser (DL) to decrease postoperative pain and achieve root canal sterility. Methods: 56 patients with anterior teeth with chronic periapical lesions in upper anterior teeth were randomly divided into two groups (n = 28). All patients were treated with two visits of conventional root canal treatment with ProTaper Universal. The DL group: root canals were irradiated with 200 µm fiber optic at both visits; the control group (Endo): the DL fiber was placed in root canal with no activation. Bacterial samples were collected from all the cases at each step of the treatment. Pain levels were evaluated using a numerical rating scale preoperatively, and after 6, 12, 24, 48 hours and 7 days. Bacterial count was used to detect both aerobic and anaerobic bacterial load. Results: The qualitative pain scores revealed statistically significant lower pain levels in the DL group compared with the Endo group at all time intervals (P<0.001), except preoperatively where there was no significant difference. There was a statistically significant lower bacterial count for both aerobic and anaerobic bacteria in the DL group compared with the Endo group in both S3 samples (after laser application) and S4 samples (bacterial colonization) (P<0.001). Conclusion: The 980 nm diode laser may be a successful adjunct to conventional endodontic treatment of necrotic cases with chronic periapical lesions in terms of postoperative pain and root canal disinfection.\n\nTrail registration: PACTR201511001275414 (date: 23rd September 2015)", "keywords": [ "Diode laser", "intracanal irradiation", "necrotic pulp", "periapical lesion", "endodontic" ], "content": "Introduction\n\nThe therapeutic goal of endodontic treatment in cases of necrotic teeth with chronic periapical lesions is the creation of a sterile, bacteria-free environment in the tooth and at the apex, including the periodontal tissue and the surrounding apical bone1. There are two main complicating factors preventing achievement of this goal: the complicated anatomical root configuration and the special characteristics of the resident bacterial flora, which makes it sometimes inaccessible even with recent available armamentarium2. In recent years, intracanal laser irradiation has been used in root canal preparation3, gaining acceptance for its disinfection ability as adjunct to the conventional mechanical instrumentation and irrigation protocols available4,5. It was also reported that the use of laser therapy may result in decreased postoperative pain6. Recently, the diode laser (DL) 980 nm was introduced and its use in root canal disinfection was established4.\n\nThis study aimed to evaluate the ability of the 980 nm DL to reduce the post-operative pain and intracanal bacteria when compared with conventional endodontic treatment.\n\n\nMethods\n\nThis study is a parallel randomized controlled trial, with an allocation ratio of 1:1 and a superiority framework.\n\nThis study was approved by the Research Ethics Committee of the Faculty of Oral and Dental Medicine, Cairo University (15-9-19).\n\nThis article was written in concordance with the CONSORT checklist 2010 (Supplementary File 1).\n\nPatients received in the outpatient clinic of the Endodontic Department, Faculty of Dentistry, Cairo University in the duration between March 2016 and March 2017 were invited to participate. In total 56 participants were included in the study after fulfilling the inclusion criteria and after signing an informed consent.\n\nInclusion criteria: Adult patients with average age between 18 and 35 years; medically free patients; patients suffering from necrotic pulp in maxillary central incisors permanent teeth with: closed apex, associated with or without sinus tract, radiographic evidence of periapical radiolucency; patients with healthy dental and periodontal status; positive patients’ acceptance for participation in the study.\n\nExclusion criteria: Illiterate patients; pregnant woman (as the hormonal changes may alter the pain perception); patients having systemic disorder; teeth that have: open apex, extra coronal restorations, greater than grade I mobility, pocket depth greater than 4 mm, non restorable tooth, previous endodontic treatment; patients taking analgesics 12 hours before the intervention; patients who had received antibiotics in the last month; patients with acute pain at the time of intervention.\n\n\nSample size\n\nTo assess DL versus conventional endodontic regarding the postoperative pain, an independent t test was done. It was estimated that a total of 50 patients would be required for the detection of a difference between groups using a two-tailed α of 0.05 and a power of 0.80 if the absolute difference in periapical lesions is 0.37 mm with SD 0.46 as reported in Markovick et al. in 2006. To compensate losses during the follow-up this number should be increased to 56 patients (10% more than the calculated). Sample size was calculated using G* Power program (2).\n\nAll patients were treated by a single endodontic over two visits after signing an informed consent.\n\nFirst visit. At the first visit, all patients recorded their pain level preoperatively using a numerical rating scale (NRS; see below for details). The teeth were locally anaesthetized (Articaine in 4% solution with epinephrine in concentration of 1:100000 (Ultracaine- dental forte, Germany)\n\nBefore isolation, antisepsis of the oral cavity was performed by rinsing for 1 min with 10 mL chlorhexidine gluconate mouth wash 0.2 %. The teeth were properly isolated with rubber dam1. An access cavity was performed. Patency of the root canal was obtained using stainless steel hand k- files size #15 (MANI- MANI, INC. Industrial Park, Utsunomiya, Tochigi, Japan). The root canals were irrigated with 1 ml sterile saline solution. The first microbial samples (S1) were collected to assess the initial colonizers of the root canals using 3 sterile paper points which were inserted into the root canal for 1 min each with pumping movements. They were immediately placed inside sterile tubes containing a reduced transport medium of thioglycolate. Working length was determined using an electronic apex locator (DENTA PORT ZX (J.Morita, Irvine, Japan)), then confirmed with intraoral periapical radiograph. Mechanical preparation was performed with the ProTaper Universal Ni Ti system up to #F4 file for all the cases. In total 10 ml of 2.5% sodium hypochlorite was used for irrigation between each file and the next using a 25-gauge needle. 5 ml of 17% EDTA (Calix E, DHARMA research, Miami, USA) was used at the end of the procedure to remove the smear layer. 5 ml of saline solution was the final irrigant used to neutralize all the previously used solutions. The second microbial samples (S2) representing the antibacterial effect of the mechanical preparation were obtained with the same procedure as the (S1) samples.\n\nAccording to the randomization and sequence generation, the patients were allocated into two groups (n = 28/group).\n\nExperimental (DL) group: Root canals were irradiated with 980 nm diode laser coupled with optical fiber 200 µm (Lite medics, Italy) with setting 1.2-watt power, in pulsed mode. The irradiation protocol was a 5 sec irradiation followed by a 10 sec pause, which constituted one lasing cycle. The lasing cycle was performed four times for each tooth. The tip was positioned 1 mm short of the apex. This was followed by activation during which it was slowly dragged at a speed of approximately 2 mm/ sec in a way that the root canals were irradiated from the apical to the coronal portion, in a helicoidal movement touching the canal walls. This was done to ensure equal diffusion of light inside the root canal lumen (Figure 1). The third microbial samples (S3) were collected as mentioned before to evaluate the effect of DL on the bacterial count.\n\nControl (Endo) group: Conventional endodontic treatment as above, after which the fiber optic was placed inside the root canals without activation (placebo) with no bacterial sample collected at this stage.\n\nAt the end of the first visit a piece of sterile cotton was placed in the pulp chamber and all teeth was dressed with intermediate restorative material (IRM) as a temporary filling (Dentsply, Latin America)\n\nSecond visit. At the second visit, one week later; the canals of both groups were accessed under rubber dam. Canals were irrigated with 1 ml sterile saline solution. The fourth microbial samples (S4) were taken to assess the recolonization of the bacteria and were collected from both groups. The same procedures of the first visit, the intracanal irradiation for the DL group and placebo for the Endo group was performed. The fifth microbial samples (S5) were collected to assess the status of root canal just before the obturation.\n\nEach root canal was obturated with the modified single cone technique using the ProTaper Universal gutta-percha points size F4 and gutta-percha points size 25 (META, Biomed, Republic of Korea) as auxiliaries with ADSEAL resin-based root canal sealer (META BIOMED CO., LTD. Chungbuuk, Korea). All the teeth were restored with IRM as a temporary filling. At the end of the second visit, all patients were instructed to record pain level on the pain scale chart after 6, 12, 24 and 48 hours and after 7 days. The patients were instructed to submit the pain scale charts after the 7th day. Patients were referred for final restorations. Any patient who reported the intake of an analgesic during this period was excluded from the study.\n\nThe NRS consisted of a line anchored by two extremes \"No pain\" and \"the worst pain\". The patients were asked to mark the chart at the point that represented their level of pain from 0 to 10.\n\nPain level was assigned to one of 4 categorical scores: No pain (0), Mild (1–3), Moderate (4–6) and Severe (7–10).\n\nThe bacterial count method was used2. Once the samples arrived to the microbiology department, Cairo University, the tubes containing the thioglycolate (transport medium) (Thioglycollate broth U.S.P alternative, Oxoid microbiology product, LTD, England) with the paper points were placed in microcentrifuge and vortexed for 30 sec. 100 µl aliquots of the vortexed samples were placed in a new sterile tube containing 1 ml of thioglycolate to obtain 1/10 concentration to assess the microbial load of common aerobes and anaerobes found in each root canal. However, no attempt was made to identify the specific microbial flora during the process. The effect of the treatment in each group, the mechanical preparation (S2), the Diode laser irradiation (S3) in the DL group only, the bacterial recolonization (S4) and the bacterial count just before obturation (S5), were compared.\n\nAerobic bacterial culture: 50 µl of these diluted samples were transferred to BHI agar plates (Oxoid microbiology product, LTD, England) and cultured under aseptic conditions, followed by incubation at 37o C for 24 hours for the aerobic bacteria. The number of bacterial colonies in each plate was counted and reported as colony forming units per milliliter (CFU/ml).\n\nAnaerobic bacterial culture: The other 50 µl of these diluted samples were transferred to BHI agar plates under aseptic conditions, the agar plates were placed in an anaerobic sealed jar with Gas-Pak (Gas-Pak system) (Oxoid microbiology product, LTD, Basingstoke, Hants, England) and anaerobic indicator (Anaerobic indicator, BR0055B.Oxoid Ltd. Basingstoke, Hants, England) were incubated for 48 hours at 37o C. Eventually, the number of bacterial colonies in each plate was counted and reported as CFU/ml.\n\nDouble blinding was implemented in this study by the assessor and the statistician concerning evaluation of the post-operative pain intensity and microbiological evaluation. However, single blinding was implemented only by the statistician concerning the results of periapical lesion size. The blinded assessors were asked to fill a chart for each outcome with the number corresponding to each patient without knowing which group the participants were related to.\n\nA random sequence was generated by computer software (http://www.random.org/) in the Center of Evidence Based Dentistry, Cairo University. The table was kept with the assistant supervisor. Four-folded numbered papers were packed in opaque sealed envelopes to be chosen by the patients after entering the study. The opaque envelopes contained the numbers of each random sequence to assign the patient to either the experimental (DL) or control group (Endo).\n\nThe assistant supervisor assigned the participants to the experimental or control groups according to the randomization table. After confirmation on patient eligibility with the assistant, the operator applied the treatment procedure assigned to that patient.\n\nStatistical analysis was performed using SPSS 19 (SPSS, Chicago, IL, USA). As data related to patients’ age and bacterial colony formation were parametric, significance of the difference between both groups was evaluated using unpaired t test. Chi square test was used to compare the qualitative pain scores. The level of significance was set at P<0.05.\n\n\nResults\n\nIn total, 56 patients were included in the study. A CONSORT flow diagram can be seen in Supplementary File 2.\n\nDemographic data, age and gender, had no significant difference between the two groups (P=0.1967 and 0.053, respectively; Table 1 and Table 2).\n\nSD= standard deviation, Min= minimum, Max= Maximum, t=unpaired t test, : p<0.05 (Significant), ns: P> 0.05 (not significant)\n\nX2: Chi square test, Significance level: P<0.05, ns: non-significant. SD= standard deviation, Min= minimum, Max= Maximum, t=unpaired t test, : p<0.05 (Significant), ns: P> 0.05 (not significant)\n\nThe qualitative pain scores revealed statistically significant lower pain levels in the DL group compared with the Endo group at 6, 12 and 24 hours (P<0.001), and at 48 hours and 7 days (P=0.002 and 0.044, respectively), while preoperatively there was no statistical significance difference (P=1.0) (Figure 2 and Table 3). The results of the bacterial count of both aerobic and the anaerobic bacteria of the DL group showed statistically significant reduction in the bacterial count than the Endo group at S4 and S5 (Aerobic: P< 0.01 and 0.002, respectively; anaerobic: P< 0.002 and 0.012, respectively; Figure 3 and Table 4 and Table 5).\n\nMeans with different small letters in the same column indicate statistically significance difference, means with different capital letters in the same row indicate statistically significance difference. ; significant (p<0.05), ns; non-significant (p>0.05).\n\nDL: Diode laser group, ES: endosurgery group, hrs: hours, n: number of patients, %: percentage of patients, mod: moderate pain, SD, standard deviation. X2: Chi square test, significance level: P<0.05, significant\n\nLine chart showing bacterial count (a) aerobic and (b) anaerobic. Means with different small letters in the same column indicate statistically significance difference, means with different capital letters in the same row indicate statistically significance difference. *; significant (p<0.05), ns; non-significant (p>0.05).\n\n\nDiscussion\n\nThis study aimed to evaluate the ability of the 980nm diode laser to decrease postoperative pain and disinfect the root canal and to evaluate whether the diode laser could be a successful adjunctive aid to conventional endodontic treatment.\n\nOver time, various laser types have been developed and are used in different dentistry fields7. Among the various lasers, diode lasers are the most frequently used8. The active medium of the 980 nm diode laser is a solid-state semiconductor made of indium, gallium and arsenide. Diode lasers have several advantages: extreme compactness, affordability, ease of operation, simple setting-up, versatility and small size8. Diode wavelengths are highly absorbed in hemoglobin and melanin and have little absorption in dental hard tissue. They are also highly absorbed by water9, which provides the laser with the advantage of acting selectively and precisely8.\n\nIn the present study, the intracanal irradiation was done using the pulsed mode to decrease the risk of thermal damage on external root surface and thus decrease the postoperative pain and favor healing of periapical area10. The temperature on the root canal walls rapidly decreases as the intracanal irradiation with the activated 200 µm fiber-optic is directed from apical to coronal direction rapidly. Thus, guaranteeing that the surrounding tissue is only marginally affected and damage of periodontal tissues or the underlying bone should not be expected9.\n\nThe 980 nm diode laser use an optical flexible fiber 200 µm to deliver the beam to the target area, probably distributing homogenously the light inside the root canal for a more efficient photoreaction7. Garcez et al.11 achieved higher antimicrobial effect when they used the optical fiber in disinfection of the root canal. Diode lasers have demonstrated excellent clinical benefits12.\n\nIn this study, the NRS scale was used to record postoperative pain. Jamison et al.13, reported that, the NRS has greater sensitivity to change in pain intensity. Amelia and Barbara14 reported that the NRS represents interval levels so it can provide data for parametric analysis.\n\nThe results of this study showed that the DL group had statistically significant lower pain levels than the Endo group at all tested time intervals (6, 12, 24, 48 hours and 7 days). These results are in accordance with the findings of Berk et al.15 and Pawar et al.16, who reported that the use of diode laser in root canal irradiation showed significantly lower pain at 8, 24, 48 hours and 7 days postoperatively when compared to conventional treatment. Tuner et al.17, found that, usually after conventional RCT of chronic cases, which is the situation in this trial, the case become acute when the process of healing starts, thus patients are at risk of experiencing postoperative pain which was not the situation in the DL group. The exact mechanism by which the use of laser results in decreasing post-operative pain is still unknown. Some authors proposed some mechanisms by which the diode laser relieves pain: Pawar et al.16 and Bjordal et al.18 found that the diode laser acts on chronic pain and has an anti-inflammatory effect by decreasing PGE2, bradykinin, histamine, acetyl choline and serotonin, also the diode laser was proved to decrease the production of substance P. Our results concerning postoperative pain in the cases of chronic apical periodontitis, when treated with diode laser, showed promising results.\n\nIt is generally accepted that the development of periapical diseases are pathologic features of polymicrobial bacterial infection and its components which stimulate bone resorption19. Thus, in this study the effect of the treatment on both the aerobic and the anaerobic bacterial count were assessed following the same methodology of Garcez et al.20 and Bonsor et al.2. This technique was selected due to its ability to detect exclusively the viable bacteria and also the correlation proved previously, by many studies21,22, between negative cultures at time of obturation and more favorable treatment outcomes.\n\nIn this study, the antibacterial results showed statistically significant lower bacterial count in the S3 samples of DL group than the S2 samples of both groups. It also showed significantly lower bacterial recolonization in S4 samples of the DL group than the Endo group. The S5 samples of the DL group resulted in significantly lower bacterial count than the Endo group, which may favor the treatment outcomes21,22. Our findings are in accordance with the findings of Garcez et al.11 and Gutknecht et al.4, who found that the use of diode laser resulted in significant decrease of the intracanal bacterial load. The high antibacterial effect of diode laser may be explained by the fact that the near infrared lasers are absorbed to small extent by dentin. This is important for the efficient disinfection as the laser is not absorbed by the superficial dentin but rather penetrates deep into the intertubular dentin1. Vaarkamp et al.23 and Odor et al.24 provided an explanation for this way of light propagation, as they described the ability of enamel prisms and dentinal tubules to act as an optical fiber and thus allowing the diode laser to be more effective in deep layers of dentin. According to Gutknecht et al. in 20089, the ND: YAG, 810 nm diode laser and 980 nm diode laser are the only wavelengths that showed high transmission through hydroxyapatite and water. Thus, it can be used successfully for the disinfection of root canals.\n\nDiode laser radiation has a bactericidal effect by altering the bacterial cell wall. Microbiologists25 talk about a permanent destruction of the cell membrane, which is commonly in correlation with direct heat having an impact on the bacteria. The diode laser exerts a photo-thermal effect on the bacteria26. It also, exerts a photo-disruptive effect on the unreachable bacteria26.\n\nGuteknecht et al.4 demonstrated that diode laser light can penetrate up to >1000 μm into the dentin. Thus, it can be an effective means for disinfection of the root canal system together with conventional biomechanical instrumentation reaching areas, which were considered earlier non-reachable.\n\nThis study is a randomized clinical trial conducted on a relatively big sample size patients, in real clinical settings and was conducted efficiently. It proposes an alternative way for treatment of necrotic teeth with chronic periapical lesions efficiently and without postoperative pain.\n\nThe following limitations should be considered: this study didn’t evaluate if there is a difference between single visit or two visit approach when the intracanal diode laser is used on postoperative pain and bacterial count of necrotic teeth with chronic periapical lesions.\n\nFurther in vivo and immunological studies are needed to identify the exact mechanism by which the intracanal Diode laser resulted in decreasing postoperative pain.\n\n\nConclusions\n\nIntracanal diode laser irradiation has the ability to decrease the postoperative pain experienced after conventional root canal treatment in cases of necrotic teeth with periapical lesions. Implementation of suitable wavelengths, together with conventional methods of cleaning and shaping, can effectively sterilize the root canals, dentin and periapical area and decrease the bacterial recolonization. Thus, based on the findings of this study, it may be concluded that the 980 nm diode laser can be used as an adjunct to conventional endodontic therapy.\n\n\nData availability\n\nF1000Research: Dataset 1. Full de-identified data for each participant, including demographic data, pain scores at all time intervals, and bacterial count of both aerobic and anaerobic bacteria. , https://doi.org/10.5256/f1000research.16794.d22435127", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nSupplementary material\n\nSupplementary File 1: Completed CONSORT checklist.\n\nClick here to access the data\n\nSupplementary File 2: CONSORT flow diagram.\n\nClick here to access the data\n\nSupplementary File 2: Full study protocol.\n\nClick here to access the data\n\n\nReferences\n\nAsnaashari M, Asnaashari N: Clinical application of 810nm diode laser and low-level laser therapy for treating an endodontic problem: A case presentation. Lasers Med Sci. 2011; 2: 82–86. Publisher Full Text\n\nBonsor SJ, Nichol R, Reid TM, et al.: Microbiological evaluation of photo-activated disinfection in endodontics (an in vivo study). Br Dent J. 2006; 200(6): 337–41, discussion 329. PubMed Abstract \| Publisher Full Text\n\nMoritz A, Gutknecht N, Goharkhay K, et al.: In vitro irradiation of infected root canals with a diode laser: results of microbiologic, infrared spectrometric, and stain penetration examinations. Quintessence Int. 1997; 28(3): 205–9. PubMed Abstract\n\nGutknecht N, Franzen R, Schippers M, et al.: Bactericidal effect of a 980-nm diode laser in the root canal wall dentin of bovine teeth. J Clin Laser Med Surg. 2004; 22(1): 9–13. PubMed Abstract \| Publisher Full Text\n\nBeer F, Buchmair A, Wernisch J, et al.: Comparison of two diode lasers on bactericidity in root canals--an in vitro study. Lasers Med Sci. 2012; 27(2): 361–4. PubMed Abstract \| Publisher Full Text\n\nKoba K, Kimura Y, Matsumoto K, et al.: Post-operative symptoms and healing after endodontic treatment of infected teeth using pulsed Nd:YAG laser. Endod Dent Traumatol. 1999; 15(2): 68–72. PubMed Abstract \| Publisher Full Text\n\nPirnat S: Versatility of an 810-nm diode laser in dentistry: An overview. Journal of the Laser and Health Academy. 2007; 4: 1–9. Reference Source\n\nMaturo P, Perugia C, Docimo R: Versatility of an 810 Nm Diode laser in pediatric dentistry. International Journal of Clinical Dentistry. 2013; 6: 161–72. Reference Source\n\nGutknecht N: Lasers in endodontics. Journal of the Laser and Health Academy. 2008; 4: 1–5. Reference Source\n\nGutknecht N, Franzen R, Meister J, et al.: Temperature evolution on human teeth root surface after diode laser assisted endodontic treatment. Lasers Med Sci. 2005; 20(2): 99–103. PubMed Abstract \| Publisher Full Text\n\nSilva Garcez A, Núñez SC, Lage-Marques JL, et al.: Efficiency of NaOCl and laser-assisted photosensitization on the reduction of Enterococcus faecalis in vitro. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2006; 102(4): e93–8. PubMed Abstract \| Publisher Full Text\n\nGoharkhay K, Moritz A, Wilder-Smith P, et al.: Effects on oral soft tissue produced by a diode laser in vitro. Lasers Surg Med. 1999; 25(5): 401–6. PubMed Abstract \| Publisher Full Text\n\nJamison RN, Gracely RH, Raymond SA, et al.: Comparative study of electronic vs. paper VAS ratings: a randomized, crossover trial using healthy volunteers. Pain. 2002; 99(1–2): 341–7. PubMed Abstract \| Publisher Full Text\n\nAmelia W, Barbara H: Pain: A review of three commonly used pain rating scales. Journal of Clinical Nursing. 2005; 14(7): 789–804. PubMed Abstract \| Publisher Full Text\n\nBerk G, Ulucam U, Berk N: Clinical healing process and symptoms of two cases of chronic periapical lesions treated with Er, Cr:YSGG laser. Journal of Oral Laser Application. 2004; 4: 211–5. Reference Source\n\nPawar S, Pujar M, Makandar S, et al.: Postendodontic treatment pain management with low-level laser therapy. Journal of Dental Lasers. 2014; 8(2): 60–3. 10.4103/0976-2868.145141\n\nTuner J, Hode L: The laser therapy handbook, Sweden, Prima Book AB. 2007.\n\nBjordal JM, Johnson MI, Iversen V, et al.: Low-level laser therapy in acute pain: A systematic review of possible mechanisms of action and clinical effects in randomized placebo-controlled trials. Photomedicine and Laser Surgery. 2006; 24(2): 158–68. PubMed Abstract \| Publisher Full Text\n\nStashenko P: Role of immune cytokines in the pathogenesis of periapical lesions. Endod Dent Traumatol. 1990; 6(3): 89–96. PubMed Abstract \| Publisher Full Text\n\nGarcez AS, Nuñez SC, Hamblin MR, et al.: Antimicrobial effects of photodynamic therapy on patients with necrotic pulps and periapical lesion. J Endod. 2008; 34(2): 138–42. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSjögren U, Figdor D, Perssom S, et al.: Influence of infection at the time of root filling on the outcome of endodontic treatment of teeth with apical periodontitis. Int Endod J. 1997; 30(5): 297–306. PubMed Abstract \| Publisher Full Text\n\nWaltimo T, Trope M, Haapasalo M, et al.: Clinical efficacy of treatment procedures in endodontic infection control and one year follow-up of periapical healing. J Endod. 2005; 31(12): 863–6. PubMed Abstract \| Publisher Full Text\n\nVaarkamp J, Bosch JJ, Verdonschot EH: Propagation of light through human dental enamel and dentine. Caries Res. 1995; 29(1): 8–13. PubMed Abstract \| Publisher Full Text\n\nOdor TM, Watson TF, Pitt Ford TR, et al.: Pattern of transmission of laser light in teeth. Int Endod J. 1996; 29(4): 228–34. PubMed Abstract \| Publisher Full Text\n\nAhmeduddin M, Nagesh B, Reddy KN, et al.: An assessment of bactericidal effect of two different types of lasers on enterococcus faecalis: An in vitro study. Journal of Dental Lasers. 2012; 6(1): 2–6. Publisher Full Text\n\nPreethee T, Kandaswamy D, Arathi G, et al.: Bactericidal effect of the 908 nm diode laser on Enterococcus faecalis in infected root canals. J Conserv Dent. 2012; 15(1): 46–50. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMorsy DA, Negm M, Diab A, et al.: Dataset 1 in: Postoperative pain and antibacterial effect of 980 nm diode laser versus conventional endodontic treatment in necrotic teeth with chronic periapical lesions: A randomized control trial. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16794.d224351" }	[ { "id": "40677", "date": "03 Dec 2018", "name": "Eman H.A. Aboulezz", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAfter reviewing the manuscript entitled 'Posteroperative pain and antibacterial effect of 980nm diode laser versus conventional nedodontic treatment in necrotic teeth with chronic periapical lesions:A randomized control trial', I have found that it is such an outstanding article overall. It provides a comprehensive examination of a complex subject; providing new options in treatment of such an inevitable problem. A detailed description of the manuscript will be as follows:\nIntroduction: which was a well-written and covering more than enough of the relevant theoretical data of literature. Aim of the study: Here the author stated the objectives and goals of the study in an appropriate manner. In the Methods: the author gave a detailed description of the employed methodology of the investigation. It was very well elaborated and relevant to the aim of the study. Results: The results were well organized into different categories; this section included tables and charts. All of which help the reader get a clear detailed insight of the different aspects occurring within the two groups of patients. Discussion: the author in this chapter was successful in showing logical overall arguments and used appropriate number of bibliography sources. This shows that the author developed a deep understanding of the theoretical knowledge and the discussed problems. Conclusion: The manuscript provides a relevant summary of the subject, methodology and results. References: They were organized and included a lot of recent publications.\nGenerally, this manuscript meets all the basic requirements of good research and demonstrates coherence between theoretical position, problem presented and the methodology appointed. The author showed in this manuscript a significant level of understanding and analytical ability. I find the work innovative and recommend indexing it.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "49584", "date": "04 Jul 2019", "name": "Patricia da Ana", "expertise": [ "Reviewer Expertise Dentistry and lasers applications in dental tissues" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a well-written and well-conducted article with clear problem and objective. For indexing, it is suggested some changes, described below:\nThere are many recent published papers that report the clinical efficacy of diode laser in endodontic treatment (see Genc Sen and, Kaya 20191). In this way, the authors need to make clear in the introduction what is the innovation that this manuscript presents.\n\nReferences need to be more up-to-date. There is no citation of any article published in the last 5 years.\n\nThe authors should pay attention to the citations. In the last sentence of the introduction, they comment that the use of the diode laser in endodontics was recently established, but they cite a 2004 reference.\n\nPlease replace the abbreviation of units using capital letters; for example, ml to mL, watt to Watt.\n\nPlease give references to the laser protocol used. Are there studies in the literature based on temperature increases generated in the root canal using such parameters? What are the observed values? Comment on their safety for the periodontal ligaments in the discussion.\n\nThe authors do not explain how they ensured periods of pain assessment.\n\nPlease correct the neodymium laser nomenclature. The correct one is Nd:YAG.\n\nPlease explain why there was no bacterial collection in the S3 period of the placebo group. Discuss whether the insertion of the laser fiber may have exerted some mechanical removal of the biofilm.\n\nPlease correct the captions of all the figures and tables. There are legends that mention letters and asterisks, but there is nothing in the figure (e.g., Figure 2), while Tables 4 and 5 do not explain what upper and lower case letters refer to.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "49690", "date": "25 Jul 2019", "name": "Ibrahim Ethem Yaylali", "expertise": [ "Reviewer Expertise Evidence-based medicine." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIntroduction: There is a lack of information about the previously performed studies. The authors should clearly and briefly describe the existing studies, which have been done so far. Why did the authors perform this study? Which gaps was/were exist in the field of postoperative pain evaluation when diode lasers are used? What was the Null hypothesis? All should be described in the Introduction. The power analysis is adequate.\nMethods: Allocation concealment and sequence generation were described adequately.\n\nResults: Demographic and baseline characteristics (mean age, sex, education, etc) of the patients can be showed in one Table so that Table 1 and Table 2 can be removed.\nDiscussion: In the Discussion, the second paragraph is unnecessary. It can be completely removed and can be add to the Introduction.\n\nIs the work clearly and accurately presented and does it cite the current literature? No\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1795
https://f1000research.com/articles/7-1794/v1	15 Nov 18	{ "type": "Research Article", "title": "Towards cardiovascular risks in children with chronic kidney disease: a prospective cohort study", "authors": [ "Taghreed Fadhil Al-Doori", "Abd El-Salam Dawood Al-Ethawi", "Jessar Saleem Hasan", "Ban Adil Al-Kaaby", "Taghreed Fadhil Al-Doori", "Jessar Saleem Hasan", "Ban Adil Al-Kaaby" ], "abstract": "Background: Children with chronic kidney disease (CKD) are at substantially high risk of morbidity and mortality from cardiovascular disease (CVD). Although this issue has been extensively studied in adults, little is known whether similar associations exist in the paediatric population. We therefore aimed to evaluate the cardiac structure and function in children with CKD, and investigate the factors that contribute to the development of CVD. Methods: A prospective cohort was established following 40 children with CKD treated in the nephrology unit at a high-volume, tertiary, teaching hospital and compared to age- and gender-matched controls of the same number. We reviewed the patients’ medical records, assessed growth parameters, measured blood pressure, took blood samples, and performed echocardiography. Results: More than half of the CKD patients are hypertensive (N = 22; 55%) and were found to have a higher proportion of increased left ventricular mass index (LVMI) (75.5%; P = 0.001). In contrast, we did not find any significant association between CKD and both valvular calcification & left ventricular (LV) fractional shortening (FS) in children (P = 0.314). Furthermore, high LVMI is found to be correlated well with the following: anaemia, hypertension, CKD duration > one year, hyperparathyroidism, estimated glomerular filtration rate (eGFR) < 15ml/minute/1.73 m2 and death (P < 0.05). Conclusions: Despite the fact that LV systolic function is preserved and valvular calcification is usually absent, left ventricular hypertrophy (LVH) is common in children with CKD. Hypertension, anaemia, hyperparathyroidism, as well as the severity and duration of renal impairment, are amongst the additional risk factors that predispose to LVH. We contribute this study to the growing information of the review articles regarding the association between CKD and CVD in paediatrics.", "keywords": [ "Chronic Kidney Disease (CKD)", "ESRD", "Cardiovascular Disease (CVD)", "Left Ventricular Hypertrophy (LVH)", "Left Ventricular Mass Index (LVMI)", "paediatric", "children" ], "content": "Introduction\n\nChildren with chronic kidney disease (CKD) are at substantially higher risk of morbidity and mortality from cardiovascular disease (CVD)1. CVD is regarded by many as the most important cause of death among patients receiving chronic renal replacement therapy; accounting for approximately 45 per cent of overall mortality2. A Task Force on CVD of the National Kidney Foundation has compared the mortality rate from CVD in dialysis patients (≈50 thousand deaths) with those in the general population (≈2 million deaths). They showed that annual mortality from CVD is much greater in dialysis patients, in spite of stratification for race, gender, or age group1. Two pathological causes of the raised risk for CVD deaths have been identified. First, the high prevalence of CVD in dialysis patients. Second, is the elevated case fatality rate among this group1.\n\nHerzog et al. have examined the US Renal Data System between 1977 and 1995. They evaluated outcomes of more than 34 thousand patients on long-term dialysis with myocardial infarction (MI). Interestingly, they found cardiac-related deaths were ≈52 per cent at 2 years and ≈70 per cent at 5 years post-MI follow-up2.\n\nResearchers have described three pathological forms of CVD prevalent in patients with CKD. The first one is atherosclerosis which is considered as the primary aetiology of ischemic heart disease (IHD) in that subset of patients3. Further, coronary artery plaques, in CKD patients, tend to be more advanced, with higher degrees of calcification and media thickening4. The second pathology is an alteration in heart geometry. It includes left ventricular (LV) remodelling, concentric LV hypertrophy (LVH), and eccentric LVH.\n\nIn regards to concentric LVH, the LV wall thickness rises to a greater extent than its diameter; while in the eccentric type, the magnitude of LV wall increase is in proportion to the increase in LV diameter. Concentric LVH, on the one hand, is predominantly seen in cases with pressure overload secondary to aortic stenosis, arteriosclerosis, or hypertension. On the other hand, the eccentric form can be depicted in scenarios with volume overload, such as anaemia, fluid retention, and arteriovenous fistulae5. The third pathological form is arteriosclerosis – a large vessel disease. This process involves loss of elasticity, vessel remodelling, and development of non-compliant arteries6. This results in elevated pulse pressure that, in turn, has been regarded, by many, as a risk factor for CVD deaths in dialysis patients7.\n\nTwo kinds of risk factors have been identified in CKD patients – traditional and uremia-related ones8. The traditional risk factors, that are included in the Framingham Heart Study Offspring Cohort population, include, among others, hyperlipidemia, hypertension, diabetes, male sex, LVH, and smoking. Despite the fact that CKD patients have a higher incidence of many of these risk factors, they are exposed to uremia-related ones as well8. A large number of dialysis patients have high levels of oxidative stress, homocysteine, lipoprotein(a), and lipoprotein remnants. Levels of inflammatory markers (e.g. C-reactive protein) and thrombogenic factors (such as fibrinogen) also are elevated. Furthermore, recent studies have shown that hemodialysis patients have a higher prevalence of sleep disorders, including apnea, that might contribute to the elevated risk for CVD9.\n\nHowever, although the issue of CVD risks in CKD patients has been extensively studied and discussed in adults, little is known whether similar associations exist in the paediatric population. The aim of this study, therefore, is to evaluate the cardiac structure & function in children with CKD and investigate for factors that contribute to the development of CVD in that set of patients.\n\n\nMethods\n\nIn this prospective cohort study, we enrolled 45 patients with CKD, treated in the nephrology unit of the Central Child Teaching Hospital-Baghdad, a high-volume, tertiary, teaching hospital over a 13-month period (from October 2016 to November 2017).\n\nWe included all the patients attending the unit during the period of the study (the sample size) with the following inclusion criteria in mind: age between 2–14 years, GFR<60 ml/min/1.73 m2 for more than 3 months, and absence of confounders, such as myocardial disease (congenital, structural or primary), diabetes mellitus (DM) or corticosteroid use during the period of study. As such, five patients were excluded as two had congenital heart diseases, another two had DM, and one was receiving corticosteroid treatment.\n\nFor comparison, we selected a control group of 40 patients (1:1 ratio) who were age- and gender-matched to the study group. Age matching was considered present if the difference was within two years. The controls were patients in the same hospital but with unrelated conditions or diseases. They have been recruited randomly from different units of the hospital. Informed written consent was taken and the following inclusion criteria were considered: normal systemic blood pressure (BP), renal function tests, general urine exam (GUE), and abdominal ultrasound (US) of the genitourinary tract with no clinical evidence of cardiovascular system issues (such as cyanosis, shortness of breath, chest pain, palpitation, and cardiac murmur).\n\nAs per this study, CKD is defined as renal damage or glomerular filtration rate (GFR) of less than 60 ml/minute/1.73 m2 body surface area (BSA) for three months, regardless of the cause10.\n\nFor all the patients (cases and controls) included in this study, we measured the following parameters: body weight (BW), height (Ht), height percentile, body mass index (BMI) (BW/Ht2), and surface area (SA). The latter is calculated as follows: SA = Square Root [BW (kg) x Ht (cm) / 3600].\n\nThe patients' medical records were reviewed for the presence of primary renal disease, its duration, underlying cause, and the medications used. In regards to the cause, we categorized the patients into four groups. Group I: congenital renal anomalies, group II: acquired renal diseases, group III: hereditary nephropathies and group IV: CKD of unknown aetiology. We recorded the mode of treatment into one of three modalities: hemodialysis (HD), continuous ambulatory peritoneal dialysis (CAPD) or conservative treatment. The outcome of the patients, during the period of the study, have been identified.\n\nOn the day of the echocardiographic study, we took the following blood samples for each patient: haemoglobin, blood urea, serum creatinine, serum calcium and serum phosphorus. All these tests were analyzed using the standard, local laboratory methods.\n\nThe estimated GFR (eGFR) has been calculated using Schwartz Formula11 (eGFR= Height (cm)/serum creatinine (mg/dl) * K). Where K is a “constant” based on age as follows: 0.70 in adolescent boys; 0.55 in adolescent females & children; 0.45 in full term infants and 0.33 in preterm newborns.\n\nAccording to the eGFR, we divided the patients into two groups: those < 15 ml/min/1.73 m2 and those ≥ 15 ml/min/1.73 m2. Furthermore, we divided the duration of the CKD into two groups–either within one year or those more than or equal to one year.\n\nBlood pressure was measured three times during the same day; in sitting position and using mercurial sphygmomanometer with an appropriate cuff size12. We took the average of the three measurements and compare the results to nomograms. Hypertension is said to be present when the averaged systolic blood pressure (SBP) and or diastolic blood pressure (DBP) equal to or exceeded 95th percentile for sex, age & height12.\n\nAnaemia is defined as a haemoglobin level of less than 11 g/dl in prepubertal children with CKD11.\n\nIntact parathyroid hormone (iPTH) was assessed using a Cobas e 411 Analyzer (Roche Diagnostics) at a local private lab. Hyperparathyroidism is said to be present when the iPTH exceeds 65 pg/ml13.\n\nTwo board-certified, pediatric cardiologists performed echocardiographic assessments using a Philips CX50 Ultrasound Machine (Philips Healthcare, USA) with the S5-1 transducer. They examined for valvular calcification using 2D Mode; while M-Mode has been used to assess the following: Interventricular Septum thickness in diastole (IVSd), Left Ventricular Internal Diameter in diastole (LVIDd), Left Ventricular Internal Diameter in systole (LVIDs) and Left Ventricular Posterior Wall in diastole (LVPWd). A built-in software tool was used to calculate the fractional shortening (FS %) [(LVIDd – LVIDs) / LVIDd * 100]. Left Ventricular Mass (LVM) was calculated using a computerized Microsoft Access 2016 model using Devereux & Reichek cube formula14 : LV Mass = 0.8 x 1.04 x [(LVIDd + LVPWd + IVSd)3 – LVIDd3] + 0.6. The calculated LVM is then indexed to body surface area (LVMI).\n\nFor this study an FS of 25–43% was considered to be a normal value. However, for the IVSd, LVIDd, LVIDs, and LVPWd we depend on values indexed to body surface area accordingto Park 200815.\n\nLeft ventricular hypertrophy (LVH) is defined as LVMI greater than the 95th percentile for gender and age16.\n\nStatistical analysis was performed with IBM SPSS Statistics 23.0. Suitable graphs & tables were used to describe the data. Student's T-test was used for comparison between continuous variables. Chi-square & Fisher's exact tests were used to examine qualitative and frequency data and to find the relation between risk factors and different outcomes. P value < 0.05 was considered significant.\n\n\nResults\n\nThe cohort consisted of 80 patients divided into two groups: 40 children with CKD (cases group) and 40 children without CKD (controls group) (Datasets 1 & 2).\n\nSome of the general characteristics of the study population are shown in Table 1.\n\nAbbreviations: BMI: Body Mass Index = body weight (Kg)/ Height2 (m2); CI: Confidence Interval\n\n\nBlood pressure\n\nAmong the study patients, 72.5% were normotensive. However, all the control group were normotensive and 55% of cases group were hypertensive with a significant association between CKD & high blood pressure (P = 0.001) (Figure 1)\n\n\nHaemoglobin level\n\nAbout 40% of the study population, 7.5% in the control group and 72.5% of the cases group were anaemic (Figure 2). There is a significant association between CKD and anaemia (P = 0.001).\n\n\nBiochemical and hormonal studies\n\nWith regards to the biochemical & hormonal studies, blood urea, serum (S.) creatinine, S. phosphorus (Po), Calcium * Phosphorus (Ca * Po) product and parathyroid hormone (PTH) levels were higher amongst patients in the case group than in the controls. However, S. Ca is lower in the cases (Table 2).\n\nAbbreviations: Bl. – Blood; Ca – Calcium; CI – Confidence Interval; Po – Phosphorus; PTH – Parathyroid hormone; S. – Serum\n\nMost of the patients in the cases group had CKD for more than a year (62.5%), and the majority (82.5%) were alive during the period of the study (Figure 3 & Figure 4). More than half of them were treated conservatively, however, other modalities were utilized (Figure 5).\n\nCAPD – continuous ambulatory peritoneal dialysis.\n\nCongenital renal abnormalities constitute 37.5% of the causes of CKD in this study, with vesicoureteral reflux (VUR) being the most common anomaly seen (32.5%) (Table 3).\n\nAbbreviations: FSGS – Focal Segmental Glomerular Sclerosis; PUV – Posterior Urethral Valve; RPGN - Rapid Progressive Glomerular Nephritis; VUR – Vesicoureteral Reflux\n\n\nEchocardiographic findings\n\nCKD Patients were found to have a higher left ventricular mass index (LVMI) when compared to the control group (75.5% vs. 0%) with a statistically significant association between CKD and LVH (P = 0.001). However, we did not find any significant association between CKD and both valvular calcification, and fractional shortening (FS) (P = 0.314) (Table 4).\n\nAbbreviations: FS – Fractional Shortening; LVMI – Left Ventricular Mass Index.\n\nIn terms of the cardiac measurements, interventricular septal diameter in diastole (IVSd), left ventricular posterior wall diameter in diastole (LVPWd), and LVMI are higher (P < 0.05) among CKD patients than the controls. On the other hand, there is no statistical significant difference (P > 0.05) in term of FS, LV internal diameter in diastole (LVIDd) and LV internal diameter in systole (LVIDs) between the CKD patients and the controls (Table 5).\n\nAbbreviations: CI – Confidence Interval; FS – Fractional Shortening; IVSd – Interventricular Septal thickness in Diastole; LVIDd – Left Ventricular Internal Diameter in Diastole; LVIDs – Left Ventricular Internal Diameter in Systole; LVMI – Left Ventricular Mass Index; LVPWd – Left Ventricular Posterior Wall in Diastole\n\nHigh LVMI was seen in patients with all of the following: anaemia, hypertension, CKD duration > one year, high PTH and eGFR < 15 ml/min/1.73 m2. All the patients who died, as well as those with acquired or unknown cause of CKD, have increased LVMI. On the other hand, no association has been discovered between gender or mode of treatment, and the risk of high LVMI (Table 6).\n\nAbbreviations: CAPD – Chronic Ambulatory Peritoneal Dialysis; CKD – Chronic Kidney Disease; eGFR - Estimated Glomerular Filtration Rate; LVMI – Left Ventricular Mass Index; PTH – Parathyroid Hormone\n\n\nDiscussion\n\nIn this study, 19 patients (47.5%) – in the cases group – were males and 21 cases (52.5%) were females; with a male: female ratio of 0.9: 1. However, this contradicts results obtained in other studies, like Harambat et al.17 where the higher prevalence of CKD in males has been attributed to the higher frequency of congenital abnormalities of the kidney and urinary tract (CAKUT) in this group.\n\nBody mass index (BMI) does not reflect body composition. Furthermore, a patient who has an appropriate BMI for age might not have a typical body composition17. In this study, BMI shows no significant difference (P = 0.812) between the cases and the controls groups. This is consistent with previous studies18,19\n\nIn terms of blood pressure, 55% of patients in the cases group were hypertensive (P = 0.001), though other studies show a higher prevalence20. Many factors contribute to the raised incidence of hypertension in CKD patients. They include sodium retention, increased activity of the renin-angiotensin-aldosterone system, an exaggerated activity of the sympathetic nervous system, secondary hyperparathyroidism, deficient nitric oxide and endothelium-mediated vasodilation, and treatment with erythropoietin – if present20–24. In addition, hypertension might be a causative (e.g. hypertensive nephrosclerosis) or contributary pathology in the development of CKD25.\n\nAnaemia was found in 29 patients (72.5%) in the cases group (P = 0.001). This is consistent with results obtained by other researchers26–29. A hypoproliferative state is thought to be the underlying mechanism. The latter might result from the following: First, changes in iron metabolism which is hepcidin-induced that causes decreased iron absorption from the gut, in addition to iron macrophage trapping30,31. Second, shortened red blood cells (RBC) survival, resulting from raised macrophage activity32. Third, a high apoptotic death of RBC precursors in the bone marrow33. Finally, a \"relative\" decline in erythropoietin (EPO) production31.\n\nIn the United States and Europe, many studies have shown that congenital genitourinary anomalies account for almost 60% of CKD cases in younger children; while glomerular diseases are more common in older children & adolescents34–36. In the first subset of patients, obstructive uropathy (21%) is the most common pathology, followed by renal dysplasia/hypoplasia (18%) and reflux nephropathy (8%). On the other hand, focal segmental glomerulosclerosis (FSGS) is the most prevalent glomerular lesion, accounting for 9% of cases. However, in our study, congenital renal anomalies were found in 15 patients (37.5%) with vesicoureteral reflux (VUR) being the most common pathology (N = 13; 32.5%). Both FSGS and rapidly progressive glomerular nephritis (RPGN) have the same prevalence as a cause of acquired glomerular diseases (N = 2; 5%).\n\nLeft ventricular hypertrophy (LVH) is regarded by many as an adaptive response to long-lasting volume or pressure overload. Initially, it is beneficial because it allows for increasing work capacity, maintains systolic function and decreases energy consumption & wall stress. However, with time, LVH becomes detrimental as it alters LV diastolic function, decreases coronary perfusion reserve and predisposes to arrhythmias & sudden death37–39. Some researchers40,41 have found that an increase in LV mass by 1 gm/m2.7/month is associated with a 62% rise in the risk of fatal and non-fatal cardiovascular events in dialysis patients. In our study, all the seven children who died had LVH.\n\nIn this study, CKD patients had a thicker diastolic interventricular septum (IVSd) (10.5 vs 6.4, P = 0.001), diastolic LV posterior wall (LVPWd) (10.7 vs 6.2, P = 0.001) and LV Mass Index (LVMI) (150.1 vs 54.7, P = 0.001) in comparison to controls. Furthermore, LVMI was increased in 27 (75.5%) patients in the cases group (P = 0.001). These results are consistent with those obtained by Malikenas et al.42 and van Huis et al.43.\n\nDespite the fact that LV systolic dysfunction carries a poor prognosis in end-stage renal disease (ESRD) in adults, and is considered by many as an important co-morbidity, LV function is usually well preserved in the pediatric population43. In our study, only one case (2.5%) shows low LV fractional shortening with no statistically significant association between CKD & systolic dysfunction. This is consistent with other researchers in other parts of the world43,44. It may be that LV systolic dysfunction only becomes clinically evident after a long period of time.\n\nMany studies have shown that vascular & valvular calcification is associated with cardiovascular disease, which is by far the most common cause of death in CKD45,46. Calcification is prevalent among CKD patients, particularly those on dialysis. Prevalence is highest among patients with lower eGFR, and it is presumed that, as the eGFR declines, the prevalence as well as the severity of the calcification rise47–49. It has been seen that 80% of dialysis patients show calcification versus 47–83% among those who are not on dialysis48–56.\n\nIn our study however, valvular calcification was present in only one patient (2.5%). This low prevalence can be explained as follows: first, all the studied patients are under the age of 14 years. Knowing that calcification is a slowly-developing process that takes time, the younger the patient, the less likely this pathology can be clinically detected. This was shown in a comprehensive systematic meta-analysis review of 30 studies over a 20-year-period that demonstrated that the age of the patient is one of the most important factors associated with calcification in CKD50. Second, we relied on echocardiography in detection and quantification of valvular calcification. However, computed tomography (CT), using special software & scanners is considered, by many, as a more sensitive tool in this context, especially when coronary arteries calcification needs to be assessed as well57,58. Third, preventive measures, in our centre, may play a role, which includes the cautious use of vitamin D therapy and achieving net calcium balance by using non-calcium-containing phosphate binders or by modifying the dialysate bath calcium concentration in dialysis patients59,60.\n\nIn adults, many studies have shown an association between LVH and parathyroid hormone (PTH) concentration in patients with CKD61,62. The underlying mechanisms of PTH-induced LVH include an indirect effect – by elevating systemic blood pressure, and a direct effect of this hormone on the cardiac myocytes62. Furthermore, in vitro studies have revealed that PTH has inotropic, chronotropic as well as hypertrophic effects on cardiac cells63,64. In this study, high LVMI was present in 26 out of 30 cases (86.7%) with high PTH. The latter is consistent with results obtained by Mitsnefes et al.65 who showed that raised PTH level is associated with LVH progression in CKD stages 2–4.\n\nIn our study, LVH was found in 24 out of 29 cases (82.8%) with anaemia. This is consistent with most studies in children, which revealed a significant relationship between anaemia and raised LVMI65–68. However, recent studies in adults with mild-moderate CKD or on long-term dialysis showed that increasing haemoglobin (Hb) levels was not associated with LVH regression69–72. Therefore, the authors have suggested that the relationship between LVH and anaemia could not be causal. Of note, these studies recruited subjects with relatively mild anaemia and, thus, not able to answer the question of whether correcting more severe anaemia may lead to a decline of LVMI. On the contrary, Morris et al.73 observed a substantial reduction in LVMI with the treatment of severe anaemia in children on dialysis.\n\nRegarding the mode of treatment of CKD and its association with LVH, research74,75 has shown that the prevalence of LVH varies with the mode of treatment of CKD; being 16–31 % in those with eGFR > 30 ml/min/1.73 m2, 60–75% when renal replacement therapy is started, and 70–90% in patients with regular dialysis. Our study, however, reveals rather different statistics. Up to 50% of patients on conservative treatment had LVH. Refusal to start renal replacement therapy by the family has caused a substantial delay in the treatment and this might be the most likely explanation of these figures.\n\nThe lower the eGFR, the more likelihood of having LVH. In this study, more than 85% of patients with eGFR of < 15 ml/min/1.73 m2, in comparison to 47% of those whom eGFR ≥ 15 ml/min/1.73 m2. These findings are consistent with those of researchers elsewhere in the world44,74,75.\n\nMalikenas et al.42 have noticed an association between the CKD causes and LVH. They found that patients with acquired renal diseases had significantly higher LVMI than those with congenital renal abnormalities. In our study, 100% of patients with acquired causes had high LVMI versus 53% with congenital anomalies. An interesting point, however, is that 100% of those with unknown cause of CKD had LVH. This can be explained by the fact that these cases are usually presented late to our centre.\n\n\nLimitations of the study\n\n1. Because of time constraints, only one assessment has been done to both the cases and the controls groups. We believe that with frequent, lengthy follow up on the cases arm, many parameters would change with time, such as LV mass & function, calcification, outcome, etc. Thus, an extension of this study is in progress to include months of follow-up.\n\n2. To detect cardiovascular calcification, we depended on echocardiographic detection. However, computed tomography (CT) is considered, by many, as a more sensitive tool in this context.\n\n3. Only left ventricular (LV) systolic function has been assessed, using LV fractional shortening (FS). While neither LV diastolic, nor right ventricular (RV) function has been evaluated, though we think both might be influenced in patients with CKD.\n\n\nConclusions\n\nDespite the fact that LV systolic function is usually well-preserved & valvular calcification is usually absent, LVH is common in children with CKD. LVH is associated with higher mortality in that subset of patients; with the severity & duration of renal impairment, hypertension, anaemia, and hyperparathyroidism are amongst the additional risk factors that predispose to LVH. We contribute this study to the growing information of the review articles regarding the association between CKD and cardiovascular complications in paediatrics. However, additional clinical trials are urgently required with a goal to reduce CVD in CKD children.\n\n\nData availability\n\nAll the raw data required for reproducibility of this study are available on Open Scientific Framework\n\nOSF. Dataset 1 & 2: Replication Data for: Toward Cardiovascular Risks in Children with Chronic Kidney Disease, https://doi.org/10.17605/OSF.IO/WY28B76\n\nData is available under a CC0 1.0 Universal license\n\n\nEthical approval\n\nThe Institutional Ethics Committee of the Paediatric Department, College of Medicine, Al-Mustansiriyah University has reviewed & discussed the initial application to conduct this study, informed about its progress, and any revision in the protocol and patient information/informed consent. The Committee has received & approved the final report of the study on April, 2018.\n\nNone of the investigators participating in this study took part in the decision making & voting for this study.\n\n\nPatients’ consent\n\nInformed written consent was taken from the parents of the children to participate in this study and for publication of the clinical details.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nFoley RN, Parfrey PS, Sarnak MJ: Clinical epidemiology of cardiovascular disease in chronic renal disease. Am J Kidney Dis. 1998; 32(5 Suppl 3): S112–9. PubMed Abstract \| Publisher Full Text\n\nHerzog CA, Ma JZ, Collins AJ: Poor long-term survival after acute myocardial infarction among patients on long-term dialysis. N Engl J Med. 1998; 339(12): 799–805. PubMed Abstract \| Publisher Full Text\n\nSarnak MJ: Cardiovascular complications in chronic kidney disease. Am J Kidney Dis. 2003; 41(5 Suppl): 11–7. PubMed Abstract \| Publisher Full Text\n\nSchwarz U, Buzello M, Ritz E, et al.: Morphology of coronary atherosclerotic lesions in patients with end-stage renal failure. Nephrol Dial Transplant. 2000; 15(2): 218–23. PubMed Abstract \| Publisher Full Text\n\nSchunkert H, Hense HW: A heart price to pay for anaemia. Nephrol Dial Transplant. 2001; 16(3): 445–8. PubMed Abstract \| Publisher Full Text\n\nLondon GM, Marchais SJ, Guerin AP, et al.: Arterial structure and function in end-stage renal disease. Nephrol Dial Transplant. 2002; 17(10): 1713–24. PubMed Abstract \| Publisher Full Text\n\nKlassen PS, Lowrie EG, Reddan DN, et al.: Association between pulse pressure and mortality in patients undergoing maintenance hemodialysis. JAMA. 2002; 287(12): 1548–55. PubMed Abstract \| Publisher Full Text\n\nSarnak MJ, Levey AS: Cardiovascular disease and chronic renal disease: a new paradigm. Am J Kidney Dis. 2000; 35(4 Suppl 1): S117–31. PubMed Abstract \| Publisher Full Text\n\nDe Santo NG, Cirillo M, Perna A, et al.: The heart in uremia: role of hypertension, hypotension, and sleep apnea. Am J Kidney Dis. 2001; 38(4 Suppl 1): S38–46. PubMed Abstract \| Publisher Full Text\n\nBello A, Kawar B, El Kossi M: Comprehensive Clinical Nephrology. 4th ed. Floege J, Richard J, Johnson JF, editors: Elsevier; 2010; 907–1015. Publisher Full Text\n\nMistry K: Clinical Pediatric Nephrology. 3rd ed. Kher K, Schnaper HW, editors: CRC Press Taylor & Francis Group; 2016; 601–27. Reference Source\n\nFalkner B, Daniels SR: Summary of the Fourth Report on the Diagnosis, Evaluation, and Treatment of High Blood Pressure in Children and Adolescents. Hypertension. 2004; 44(4): 387–8. PubMed Abstract \| Publisher Full Text\n\nMarcocci C, Brandi ML, Scillitani A, et al.: Italian Society of Endocrinology Consensus Statement: definition, evaluation and management of patients with mild primary hyperparathyroidism. J Endocrinol Invest. 2015; 38(5): 577–93. PubMed Abstract \| Publisher Full Text\n\nDevereux RB, Reichek N: Echocardiographic determination of left ventricular mass in man. Anatomic validation of the method. Circulation. 1977; 55(4): 613–8. PubMed Abstract \| Publisher Full Text\n\nPark MK: Appendix D: Normal Echocardiographic Values and Images. In: Pediatric Cardiology for Practitioners [Internet]. USA: Mosby-Elsevier. Fifth Edition. 2008. Reference Source\n\nHoughton A: The left ventricle & its systolic function. In: Houghton A, editor. Making sense of echocardiography: A hands-on guide. Second ed. United Kingdom CRC Press; 2014; 117–33. Reference Source\n\nHarambat J, van Stralen KJ, Kim JJ, et al.: Epidemiology of chronic kidney disease in children. Pediatr Nephrol. 2012; 27(3): 363–73. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGao T, Leonard MB, Zemel B, et al.: Interpretation of body mass index in children with CKD. Clin J Am Soc Nephrol. 2012; 7(4): 558–64. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRodig NM, McDermott KC, Schneider MF, et al.: Growth in children with chronic kidney disease: a report from the Chronic Kidney Disease in Children Study. Pediatr Nephrol. 2014; 29(10): 1987–95. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWhaley-Connell AT, Sowers JR, Stevens LA, et al.: CKD in the United States: Kidney Early Evaluation Program (KEEP) and National Health and Nutrition Examination Survey (NHANES) 1999-2004. Am J Kidney Dis. 2008; 51(4 Suppl 2): S13–20. PubMed Abstract \| Publisher Full Text\n\nBuckalew VM Jr, Berg RL, Wang SR, et al.: Prevalence of hypertension in 1,795 subjects with chronic renal disease: the modification of diet in renal disease study baseline cohort. Modification of Diet in Renal Disease Study Group. Am J Kidney Dis. 1996; 28(6): 811–21. PubMed Abstract \| Publisher Full Text\n\nNeumann J, Ligtenberg G, Klein, II, et al.: Sympathetic hyperactivity in chronic kidney disease: pathogenesis, clinical relevance, and treatment. Kidney Int. 2004; 65(5): 1568–76. PubMed Abstract \| Publisher Full Text\n\nPassauer J, Pistrosch F, Bussemaker E, et al.: Reduced agonist-induced endothelium-dependent vasodilation in uremia is attributable to an impairment of vascular nitric oxide. J Am Soc Nephrol. 2005; 16(4): 959–65. PubMed Abstract \| Publisher Full Text\n\nRaine AE, Bedford L, Simpson AW, et al.: Hyperparathyroidism, platelet intracellular free calcium and hypertension in chronic renal failure. Kidney Int. 1993; 43(3): 700–5. PubMed Abstract \| Publisher Full Text\n\nBakris GL, Ritz E: The message for World Kidney Day 2009: hypertension and kidney disease: a marriage that should be prevented. Kidney Int. 2009; 75(5): 449–52. PubMed Abstract \| Publisher Full Text\n\nCullis JO: Diagnosis and management of anaemia of chronic disease: current status. Br J Haemato. 2011; 154(3): 289–300. PubMed Abstract \| Publisher Full Text\n\nDrey N, Roderick P, Mullee M, et al.: A population-based study of the incidence and outcomes of diagnosed chronic kidney disease. Am J Kidney Dis. 2003; 42(4): 677–84. PubMed Abstract \| Publisher Full Text\n\nGangat N, Wolanskyj AP: Anemia of chronic disease. Semin Hematol. 2013; 50(3): 232–8. PubMed Abstract \| Publisher Full Text\n\nYilmaz MI, Solak Y, Covic A, et al.: Renal anemia of inflammation: the name is self-explanatory. Blood Purif. 2011; 32(3): 220–5. PubMed Abstract \| Publisher Full Text\n\nMeans RT Jr: Advances in the anemia of chronic disease. Int J Hematol. 1999; 70(1): 7–12. PubMed Abstract\n\nTheurl I, Mattle V, Seifert M, et al.: Dysregulated monocyte iron homeostasis and erythropoietin formation in patients with anemia of chronic disease. Blood. 2006; 107(10): 4142–8. PubMed Abstract \| Publisher Full Text\n\nWeiss G, Goodnough LT: Anemia of chronic disease. N Engl J Med. 2005; 352(10): 1011–23. PubMed Abstract \| Publisher Full Text\n\nMeans RT Jr, Krantz SB: Progress in understanding the pathogenesis of the anemia of chronic disease. Blood. 1992; 80(7): 1639–47. PubMed Abstract\n\nArdissino G, Dacco V, Testa S, et al.: Epidemiology of chronic renal failure in children: data from the ItalKid project. Pediatrics. 2003; 111(4 Pt 1): e382–7. PubMed Abstract \| Publisher Full Text\n\nNg DK, Robertson CC, Woroniecki RP, et al.: APOL1-associated glomerular disease among African-American children: a collaboration of the Chronic Kidney Disease in Children (CKiD) and Nephrotic Syndrome Study Network (NEPTUNE) cohorts. Nephrol Dial Transplant. 2017; 32(6): 983–90. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPloos van Amstel S, Noordzij M, Warady BA, et al.: Renal replacement therapy for children throughout the world: the need for a global registry. Pediatr Nephrol. 2018; 33(5): 863–71. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHaider AW, Larson MG, Benjamin EJ, et al.: Increased left ventricular mass and hypertrophy are associated with increased risk for sudden death. J Am Coll Cardiol. 1998; 32(5): 1454–9. PubMed Abstract \| Publisher Full Text\n\nLevy D, Garrison RJ, Savage DD, et al.: Prognostic implications of echocardiographically determined left ventricular mass in the Framingham Heart Study. N Engl J Med. 1990; 322(22): 1561–6. PubMed Abstract \| Publisher Full Text\n\nVerdecchia P, Carini G, Circo A, et al.: Left ventricular mass and cardiovascular morbidity in essential hypertension: the MAVI study. J Am Coll Cardiol. 2001; 38(7): 1829–35. PubMed Abstract \| Publisher Full Text\n\nWhyte DA, Fine RN: Chronic kidney disease in children. Pediatr Rev. 2008; 29(10): 335–41. PubMed Abstract \| Publisher Full Text\n\nZoccali C, Benedetto FA, Mallamaci F, et al.: Left ventricular mass monitoring in the follow-up of dialysis patients: prognostic value of left ventricular hypertrophy progression. Kidney Int. 2004; 65(4): 1492–8. PubMed Abstract \| Publisher Full Text\n\nMalikenas A, Cerniauskiene V, Jakutovic M, et al.: Left ventricular geometry in children with chronic renal failure. Medicina (Kaunas). 2005; 41Suppl 1: 5–11. PubMed Abstract\n\nvan Huis M, Schoenmaker NJ, Groothoff JW, et al.: Impaired longitudinal deformation measured by speckle-tracking echocardiography in children with end-stage renal disease. Pediatr Nephrol. 2016; 31(9): 1499–508. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDogan CS, Akman S, Simsek A, et al.: Assessment of left ventricular function by tissue Doppler echocardiography in pediatric chronic kidney disease. Ren Fail. 2015; 37(7): 1094–9. PubMed Abstract \| Publisher Full Text\n\nHouben E, Neradova A, Schurgers LJ, et al.: The influence of phosphate, calcium and magnesium on matrix Gla-protein and vascular calcification: a systematic review. G Ital Nefrol. 2016; 33(6): pii: gin/33.6.5. PubMed Abstract\n\nKim HJ, Kang E, Oh YK, et al.: The association between soluble klotho and cardiovascular parameters in chronic kidney disease: results from the KNOW-CKD study. BMC Nephrol. 2018; 19(1): 51. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBover J, Evenepoel P, Urena-Torres P, et al.: Pro: cardiovascular calcifications are clinically relevant. Nephrol Dial Transplant. 2015; 30(3): 345–51. PubMed Abstract \| Publisher Full Text\n\nBraun J, Oldendorf M, Moshage W, et al.: Electron beam computed tomography in the evaluation of cardiac calcification in chronic dialysis patients. Am J Kidney Dis. 1996; 27(3): 394–401. PubMed Abstract \| Publisher Full Text\n\nGoodman WG, Goldin J, Kuizon BD, et al.: Coronary-artery calcification in young adults with end-stage renal disease who are undergoing dialysis. N Engl J Med. 2000; 342(20): 1478–83. PubMed Abstract \| Publisher Full Text\n\nCozzolino M, Gallieni M, Brancaccio D: Vascular calcification in uremic conditions: new insights into pathogenesis. Semin Nephrol. 2006; 26(1): 33–7. PubMed Abstract \| Publisher Full Text\n\nFukagawa M, Kazama JJ: The making of a bone in blood vessels: from the soft shell to the hard bone. Kidney Int. 2007; 72(5): 533–4. PubMed Abstract \| Publisher Full Text\n\nJablonski KL, Chonchol M: Vascular calcification in end-stage renal disease. Hemodial Int. 2013; 17Suppl 1: S17–21. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMoe SM, O'Neill KD, Duan D, et al.: Medial artery calcification in ESRD patients is associated with deposition of bone matrix proteins. Kidney Int. 2002; 61(2): 638–47. PubMed Abstract \| Publisher Full Text\n\nShanahan CM, Crouthamel MH, Kapustin A, et al.: Arterial calcification in chronic kidney disease: key roles for calcium and phosphate. Circ Res. 2011; 109(6): 697–711. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShantouf R, Kovesdy CP, Kim Y, et al.: Association of serum alkaline phosphatase with coronary artery calcification in maintenance hemodialysis patients. Clin J Am Soc Nephrol. 2009; 4(6): 1106–14. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVervloet M, Cozzolino M: Vascular calcification in chronic kidney disease: different bricks in the wall? Kidney Int. 2017; 91(4): 808–17. PubMed Abstract \| Publisher Full Text\n\nAgatston AS, Janowitz WR, Hildner FJ, et al.: Quantification of coronary artery calcium using ultrafast computed tomography. J Am Coll Cardiol. 1990; 15(4): 827–32. PubMed Abstract \| Publisher Full Text\n\nMao SS, Pal RS, McKay, et al.: Comparison of coronary artery calcium scores between electron beam computed tomography and 64-multidetector computed tomographic scanner. J Comput Assist Tomogr. 2009; 33(2): 175–8. PubMed Abstract \| Publisher Full Text\n\nReynolds JL, Joannides AJ, Skepper JN, et al.: Human vascular smooth muscle cells undergo vesicle-mediated calcification in response to changes in extracellular calcium and phosphate concentrations: a potential mechanism for accelerated vascular calcification in ESRD. J Am Soc Nephrol. 2004; 15(11): 2857–67. PubMed Abstract \| Publisher Full Text\n\nStenvinkel P, Ketteler M, Johnson RJ, et al.: IL-10, IL-6, and TNF-alpha: central factors in the altered cytokine network of uremia--the good, the bad, and the ugly. Kidney Int. 2005; 67(4): 1216–33. PubMed Abstract \| Publisher Full Text\n\nLondon GM, De Vernejoul MC, Fabiani F, et al.: Secondary hyperparathyroidism and cardiac hypertrophy in hemodialysis patients. Kidney Int. 1987; 32(6): 900–7. PubMed Abstract \| Publisher Full Text\n\nRostand SG, Drueke TB: Parathyroid hormone, vitamin D, and cardiovascular disease in chronic renal failure. Kidney Int. 1999; 56(2): 383–92. PubMed Abstract \| Publisher Full Text\n\nBogin E, Massry SG, Harary I: Effect of parathyroid hormone on rat heart cells. J Clin Invest. 1981; 67(4): 1215–27. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKatoh Y, Klein KL, Kaplan RA, et al.: Parathyroid hormone has a positive inotropic action in the rat. Endocrinology. 1981; 109(6): 2252–4. PubMed Abstract \| Publisher Full Text\n\nMitsnefes MM, Kimball TR, Kartal J, et al.: Progression of left ventricular hypertrophy in children with early chronic kidney disease: 2-year follow-up study. J Pediatr. 2006; 149(5): 671–5. PubMed Abstract \| Publisher Full Text\n\nEl-Husseini AA, Sheashaa HA, Hassan NA, et al.: Echocardiographic changes and risk factors for left ventricular hypertrophy in children and adolescents after renal transplantation. Pediatr Transplant. 2004; 8(3): 249–54. PubMed Abstract \| Publisher Full Text\n\nMatteucci MC, Wuhl E, Picca S, et al.: Left ventricular geometry in children with mild to moderate chronic renal insufficiency. J Am Soc Nephrol. 2006; 17(1): 218–26. PubMed Abstract \| Publisher Full Text\n\nMitsnefes MM, Daniels SR, Schwartz SM, et al.: Severe left ventricular hypertrophy in pediatric dialysis: prevalence and predictors. Pediatr Nephrol. 2000; 14(10–11): 898–902. PubMed Abstract \| Publisher Full Text\n\nFoley RN, Parfrey PS, Morgan J, et al.: Effect of hemoglobin levels in hemodialysis patients with asymptomatic cardiomyopathy. Kidney Int. 2000; 58(3): 1325–35. PubMed Abstract \| Publisher Full Text\n\nLevin A, Djurdjev O, Thompson C, et al.: Canadian randomized trial of hemoglobin maintenance to prevent or delay left ventricular mass growth in patients with CKD. Am J Kidney Dis. 2005; 46(5): 799–811. PubMed Abstract \| Publisher Full Text\n\nPaoletti E, Cassottana P, Bellino D, et al.: Left ventricular geometry and adverse cardiovascular events in chronic hemodialysis patients on prolonged therapy with ACE inhibitors. Am J Kidney Dis. 2002; 40(4): 728–36. PubMed Abstract \| Publisher Full Text\n\nRoger SD, McMahon LP, Clarkson A, et al.: Effects of early and late intervention with epoetin alpha on left ventricular mass among patients with chronic kidney disease (stage 3 or 4): results of a randomized clinical trial. J Am Soc Nephrol. 2004; 15(1): 148–56. PubMed Abstract \| Publisher Full Text\n\nMorris KP, Skinner JR, Hunter S, et al.: Cardiovascular abnormalities in end stage renal failure: the effect of anaemia or uraemia? Arch Dis Child. 1994; 71(2): 119–22. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFoley RN, Parfrey PS, Harnett JD, et al.: Clinical and echocardiographic disease in patients starting end-stage renal disease therapy. Kidney Int. 1995; 47(1): 186–92. PubMed Abstract \| Publisher Full Text\n\nLondon GM, Pannier B, Guerin AP, et al.: Alterations of left ventricular hypertrophy in and survival of patients receiving hemodialysis: follow-up of an interventional study. J Am Soc Nephrol. 2001; 12(12): 2759–67. PubMed Abstract\n\nAL-ETHAWI AESD: Replication Data for: Toward Cardiovascular Risks in Children with Chronic Kidney Disease. 2018. http://www.doi.org/10.17605/OSF.IO/WY28B" }	[ { "id": "46500", "date": "11 Apr 2019", "name": "Roy Oomen Mathew", "expertise": [ "Reviewer Expertise Chronic kidney disease", "cardiovascular disease in chronic kidney disease", "community acquired acute kidney injury." ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nFadhil Al-Doori et al. have written a report of a prospective cohort analysis of echocardiographic features in CKD. The main finding is that of greater LVMI in patients with CKD that seems to correlate with duration of CKD and lack of significant valvular calcification.\n\nThe study design is reported as prospective cohort but there was no clear prospective analysis completed. Most of the analyses demonstrated is cross sectional. In the limitations section it is mentioned that repeated analyses to assess changes over time is planned. It is not clear to me why this preliminary analysis would be submitted with very inconclusive data.\nMany issues with presentation are present: Introduction: The entire introduction is about adult CVD with only a statement in the last paragraph that little is known about CVD in pediatric CKD. There needs to be more background about pediatric CKD.\nIt's not clear to me why diastolic function and LV EF were not available. This seems to be a routine parameter in most echo readings. That needs to be explained in the methods section as well as the limitation section. Was it just not recorded for the study or was it actually not in the echo report - the echocardiographer never calculated those values?\n\nResults: There are far too many sections describing baseline parameters, as well figures and graphs. Figures 1-5 and Tables 2 and 3 all need to be condensed into Table 1. Then a single section can speak to background features and lab values. Many of the lab values are not surprising in terms of differences between CKD and normal - so not clear individual sections need to be dedicated to how each lab value is different between cases and controls.\n\nThe meat of the study should be in the Echo findings.\nWhat are the differences between cases and controls? You can subsequently look at differences between your eGFR classifications (< and >= 15) and duration of CKD (< and >= 1yr). These are the interesting findings.\n\nYou have an outcome of death and it seems awkward to have the outcome as a predictor in Table 6. It seems more appropriate to perform a multivariate analysis of death as an outcome and see how echocardiographic parameters compare to CKD and HTN and other lab parameters to predict death.\nDiscussion: The first half of the discussion is spent describing results (which should not be there) and talking about biochemical abnormalities in CKD which is already well known. Essentially the appropriate discussion section should starting at page 9 paragraph 4 - which starts \"Left Ventricular Hypertrophy....\"\nOverall - This manuscript seems premature especially since follow up data is planned. It is not clear to me what can be presented except a very brief report with the follow up data.\nI would recommend revising this when the follow up data is available.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4621", "date": "03 May 2019", "name": "Abd El-Salam El-Ethawi", "role": "Author Response", "response": "Dear Dr Mathew, Thanks a lot for your scientific critical appraisal. We appreciate your time spent on reading our manuscript. We do agree with most of your comments and suggestions. They will, definitely, be addressed in the upcoming version - when the follow up data will be available. Looking forward to having your \"next go\" after revision. Sincerely, Al-Ethawi" } ] }, { "id": "83112", "date": "04 May 2021", "name": "William Primack", "expertise": [ "Reviewer Expertise Glomerular disease", "workforce", "communication" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAl-Doori et al. report on echocardiographic findings on a cross-sectional (not prospective) cohort of children aged 2-14 with CKD. Their principal finding confirms other studies showing an association between the duration of hypertension and increased LVMI. They also find an association between increased LVMI and anemia, duration of CKD, and elevated PTH.\nThe authors do a quite comprehensive review of the literature on cardiovascular disease and CKD, but nearly all the articles report on findings in adults. Since this article was submitted 2 ½ years ago there have been several studies and reviews published on CV disease and pediatric CKD. Several are listed at the bottom of this review.\nI commend the authors for performing the study in an area of limited resources.\nThe authors might consider performing multivariate analyses of their data to see if, for example, the degree of anemia is better correlated with increased LVMI apart from hypertension or similarly if duration of CKD, eGFR, or PTH levels are independently associated with increased LVMI.\nNearly all the data in the figures could better presented in tablular form with means and S.D. or S.E.The authors should consider reporting in the baseline data whether any of the study patients were receiving EPO or antihypertensives.\nMortality in the studied cohort was high (18% in 13 months). Were any of the deaths related to CV issues? Were any of the deceased study patients autopsied and, if so, were the echocardiographic findings of increased LVMI (and no calcifications) in all who died supported by the autopsy results?\nBy the time this study was done, the Fifth Report on Hypertension in Children had been published and should have been the reference values used, as was the modified Schwartz equation for eGFR based on the CKIDS data.\nIf the authors are willing/able to do a careful statistical analysis as mentioned above, this paper would have more value. A couple of thoughts:\nHow were the authors able to identify the duration of kidney disease in those patients with an unknown cause since they comment that this group often presented later in the course of their CKD?\n\nWhy do patients with hereditary nephropathy (I assume most with cytinosis and not genetically associated FSGS.), have less frequent elevated LVMI despite having CKD from birth?\nSeveral recent references (via Pubmed): Cardiovascular risk factors in children on dialysis: an update.\nQuerfeld U, Schaefer F. Pediatr Nephrol. 2020 Jan;35(1):41-57. doi: 10.1007/s00467-018-4125-x. Epub 2018 Oct 31. PMID: 30382333\n\nCardiovascular disease risk among children with focal segmental glomerulosclerosis: a report from the chronic kidney disease in children study.\nSethna CB, Ng DK, Jiang S, Saland J, Warady BA, Furth S, Meyers KE. Pediatr Nephrol. 2019 Aug;34(8):1403-1412. doi: 10.1007/s00467-019-04229-3. Epub 2019 Mar 22. PMID: 30903375\n\nPrevalence of atheromatous and non-atheromatous cardiovascular disease by age in chronic kidney disease.\nVillain C, Metzger M, Combe C, Fouque D, Frimat L, Jacquelinet C, Laville M, Briançon S, Klein J, Schanstra JP, Robinson BM, Mansencal N, Stengel B, Massy ZA. Nephrol Dial Transplant. 2020 May 1;35(5):827-836. doi: 10.1093/ndt/gfy277. PMID: 30169874\nAdiposity, Sex, and Cardiovascular Disease Risk in Children With CKD: A Longitudinal Study of Youth Enrolled in the Chronic Kidney Disease in Children (CKiD) Study.\nBrady TM, Roem J, Cox C, Schneider MF, Wilson AC, Furth SL, Warady BA, Mitsnefes M. Am J Kidney Dis. 2020 Aug;76(2):166-173. doi: 10.1053/j.ajkd.2020.01.011. Epub 2020 May 7. PMID: 32389356\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "6801", "date": "15 Jun 2021", "name": "Abd El-Salam El-Ethawi", "role": "Author Response", "response": "Thanks a lot for your revision and feedback. We do agree with most of your points. It was actually initial results of that cohort study. We are working currently on the final version of it and we will be honoured to have you as a reviewer. Thanks again; your time & effort are massively appreciated. Al-Ethawi (Corresponding author)" } ] } ]	1	https://f1000research.com/articles/7-1794
https://f1000research.com/articles/7-1459/v1	13 Sep 18	{ "type": "Research Note", "title": "Strigolactone GR24 upregulates Nrf2 target genes and may protect against oxidative stress in skeletal muscle", "authors": [ "Shalem Raju Modi", "Tarja Kokkola", "Tarja Kokkola" ], "abstract": "GR24 is a synthetic strigolactone analog, demonstrated to regulate the development of plants and arbuscular mycorrhizal fungi. GR24 possesses anti-cancer and anti-apoptotic properties, enhances insulin sensitivity and mitochondrial biogenesis in skeletal myotubes, inhibits adipogenesis, decreases inflammation in adipocytes and macrophages and downregulates the expression of hepatic gluconeogenic enzymes. Transcription factor Nrf2 (Nuclear factor (erythroid-derived 2)-like 2) is a master regulator of antioxidant response, regulating a multitude of genes involved in cellular stress responses and anti-inflammatory pathways, thus maintaining cellular redox homeostasis. Nrf2 activation reduces the deleterious effects of mitochondrial toxins and has multiple roles in promoting mitochondrial function and dynamics. We studied the role of GR24 on gene expression in rat L6 skeletal muscle cells which were differentiated into myotubes. The myotubes were treated with GR24 and analyzed by microarray gene expression profiling. GR24 upregulated the cytoprotective transcription factor Nrf2 and its target genes, activating antioxidant defences, suggesting that GR24 may protect skeletal muscle from the toxic effects of oxidative stress.", "keywords": [ "Strigolactone", "GR24", "Nrf2", "Oxidative stress", "Microarray" ], "content": "Introduction\n\nStrigolactones are carotenoid-derived phytohormones with endogenous roles in regulating plant growth and exogenous roles in establishing symbiosis of host plant with arbuscular mycorrhizal fungi1. Strigolactones induce beneficial effects in mycorrhiza, such as mitochondrial biogenesis and ATP production2–4. The anti-cancer properties5–8 and anti-inflammatory potential9 of strigolactones have recently been investigated in mammalian cells. The positive effects of strigolactone analog GR24 in enhancing insulin sensitivity, mitochondrial function and inhibiting adipogenesis and inflammation in insulin-sensitive cells have also been demonstrated10,11.\n\nNrf2 (Nuclear factor (erythroid-derived 2)-like 2) activates gene transcription by binding to the antioxidant response element (ARE) in the promoters of its target genes. It regulates multiple biological functions ranging from cellular redox metabolism, detoxification, heme, lipid and glucose metabolism, NADPH generation, autophagy, apoptosis, xenobiotic stress reponse to inflammation by interacting with its target genes, regulating an extensive antioxidant protein network. Nrf2 is associated with disease pathologies like cancer12, hepatotoxicity13, cardiovascular disease14 and neurodegenerative diseases15.\n\nMitochondria are the major sites of reactive oxygen species (ROS) production and also the targets of their toxic effects. Mitochondrial dysfunction has been associated with the development of insulin resistance, and diabetes is known to induce oxidative stress through the overproduction of ROS and ROS-induced DNA damage16. Nrf2 activation defends against mitochondrial toxins and ROS and affects mitochondrial function by regulating mitochondrial biogenesis, mitochondrial fatty acid oxidation, respiration, ATP production and mitochondrial dynamics17. With its versatile protective mechanism against oxidative stress, pancreatic β-cell apoptosis and insulin resistance, Nrf2 has become a promising therapeutic target for the treatment of type 2 diabetes18.\n\nWe have demonstrated that GR24 ameliorates insulin sensitivity, stimulates mitochondrial biogenesis and ATP production and upregulates genes regulating mitochondrial function in L6 myotubes10. Very recently, the efficacy of GR24 in promoting cytoprotective responses via Nrf2 activation was reported in hepatic and macrophage cell lines19. This work reports a transcriptomic study revealing the potential beneficial effects of GR24 in upregulating Nrf2 and its target genes involved in detoxification, carbohydrate and lipid metabolism, heme metabolism, NADPH regeneration and oxidative stress in L6 myotubes, thus contributing to metabolic homeostasis.\n\n\nMethods\n\nResults from the methods described in this study have been published previously10, although the analyses described here are published for the first time.\n\nRat L6 myoblasts (ATCC, Manassas, VA, USA) were maintained in Dulbecco’s modified Eagle’s medium (DMEM 4.5 g/l glucose, Lonza, Basel, Switzerland) supplemented with 10% FBS (Hyclone, Pasching, Austria), 2 mM L-glutamine (Lonza) and 1% penicillin/streptomycin (Lonza) at +37°C in a humidified atmosphere of 5% CO2. The cells were seeded in multiwell plates at 2 × 104 cells/cm2 one day before starting the differentiation. Myoblasts were differentiated into myotubes by switching into α-MEM media (Gibco, Paisley, UK) supplemented with 2% horse serum (Gibco) and 1% penicillin/streptomycin. GR24 (3E,3aR,8bS)-3-[[(2 S)-4-methyl-5-oxo-2H-furan-2-yl] oxymethylidene]-4,8b-dihydro-3aH-indeno[1,2-b]furan-2-one, Chiralix, Nijmegen, Netherlands) was dissolved in DMSO. The control samples had equivalent DMSO concentration.\n\nL6 myotubes were treated with 60 µM GR24 at 7 days of differentiation for 24 h in three independent experiments, resulting in three replicate microarrays in each treatment group. Total RNAs were extracted with RNeasy Mini kit (Qiagen, Hilden, Germany) and RNA quality was assessed with the Agilent Bioanalyzer 2100 (Agilent Technologies, Espoo, Finland). Total RNA (200 ng) was converted to cDNA with Agilent AffinityScript RNase Block, labelled according to manufacturer’s instructions and purified using RNeasy mini spin columns (Qiagen). RNA Spike-In Kit (Exiqon, Vedbaek, Denmark) was used to monitor the success of labelling. Samples were then mixed with blocking agent, fragmentation and hybridization buffer, and were hybridized to Agilent SurePrint G3 Rat GE 8×60 K Microarrays for 17 h at 65°C before washing and scanning with Agilent Scanner G2505C using manufacturer’s protocols. Agilent Feature Extraction software was used to extract data from raw microarray image files10.\n\nThe data was processed with limma package (version 3.28.21) of the Bioconductor software20. Microarray data are MIAME compliant. Differentially expressed transcripts were analysed by Bayes moderated t-statistics followed by the Benjamini-Hochberg correction method to control false discovery rate (FDR) with significance threshold set at p < 0.0520,21.\n\n\nResults and discussion\n\nThe top 5 GR24-upregulated genes (glutathione S-transferase alpha 1, metallothionein 1M, heme oxygenase 1, glutathione S-transferase alpha 2, and sequestosome 1) were found to be known targets of Nrf2 (Table 1). As the prototypical Nrf2 target gene NAD(P)H quinone dehydrogenase 1 (Nqo1) was also found high on the list, it was compelling to look into the extensive list of Nrf2 target genes. We found 56 known Nrf2 target genes22, including Nrf2 itself, to be upregulated by GR24 treatment (Table 1). Prior to our study, the effects of strigolactones treatment on the mammalian cell transcriptome have only been investigated in human osteosarcoma cells, where strigolactone analogs mainly upregulated the heat shock stress proteins and downregulated the cell cycle5.\n\nActivation of Nrf2 signaling is known to have beneficial effects in cancer, metabolic syndrome, obesity, nephropathy, retinopathy, neuropathy, and β-cell protection. Nrf2 regulates the transcription of genes involved in antioxidant, detoxification and metabolic processes23. In our study, GR24 enhanced the expression of the Nrf2-dependent antioxidant genes Nqo1 and heme oxygenase 1, which are known to block inflammatory pathways24. GR24 has recently been shown to alleviate inflammation and upregulate these two NRF2 target genes in hepatic and macrophage cells19.\n\nGlutathione and associated enzymes form an important antioxidant defence system. Treatment with GR24 was found to increase root glutathione content in plants25, but there are no previous reports on the effects of strigolactones on the glutathione system in mammalian cells. Our results show that GR24 treatment upregulated glutamate-cysteine ligase, the rate-limiting enzyme in glutathione synthesis (Table 1). Many components of the glutathione system are upregulated by Nrf2, and the elevations in the expression of glutaredoxin, glutathione peroxidase 4 and several glutathione S-transferases in GR24-treated cells were evident in our results (Table 1).\n\n\nConclusions\n\nGR24 upregulated the cytoprotective transcription factor Nrf2 and its target genes, which are involved in detoxification, antioxidant systems, carbohydrate and lipid metabolism, heme and iron metabolism, NADPH regeneration and regulation of transcription in L6 myotubes. The ability of GR24 to enhance key cellular defence mechanisms through the activation of Nrf2 signaling may beneficially protect cells from oxidative stress and inflammation, and may alleviate the complications associated with metabolic syndrome.\n\n\nData availability\n\nMicroarray raw data have been deposited in the NCBI Gene Expression Omnibus (GEO), accession number GSE90833: https://identifiers.org/geo/GSE90833.", "appendix": "Grant information\n\nThis work was financially supported by the Academy of Finland, Diabetes Wellness Finland, Finnish Diabetes Research Foundation and Sigrid Jusélius Foundation.\n\nThe funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe thank Mikko Kivento for carrying out the bioinformatics on microarray analysis at the Functional Genomics Unit of the University of Helsinki.\n\n\nReferences\n\nBrewer PB, Koltai H, Beveridge CA: Diverse roles of strigolactones in plant development. Mol Plant. 2013; 6(1): 18–28. PubMed Abstract \| Publisher Full Text\n\nBesserer A, Puech-Pagès V, Kiefer P, et al.: Strigolactones stimulate arbuscular mycorrhizal fungi by activating mitochondria. PLoS Biol. 2006; 4(7): e226. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBesserer A, Bécard G, Jauneau A, et al.: GR24, a synthetic analog of strigolactones, stimulates the mitosis and growth of the arbuscular mycorrhizal fungus Gigaspora rosea by boosting its energy metabolism. Plant Physiol. 2008; 148(1): 402–413. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAl-Babili S, Bouwmeester HJ: Strigolactones, a novel carotenoid-derived plant hormone. Annu Rev Plant Biol. 2015; 66: 161–186. PubMed Abstract \| Publisher Full Text\n\nPollock CB, McDonough S, Wang VS, et al.: Strigolactone analogues induce apoptosis through activation of p38 and the stress response pathway in cancer cell lines and in conditionally reprogrammed primary prostate cancer cells. Oncotarget. 2014; 5(6): 1683–1698. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMayzlish-Gati E, Laufer D, Grivas CF, et al.: Strigolactone analogs act as new anti-cancer agents in inhibition of breast cancer in xenograft model. Cancer Biol Ther. 2015; 16(11): 1682–1688. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCroglio MP, Haake JM, Ryan CP, et al.: Analogs of the novel phytohormone, strigolactone, trigger apoptosis and synergize with PARP inhibitors by inducing DNA damage and inhibiting DNA repair. Oncotarget. 2016; 7(12): 13984–14001. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHasan MN, Choudhry H, Razvi SS, et al.: Synthetic strigolactone analogues reveal anti-cancer activities on hepatocellular carcinoma cells. Bioorg Med Chem Lett. 2018; 28(6): 1077–1083. PubMed Abstract \| Publisher Full Text\n\nZheng JX, Han YS, Wang JC, et al.: Strigolactones: a plant phytohormone as novel anti-inflammatory agents. Medchemcomm. 2017; 9(1): 181–188. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nModi S, Yaluri N, Kokkola T, et al.: Plant-derived compounds strigolactone GR24 and pinosylvin activate SIRT1 and enhance glucose uptake in rat skeletal muscle cells. Sci Rep. 2017; 7(1): 17606. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nModi S, Yaluri N, Kokkola T: Strigolactone GR24 and pinosylvin attenuate adipogenesis and inflammation of white adipocytes. Biochem Biophys Res Commun. 2018; 499(2): 164–169. PubMed Abstract \| Publisher Full Text\n\nMenegon S, Columbano A, Giordano S: The Dual Roles of NRF2 in Cancer. Trends Mol Med. 2016; 22(7): 578–593. PubMed Abstract \| Publisher Full Text\n\nJadeja RN, Upadhyay KK, Devkar RV, et al.: Naturally Occurring Nrf2 Activators: Potential in Treatment of Liver Injury. Oxid Med Cell Longev. 2016; 2016: 3453926. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChen QM, Maltagliati AJ: Nrf2 at the heart of oxidative stress and cardiac protection. Physiol Genomics. 2018; 50(2): 77–97. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDinkova-Kostova AT, Kostov RV, Kazantsev AG: The role of Nrf2 signaling in counteracting neurodegenerative diseases. FEBS J. 2018. PubMed Abstract \| Publisher Full Text\n\nMontgomery MK, Turner N: Mitochondrial dysfunction and insulin resistance: an update. Endocr Connect. 2015; 4(1): R1–R15. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHolmström KM, Kostov RV, Dinkova-Kostova AT: The multifaceted role of Nrf2 in mitochondrial function. Curr Opin Toxicol. 2016; 1: 80–91. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDavid JA, Rifkin WJ, Rabbani PS, et al.: The Nrf2/Keap1/ARE Pathway and Oxidative Stress as a Therapeutic Target in Type II Diabetes Mellitus. J Diabetes Res. 2017; 2017: 4826724. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTumer TB, Yilmaz B, Ozleyen A, et al.: GR24, a synthetic analog of Strigolactones, alleviates inflammation and promotes Nrf2 cytoprotective response: In vitro and in silico evidences. Comput Biol Chem. 2018; 76: 179–190. PubMed Abstract \| Publisher Full Text\n\nRitchie ME, Phipson B, Wu D, et al.: limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015; 43(7): e47. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStorey JD, Tibshirani R: Statistical significance for genomewide studies. Proc Natl Acad Sci U S A. 2003; 100(16): 9440–9445. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHayes JD, Dinkova-Kostova AT: The Nrf2 regulatory network provides an interface between redox and intermediary metabolism. Trends Biochem Sci. 2014; 39(4): 199–218. PubMed Abstract \| Publisher Full Text\n\nSykiotis GP, Habeos IG, Samuelson AV, et al.: The role of the antioxidant and longevity-promoting Nrf2 pathway in metabolic regulation. Curr Opin Clin Nutr Metab Care. 2011; 14(1): 41–48. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAhmed SM, Luo L, Namani A, et al.: Nrf2 signaling pathway: Pivotal roles in inflammation. Biochim Biophys Acta. 2017; 1863(2): 585–597. PubMed Abstract \| Publisher Full Text\n\nMarquez-Garcia B, Njo M, Beeckman T, et al.: A new role for glutathione in the regulation of root architecture linked to strigolactones. Plant Cell Environ. 2014; 37(2): 488–498. PubMed Abstract \| Publisher Full Text" }	[ { "id": "38815", "date": "02 Oct 2018", "name": "David A. Hood", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nInteresting investigation of the effect of a plant-based compound to change gene expression in muscle cells, however the study is very limited in scope.\n1. All reported data show increases in mRNA. Are there no significant decreases that are noteworthy? 2. Have any changes in mRNA using microarrays been verified with Q-PCR? This is standard practice; 3. Could a few protein blots be included to strengthen the data and take the interpretation to the protein level? 4. The title assumes that the mRNA changes lead to protein changes which would impact oxidative stress.The title needs to be re-worded unless actual measures of oxidative stress are performed. this would be relatively easy with an oxi-blot, or ROS measure in cells.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4072", "date": "14 Nov 2018", "name": "SHALEM RAJU MODI", "role": "Author Response", "response": "We thank the referee for carefully evaluating our research note. He has correctly identified that the study has its limitations. As we have been engaged in other studies and have not been able to dedicate more time for continuing on this track, we decided to submit our microarray analyses in the current format. This Research Note format facilitates publishing small studies, such as our study with only a single table. 1. All reported data show increases in mRNA. Are there no significant decreases that are noteworthy? In the current short publication, we do not report all significant findings. A total of 16,153 transcripts were differentially regulated (9,838 up- and 6,315 downregulated) to the GR24 treatment. Some of the differentially regulated transcripts have been described in our previous publication (Modi S, Yaluri N, Kokkola T, et al.: Plant-derived compounds strigolactone GR24 and pinosylvin activate SIRT1 and enhance glucose uptake in rat skeletal muscle cells. Sci Rep. 2017; 7(1): 17606.). Nevertheless, the most interesting finding from the microarray experiments was the strong upregulation of Nrf2 target transcripts by GR24 treatment. 2. Have any changes in mRNA using microarrays been verified with Q-PCR? This is standard practice; 3. Could a few protein blots be included to strengthen the data and take the interpretation to the protein level? 4. The title assumes that the mRNA changes lead to protein changes which would impact oxidative stress.The title needs to be re-worded unless actual measures of oxidative stress are performed. this would be relatively easy with an oxi-blot, or ROS measure in cells. These are very valid questions. The suggested qPCR analyses, protein blots and measures of oxidative stress would be essential next steps to verify the changes in gene expression. For that reason, we published our preliminary analyses as a Research Note, as additional analyses would have been required in order to produce a complete full-length article. When drafting the title, we avoided making strong claims on the effects on oxidative stress. After reading the suggestions by the referee, we have reconsidered the title, and will submit a new version of the article with a title “Strigolactone GR24 upregulates target genes of the cytoprotective transcription factor Nrf2 in skeletal muscle”." } ] }, { "id": "39708", "date": "31 Oct 2018", "name": "Tugba Boyunegmez Tumer", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn current research note, a microarray study was carried out to analyze the effect of GR24, a synthetic strigolactone analog, on the gene expression profile of rat L6 myotubes. 60 µM of GR24 was used to treat L6 myotubes at 7th day of differentiation for 24 h in three independent experiments. Microarray data was processed and differentially expressed transcripts were analyzed. Upregulated genes were partly presented in Table 1. Accordingly, GR24 upregulated the cytoprotective transcription factor Nrf2 and its several target genes involved in antioxidant defense and phase II detoxification processes. The fold values especially for GSTa1 and a2, NQO1, GGt1, Hmox1, Mtm1m, Sqstm1 were considerably and significantly high (above 6 fold). Authors previously showed that GR24 (60 µM) ameliorates insulin sensitivity, stimulates mitochondrial biogenesis and ATP production and upregulates genes regulating mitochondrial function in L6 myotubes (Modi et al., 20171).\nAlso, in this previous report it was shown that 9838 transcripts were upregulated and 6315 transcripts were downregulated in L6 myotubes as a result of 60 µM GR24 treatment. Up and down regulated transcripts involved in insulin signaling and mitochondrial function in L6 myaotubes have been presented in this previous paper (Modi et al., 20171). In the present research note, authors represented up regulated transcripts involved in oxidative stress and cytoprotective signaling when myotubes were treated with 60 µM GR24. Since the work can be classified as original and the results are remarkable, data deserves to be published in F1000Research as research note. However, the data is limited to reach the conclusion suggesting that GR24 may protect skeletal muscle from the toxic effects of oxidative stress and thus contributing to metabolic homeostasis. In the scope of report, there is no any biological assay or disease modelling evaluating the beneficial effects of GR24 on myotubes. Furthermore, gene expression profiling does not always reflect the changes on the protein level. Regarding these facts, title also should not contain the assumption suggesting that GR24 may protect against oxidative stress in skeletal muscle.\nIn my opinion, both conclusion and title need revision and if possible it is better to represent additional data and interpretation in terms of down regulated genes in specified signaling pathways. Also could you please add an explanation about why you chose 60 µM concentration?\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4195", "date": "14 Nov 2018", "name": "SHALEM RAJU MODI", "role": "Author Response", "response": "We thank the referee for kindly evaluating our research note and for the valid comments and suggestions. We agree with the comments and hence, we changed the discussion part and also the title of the manuscript. Also could you please add an explanation about why you chose 60 µM concentration? We initially tested GR24 at various concentrations from 20 µM – 100 µM. We observed that at 60 µM concentration, GR24 significantly upregulated SIRT1 protein expression and enhanced glucose uptake in skeletal muscle cells (Sci Rep. 2017; 7(1): 17606.). Moreover, longer exposure to 100 µM GR24 had a small but significant cytotoxic effect. Hence, we have chosen 60 µM to be an ideal concentration for all our experiments." } ] } ]	1	https://f1000research.com/articles/7-1459
https://f1000research.com/articles/7-1515/v1	21 Sep 18	{ "type": "Research Article", "title": "Young key affected population in Myanmar: are there any challenges in seeking information and care for HIV/sexually transmitted infections and reproductive health?", "authors": [ "Kyaw Min Htut", "Myo Myo Mon", "Zin Mar Aye", "Lwin Lwin Ni", "Myo Myo Mon", "Zin Mar Aye", "Lwin Lwin Ni" ], "abstract": "Background: Unmet needs and barriers in seeking HIV/STI and RH information and care are present especially among young key affected population (YKAP). Therefore, the study was conducted to determine the health seeking behaviors of YKAP regarding HIV/STI and RH, and challenges in seeking health information and care. Methods: A cross-sectional, mixed-methods study was conducted at two large cities in Myanmar. Face-to-face interviews were conducted with YKAP aged 15-24 years. In-depth interviews and key informant interviews were done with YKAP and health care providers. Descriptive statistics and bivariate analyses were done for quantitative data and thematic analysis was applied for qualitative data. Results: A total of 119 young men who have sex with men (YMSM) and 123 young female sex workers (YFSW) included in the study. Mean age of YMSM and YFSW were 20.9±2.4 and 21.7±2.2 years. Over 30% of YMSM and 49.3% of YFSW had experience of any STI symptom. Particularly, 17% of YMSM and 10% of YFSW had genital ulcer, and majority sought health care at NGO clinics. About 37% of YMSM and 40% of YFSW visited Drop-in-center (DIC) within one to six months. Over 13% of YMSM and 14.6% of YFSW had challenges in seeking HIV/STI and RH information. YMSM/YFSW type and age of YMSM were associated with visit to DIC. Lesser proportions of Tha-nge (43.5%), younger age YMSM (66.7%), brothel-based YFSW (47.9%) visited DIC than others (p<0.05). Challenges and unmet needs expressed by YKAP were reluctance in asking health information, worry for future fertility, consequences of anal sex and contraception. Challenges expressed by providers were limited time during outreach service and difficulty in reaching entertainment-based sex workers. Conclusions: Special attention in provision of health information should be paid to YKAP since there is a considerable proportion of YKAP with unmet need in seeking HIV/STI/RH information and care.", "keywords": [ "Young Key Affected Population", "Men who have sex with men", "Female sex workers", "HIV", "Sexually Transmitted Infection", "Reproductive Health", "Myanmar" ], "content": "Introduction\n\nGlobally, one-fourth of the total population is young people aged 10–24 years, and they are most vulnerable from the global epidemic of HIV. Around one-third of all new HIV infections worldwide occurred among youth aged 15–24 years and about five million people aged 10–24 years were infected with HIV1. As described in latest census of Myanmar in 2014, the total population is 51.4 million, with 16 million young people, which accounts for 28% of the population2. One of the main health problems faced by young people results from sexual and reproductive health risk-taking behaviors, leading to unintended pregnancies and HIV/AIDS.\n\nKey affected populations in relation to HIV transmission were men who have sex with men (MSM), sex workers, people who inject drugs, and people in prisons3. In Myanmar, according to sentinel surveillance data, HIV prevalence among young key populations was higher than that of other populations. In particular, 5.5% and 7.9% among female sex workers aged 15–19 years and 20–24 years, respectively; and 9.1% and 8.6% among men who have sex with men aged 15–19 years and 20–24 years, respectively4. As mentioned in National Strategic Plan on HIV/AIDS for 2016–2020, HIV prevalence among female sex workers (FSW) and MSM were 14.6% and 11.6% respectively according to findings from the Integrated Biological and Behavioral Survey in 20155.\n\nPrevious studies have identified factors related to health care-seeking behaviors of young people. These factors included stigma and discrimination6, long waiting times, inconvenient locations of clinics, not knowing where to get the services, and negative attitudes among health care providers6,7. Globally, there are a number of studies indicating the presence of HIV‑related stigma in healthcare settings8,9. Discrimination is a major problem when seeking health care for HIV-infected individuals10. Research had found that consequences of HIV-related stigma on health-seeking behavior resulted in fear of receiving HIV testing, and delaying in responses such as adhering to treatment and preventive behaviours11,12. In a study in Laos, the main barriers were related to location of health facility, lack of awareness on availability of services and unfavorable attitude of healthcare providers7.\n\nWith regards the service utilization, the percentage of FSW who received an HIV test in the last 12 months was 71%, and the percentage of MSM who received an HIV test in the last 12 months was 48%13. Reducing the incidence of HIV among priority populations like MSM and FSW was described as one of the objectives to fulfil the goal of the current National Strategic Plan. It was also stated that efforts must be made to tailor services to reach the priority population of young people5. However, very few studies have been conducted among the YKAP in Myanmar identifying health-seeking behaviors and their perceived barriers. Therefore, current study was conducted to determine the health seeking behaviors regarding HIV/STI and reproductive health (RH), challenges and the unmet needs in seeking health information among YKAP.\n\n\nMethods\n\nA cross-sectional, mixed-methods study was conducted using both quantitative and qualitative methods among the YKAP, including young FSW (YFSW) and young MSM (YMSM) in Yangon and Mandalay, Myanmar, during February and June 2017. Yangon and Mandalay are two largest business cities of Myanmar where the YKAP community is larger than that of other areas.\n\nInclusion criteria:\n\n1. YFSW aged 15 to 24 years currently working as sex workers whose sex work was based either at brothels, entertainment places (karaoke, club, bar) or on the streets.\n\n2. YMSM aged 15 to 24 years who identified themselves as apwint (open type) or apone (hidden type) or tha-nge (male partner of either apwint or apone).\n\nOperational definition of MSMs according to their types5:\n\nApwint: Those who are biological males whose public and private gender identity is generally feminine, but they may dress as men or dress and act as females. Apwint are generally more ‘open’ MSM and some could be considered ‘transgender’.\n\nApone: Those who are biological males whose gender identity may be either masculine or feminine and may or may not express themselves femininely.\n\nTha Nge: Those who are biological males whose gender identity is masculine with a sexual preference for apwint and apone as well as for women, however they are often ‘hidden’ MSM.\n\nOutcome variables\n\nHealth-seeking behavior was measured on\n\n1) Ever receive HIV testing (Yes/No)\n\n2) STI treatment (Yes/No)\n\n3) Visit to DIC (Yes/No)\n\nIndependent variables\n\n1) Age: both continuous and categorical measurement\n\n2) Type of MSM: either apwint, apone or thange (as defined above)\n\n3) Type of FSW: either brothel-based, entertainment-based or street-based\n\n4) Education: either illiterate, read & write, primary school, middle school, high school, university\n\n5) Having income earning job: either yes,not always or no\n\n6) Any STI symptoms: yes or no\n\nChallenges and unmet needs were mainly discussed during in-depth interviews and key informant interviews.\n\nPurposive sampling was applied in recruiting YMSM and YFSW. Firstly, identification of the places for recruitment of the possible participants was made after discussion with the focal persons from the networks of FSW and MSM. FSWs were recruited from brothels, massage parlors, karaokes and soliciting sites on the streets according to the inclusion criteria. MSMs were recruited at beauty salons and gathering places along the streets. No recruitment was done through clinics and drop-in centers (DIC) to prevent bias in sampling those with good health-seeking behavior.\n\nConsidering proportions of MSM and FSW who seek HIV testing service in last 12 months as 20% and 30% according to a previous study5, 95% confidence level, precision of 0.1 and design effect of 1.5, the minimum required sample size for each population were 93 (YMSM) and 122 (YFSW) by using a sample size formula for one proportion.\n\nA structured questionnaire was developed for quantitative assessment and guidelines were developed for in-depth interviews (IDIs) and key informant interviews (KIIs) (Supplementary File 1, Supplementary File 2). Research assistants were trained at the Department of Medical Research before field data collection. Face-to-face interviews were carried out using a structured, pre-tested questionnaire by trained interviewers. In-depth interviews were also conducted with YKAP and key informant interviews were carried out with the service provider to explore their opinions and experiences (Supplementary File 3). Service providers are the focal persons from National AIDS Program and non-governmental organization (NGO)/international NGO working for the key populations. These IDIs and KIIs were conducted by two principal investigators who have experience of conducting qualitative interviews. Because of confidentiality issues, the interviews were not audio recorded. However, discussions were noted down by well-trained note takers.\n\nData entry was conducted using EpiData version 3.1 and analysis was conducted with SPSS version 16 for quantitative data. Descriptive statistics were shown as frequency/percentage for categorical variables and mean/median for continuous variables. Bi-variate analysis was done using the chi-squared test. Transcripts were prepared and manual thematic analysis was also applied for qualitative information.\n\nInformed consent was obtained from each participant after thorough explanation about the objectives of the study. Anonymity and confidentiality of the information were ensured using the code numbers and only researchers have accessed to the information. Ethics approval was also obtained from the Ethics Review Committee of The Department of Medical Research (Ethics/DMR/2016/091), Ministry of Health and Sports, Myanmar.\n\n\nResults\n\nSocio-demographic characteristics and family related information of participants are shown in Table 1. A total of 119 young men who have sex with men (YMSM) and 123 young female sex workers (YFSW) included in the assessment. The mean age of YMSM and YFSW was 20.9±2.4 and 21.7±2.2 years, respectively. Nearly 60% of YMSM were apwint (open type), 21.8% were apone (hidden type) and 19.3% were tha-nge (male partner of apwint or apone) as identified by themselves. Based on the place of sex work, YFSW included in the study were identified as brothel-based (40%), entertainment-based (karaoke/restaurant/nightclub/massage) (32.5%) and street-based (28.5%) respectively.\n\nYMSM, young men who have sex with men; YFSM, young female sex workers.\n\nRegarding their education status, 72.3% and 86.2 % of YMSM and YFSW, respectively, had completed primary school education, and 8.4% of YMSM were university graduates. Around one-fourth of both YMSM (27.1%) and YFSW (24.4%) were married. Median monthly income of YMSM and YFSW were 200,000 Kyats and 300,000 Kyats, respectively. Over 50% of YMSM and about 40% of YFSW were currently living with their parents. More than half of YMSMs’ parents/guardians accepted their sexual identity as MSM while only 22% of YFSWs’ parents/guardians accepted them as sex workers.\n\nFigure 1 describes the STI symptoms experienced by YMSM and YFSW; genital ulcer was most common for YMSM while white discharge was most common for YFSW. In particular, past incidence of genital ulcers was reported by 17% of YMSM and 11% of YFSW. Over 21% of YFSW suffered from white discharge while 7.6% of YMSM suffered from urethral discharge. Additionally, lower abdominal pain was also common in YFSW (18.7%). Table 2 shows the health-seeking behaviors of YMSM and YFSW regarding RH, STI and HIV. About 70% of YMSM and 56% of YFSW had experience of health-seeking for STI symptoms and the majority of them go to NGO clinics to treat STI. Over 90% of YMSM and YFSW have received HIV testing in the past and over 80% of them had tested for HIV within 6 months. The main reason for undergoing HIV testing is that they would like to know whether they have been infected with HIV or not. Over 90% of both YMSM and YFSW went to an NGO clinic for HIV testing. Regarding the utilization of drop-in centers (DICs), 79% of YMSM and 56.9% of YFSW have ever visited a DIC. Among them, 53.2% of YMSM and 38.6% of YFSW visited a DIC within the previous month.\n\nNGO, non-governmental organization; DIC, drop-in clinic.\n\nFigure 2 shows the barriers or limitations in receiving STI/HIV and RH information, and health care seeking. Just over 13% of YMSM and 14.6% of YFSW mentioned that they experience external and personal barriers towards seeking health information on RH. Similarly, 11% of YMSM and 12% of YFSW have barriers in seeking STI/HIV information.\n\nYMSM, young men who have sex with men; YFSM, young female sex workers.\n\nVisit to DICs among YKAP and their background characteristics are shown in Table 3. Type of MSM and age were associated with visiting a DIC among YMSM. A lesser proportion of tha-nge (43.5%) visited a DIC than apwint (85.7%) and apone (92.3%) (p=0.0001). A higher proportion of older YMSM visited DICs in comparing to younger MSM (84.3% vs. 66.7%, p=0.03). Among YFSWs, visiting a DIC was associated with their place of work: a significantly higher proportion of street-based YFSW (77.1%) visited DICs than those who worked in entertainment locations (50%) and brothels (47.9%) (p=0.01).\n\nDuring in-depth interviews and key informant interviews, different challenges and unmet needs in seeking health information and services were mentioned by YKAP as shown in Table (4). Common challenges mentioned by YMSMs were “financial problems” and “discrimination from health care providers”, while YFSWs stated their challenges as “no/limited time to access health service”, “reluctance in asking health information” and “restriction to go outside”. Regarding their unmet needs, most tha-nge (male partners of apwint and apone) expressed their concerns about the health consequences from having sexual relationship with MSM and future fertility. Other MSMs would like to know the consequences of anal sex and its treatment. Similarly, YFSWs also expressed that they have unmet needs concerning their future fertility and contraception. Most providers mentioned that it was difficult to reach and provide services to the girls from entertainment locations, such as karaoke.\n\nYMSM, young men who have sex with men; YFSM, young female sex workers.\n\nSelected quotations included:\n\n“… Currently, I’m living together with “achout” (apwint MSM), but I have a plan to marry a girl in the future. I worry about my fertility status at that time… afraid that I may not able to procreate …” (IDI 3, tha-nge, 23 years old).\n\n“… We can’t go outside everyday… we’re allowed to go outside for one day per week … only for 2–3 hours …” (IDI 6, FSW, 20 years old).\n\n“… We could educate the girls from brothel … They told us their experiences frankly… But, we couldn’t communicate frankly with the girls from karaoke … most of them didn’t accept …” (KII 2, out-reach health care provider).\n\n\nDiscussions and conclusions\n\nThe current study highlights the health-seeking behaviors, challenges and unmet needs of the YKAP (YMSM and YFSW) regarding reproductive health and STI/HIV services. Experience of any STI symptom was disclosed by some YMSM and YFSW. The majority of YMSM and YFSW have received HIV testing in the past. In addition, both YMSM and YFSW had experience of health-seeking for STI symptoms and majority of them sought health care from NGO clinics. We also found barriers and unmet needs in seeking health information on RH and STI/HIV among YKAP.\n\nPrevious studies have documented on the health care seeking behavior among key population like MSM and FSW. Studies conducted in China found that 40–60% of MSM had ever done HIV testing14–16, while a similar HIV testing rate among FSW was documented in a study done in Nigeria17. Moreover, Bartelsman et al. documented that the HIV testing rate was as low as 32.7% among MSM in Amsterdam18. In contrast to other studies, much higher proportions of YMSM and YFSW from current study had been tested for HIV in the past. Different socio-economic background, sampling strategy and cultural context might contribute to this discrepancy. However, findings on high HIV testing rates among YMSM was consistent with the previous study done in two large cities of Myanmar, Yangon and Monywa, in 201511. It was also supported by the evidence from the progress report of National AIDs Program in Myanmar, which found that HIV testing rates among MSM and FSW were dramatically increased between 2006 and 2010. Specifically, testing rates tripled in MSM and quadrupled in FSW19. In the current study, although the self-reported testing rates were high, we did not ask about the quality of testing services and did not verify these rates by other methods.\n\nExperience of STIs among FSW had been noted in previous studies20,21. In Bangladesh, 41.6% of FSW had experience of any STI symptom and many of them had unmet needs for SRH care21. A lesser proportion of YFSW from our study also had experience of STI symptoms in the past. On the other hand, 21.4% and 15.4% of MSM from Tanzania and Peru had any past STI symptoms, and similar findings were also identified in the current study22,23. However, in our study, experience of STI symptoms was noted according to their responses and was not validated by blood test. Access to RH information, including that concerning HIV/STI, is important for the YKAP.\n\nUnmet needs and the barriers in seeking sexual and reproductive health care was documented in previous studies24,25. The prevalence of unmet need was 25% among hotel-based FSW and 36% among street-based FSW according to a study in Bangladesh25. Another study also reported that over 50% of FSWs have faced barriers in seeking SRH care24. Common barriers included financial problems, shame about receiving care, unwillingness and unfriendly behavior of the provider. Certain proportions of YKAP from current study mentioned that they have challenges in seeking reproductive health information. Similarly, some of them have barriers in seeking STI/HIV information.\n\nThe current study has certain limitations. Findings on the information related to STI experience and HIV testing of key population may have some bias since we have to rely on the respondents’ answers and could not validate them by other methods. However, we tried to overcome the limitation by providing a thorough explanation about the study’s objectives. Furthermore, generalization of the study findings to other areas of Myanmar may also have limitations because the study participants were only from two large major cities, where many NGOs/international NGOs are working for these populations.\n\nSpecial attention in provision of health information should be paid to the YKAP since there is a considerable proportion of the YKAP with unmet needs in seeking RH information and care. Strengthening of health education activities regarding STI is recommended for YKAP, especially for YFSW who work in entertainment-based locations.\n\n\nData availability\n\nDataset 1. Complete answers to questionnaires for young men who have sex with men and young female sex workers. A key to the coding and abbreviations is also included. DOI: https://doi.org/10.5256/f1000research.16029.d217070.\n\nTranscripts from interviews are not available to maintain the confidentiality of the study subjects.", "appendix": "Grant information\n\nThe current study was fully funded by the World Health Organization.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary material\n\nSupplementary File 1. Questionnaire used to obtain data for young men who have sex with men.\n\nClick here to access the data.\n\nSupplementary File 2. Questionnaire used to obtain data for young female sex workers.\n\nClick here to access the data.\n\nSupplementary File 3. Guidelines for interviews with service providers and the young key affected population.\n\nClick here to access the data.\n\n\nReferences\n\nWorld Health Organization: Technical Brief: HIV and Young Transgender People. 2015. Reference Source\n\nDepartment of Population, Ministry of Labour, Immigration and Population,et al.: Myanmar Census 2014: Overview of the Results of the 2014 Population and Housing Census. 2017. Reference Source\n\nWorld Health Organization: Consolidated Guidelines on HIV Prevention, Diagnosis, Treatment and Care for Key Populations. Geneva: World Health Organization; 2014. PubMed Abstract\n\nDepartment of Public Health, Ministry of Health, Myanmar: Five Year Strategic Plan for Young People’s Health (2016-2020). Reference Source\n\nNational AIDS Program, Ministry of Health and Sports, Myanmar: Myanmar National Strategic Plan on HIV and AIDS 2016-2020. Reference Source\n\nLi L, Lee SJ, Thammawijaya P, et al.: Stigma, social support, and depression among people living with HIV in Thailand. AIDS Care. 2009; 21(8): 1007–13. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPhrasisombath K, Thomsen S, Sychareun V, et al.: Care seeking behaviour and barriers to accessing services for sexually transmitted infections among female sex workers in Laos: a cross-sectional study. BMC Health Serv Res. 2012; 12: 37. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZaidi MA, Griffiths R, Newson-Smith M, et al.: Impact of stigma, culture and law on healthcare providers after occupational exposure to HIV and hepatitis C. Cult Health Sex. 2012; 14(4): 379–91. PubMed Abstract \| Publisher Full Text\n\nZamberia AM: HIV-related stigma and access to health care among people living with HIV in Swaziland. Dev South Afr. 2011; 28(5): 669–680. Publisher Full Text\n\nSteward WT, Bharat S, Ramakrishna J, et al.: Stigma is associated with delays in seeking care among HIV-infected people in India. J Int Assoc Provid AIDS Care. 2013; 12(2): 103–109. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nUNESCO and Department of Medical Research, Ministry of Health, Myanmar: Multi-level risk and protective factors and HIV-related risk behaviours among young men who have sex with men (YMSM) in Myanmar. 2015. Reference Source\n\nThai Network of People Living with HIV/AIDS (TNP+): Index of Stigma and Discrimination against People Living with HIV/AIDS in Thailand. 2008. Reference Source\n\nNational AIDS Program, Ministry of Health, Myanmar: Global AIDS Response Progress Report. 2014. Reference Source\n\nCao B, Liu C, Durvasula M, et al.: Social Media Engagement and HIV Testing Among Men Who Have Sex With Men in China: A Nationwide Cross-Sectional Survey. J Med Internet Res. 2017; 19(7): e251. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRen XL, Wu ZY, Mi GD, et al.: Uptake of HIV Self-testing among Men Who have Sex with Men in Beijing, China: a Cross-sectional Study. Biomed Environ Sci. 2017; 30(6): 407–417. PubMed Abstract \| Publisher Full Text\n\nZhang TP, Liu C, Han L, et al.: Community engagement in sexual health and uptake of HIV testing and syphilis testing among MSM in China: a cross-sectional online survey. J Int AIDS Soc. 2017; 20(1): 21372. PubMed Abstract \| Free Full Text\n\nOkafor UO, Crutzen R, Ifeanyi O, et al.: HIV prevalence and high-risk behaviour of young brothel and non-brothel based female sex workers in Nigeria. BMC Res Notes. 2017; 10(1): 380. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBartelsman M, Joore IK, van Bergen JE, et al.: HIV testing week 2015: lowering barriers for HIV testing among high-risk groups in Amsterdam. BMC Infect Dis. 2017; 17(1): 529. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOo HN, Hone S, Fujita M, et al.: Evolution of the health sector response to HIV in Myanmar: progress, challenges and the way forward. J Virus Erad. 2016; 2(Suppl 4): 20–6. PubMed Abstract \| Free Full Text\n\nChanda MM, Ortblad KF, Mwale M, et al.: Contraceptive use and unplanned pregnancy among female sex workers in Zambia. Contraception. 2017; 96(3): 196–202. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWahed T, Alam A, Sultana S, et al.: Sexual and reproductive health behaviors of female sex workers in Dhaka, Bangladesh. PLoS One. 2017; 12(4): e0174540. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPerez-Brumer AG, Konda KA, Salvatierra HJ, et al.: Prevalence of HIV, STIs, and risk behaviors in a cross-sectional community- and clinic-based sample of men who have sex with men (MSM) in Lima, Peru. PLoS One. 2013; 8(4): e59072. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRoss MW, Nyoni J, Ahaneku HO, et al.: High HIV seroprevalence, rectal STIs and risky sexual behaviour in men who have sex with men in Dar es Salaam and Tanga, Tanzania. BMJ Open. 2014; 4(8): e006175. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWahed T, Alam A, Sultana S, et al.: Barriers to sexual and reproductive healthcare services as experienced by female sex workers and service providers in Dhaka city, Bangladesh. PLoS One. 2017; 12(7): e0182249. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKatz KR, McDowell M, Green M, et al.: Understanding the Broader Sexual and Reproductive Health Needs of Female Sex Workers in Dhaka, Bangladesh. Int Perspect Sex Reprod Health. 2015; 41(4): 182–90. PubMed Abstract \| Publisher Full Text\n\nHtut KM, Mon MM, Aye ZM, et al.: Dataset 1 in: Young key affected population in Myanmar: are there any challenges in seeking information and care for HIV/sexually transmitted infections and reproductive health? F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16029.d217070" }	[ { "id": "38597", "date": "08 Oct 2018", "name": "Tasnuva Wahed", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nComments: This is an important article in the field of sexual and reproductive health research, especially on sexually transmitted infections (STIs) and AIDs. This study intended to document the health seeking behaviours including barriers in health seeking and access to SRH information among young key populations, eg., female sex workers (FSWs), men having sex with men (MSMs) who were 15 to 24 years old in two cities of Myanmar. Following clarifications are needed on this script:\nGeneral observations:\nNeed a clear operational definition of reproductive health (RH) and services? The author only described about STI and HIV. There are other RH problems of FSWs, such as-unintended pregnancies or childbirth, abortion etc. Did the authors exclude these RH problems from their study? Need a clear definition of unmet needs in seeking health information. How did they calculate unmet need of seeking health information? Need separately discuss about qualitative and quantitative data collection methods in the methodology. Need one paragraph conclusion at the end of this script. At current state, it is finished with limitations following recommendations without a conclusion.\nSpecific observations: Abstract:\n\nBackground: Please, use full abbreviation at first once before you use short term like sexually transmitted infections (STIs), reproductive health (RH) etc. Methods: What is the study period? How many face to face interviews, in-depth interviews and key informant interviews were conducted? What is the main outcome variable?\nIntroduction:\nWhat is the proportion of young key infected populations in Myanmar? If this information is not available, what proportion key infected populations of total population were reported in most recent national census or survey?\nMethods:\nOperational definition: As mentioned earlier, please, define RH seeking behaviour Variables: You have only three outcome variables which are 1) ever receive HIV testing 2) STI treatment 3) Visit to DIC. But you have an objective to determine in seeking health information, therefore, you should have one variable “access to SRH information or seeking health information”. Sampling: What is the reason of choosing purposive sampling? How sample size of qualitative interviews (eg., in-depth interviews and key informant interviews) were determined? As mention earlier, please describe separately and clearly each method of data collection including sampling, sample size, development of questionnaires/ qualitative guidelines, data collection procedure Data analysis: Please, describe details of qualitative data analysis. Ethical consideration: Were verbal or written informed consents taken? What measures were taken to take consents from participants who were aged below 18 years old?\nResults:\nYou just showed reported barriers in seeking STI/HIV information in Figure 2. Before that, did you measure what proportion of participants accessed to STI/HIV information? Table 4: “No or limited time to access health services”- this statement is not clear. Does it mean that YFSW do not have time or have limited time to access health services? If it is so, do you discuss this point how policy makers or health programme can overcome this problem? Table 4: “Reluctance in asking health information”- same comment above that this statement is not clear. Does it mean that YFSW kept reluctance in asking health information to healthcare providers? If it is so, do you discuss this point how policy makers or health programme can overcome this problem?\nDiscussion:\nLast sentence of 2nd paragraph: “In the current study, although the self-reported testing rates were high, we did not ask about the quality of testing services and did not verify these rates by other methods.”- do you please, justify why you did not take measures on quality and verification of self-reported testing services? How this limitation can be overcome in your study? Please, include a concluding paragraph.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4098", "date": "13 Nov 2018", "name": "Spring Research Team", "role": "Author Response", "response": "Comment: Need a clear operational definition of reproductive health (RH) and services? The author only described about STI and HIV. There are other RH problems of FSWs, such as-unintended pregnancies or childbirth, abortion etc. Did the authors exclude these RH problems from their study? Response: Thanks so much for your comments and clarification. You are right that there are many issues under the scope of RH including pregnancy, child birth, abortion, etc. However, in our study, we would like to focus only on STI and HIV since these are most common problems among FSWs community in Myanmar especially in relation to HIV transmission. We had revised our manuscript by adding operational definition of RH and services. We also revised the following facts according to your comment \"Need a clear definition of unmet needs in seeking health information. How did they calculate unmet need of seeking health information?\" Response: Since our main focus was on STI and HIV, participants were requested to respond about their unmet needs in seeking health information regarding STI/HIV. Operational definition of unmet needs in seeking health information was defined in our study as follows. “Although YKAP wants to know or receive STI/HIV information/care, they could not get/receive information/care as they would like to.” For example- though they want to know details about the consequences of anal sex, they do not know how to get or from whom they could get the information. Comment: Need separately discuss about qualitative and quantitative data collection methods in the methodology. Response: According to your suggestion, we revised our manuscript by describing separately about quantitative and qualitative methods. Comment: Need one paragraph conclusion at the end of this script. At current state, it is finished with limitations following recommendations without a conclusion. Response: I think, conclusion is already included together with recommendation. However, we’re revising by adding conclusion statement. “In conclusion, some YKAP have experienced of STI symptom in the past and many of them went to NGO clinic for seeking care. Moreover, many YKAP have tested for HIV within six months. Lesser proportions of Tha-Nge, younger MSM, brothel and entertainment-based YFSW visited DIC than their counterparts. A considerable proportion of YKAP perceived that they have unmet needs in seeking RH information and care.”" } ] }, { "id": "38899", "date": "16 Oct 2018", "name": "Kyu Kyu Than", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAn interesting and important article related to young key population (YKP) in Myanmar. The article mainly focused on the health seeking behaviours of YKPs (Young men having sex with men and Young female sex workers) in relation to STI treatment and HIV testing. It also seeks the utilization of Drop in centres by YKPs. Following clarifications and amendments are need on the manuscript.\nGeneral observation\nAlthough RH services were included in the objectives, the authors mainly focused on the STI symptoms, HIV testing and DIC utilization. The qualitative part of the study was weak in analysis and did not identify the main themes analysed. Was not clear what themes were ask and analysed and what was the added value to the quantitative study. Methodology needs to be more elaborated for the mix method studies using specific guidelines for qualitative and quantitative research methods. Clearly defining the unmet need for health information seeking is required.\nSpecific observation Abstract\nSample size for the qualitative and quantitative study should be moved to the methodology paragraph. Abbreviations should be avoided if possible in the abstract. If need to please specify before using it.\nIntroduction\nAs the focus of the study is youth 15 to 24 years, the first paragraph is a bit confusing shifting from 10-24 and 15-24 in the same sentence.\nMethods\nThere are three main outcome variables that seem to measure the health seeking behaviour. There is a missing variable on obtaining health information. Please clarify. Rationale for use of purposive sampling needs to be specified and more details of the sampling and data collection procedures are required. Recruitment of participants for the qualitative study was not mentioned and the sample size for KIIs and IDIs was not observed. Using note takers in the interviews for populations like MSMs and FSWs,how was the breach of confidentiality ensured. Advantage of using a mix method study would benefit the study methodology stronger. Need to specify the main themes analysed for qualitative thematic analysis.\nResults\nThere is lack of results on health information seeking. Not clear how many percent try to seek health information and what are the main barriers? Table 4 describes the qualitative analysis but the linkage is missing with the quantitative findings. Description of the qualitative findings seems more appropriate if it could be linked to the quantitative data. Very little qualitative information is observed.\nDiscussion\nDiscussions need a conclusion paragraph to draw the main findings and a way forward.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4099", "date": "13 Nov 2018", "name": "Spring Research Team", "role": "Author Response", "response": "General ObservationComment: Although RH services were included in the objectives, the authors mainly focused on the STI symptoms, HIV testing and DIC utilization.Response: Thanks so much for your comments and clarification. You are right that there are many issues under the scope of RH including pregnancy, child birth, abortion, etc. However, in our study, we would like to focus only on STI and HIV since these are most common problems among MSMs and FSWs in Myanmar especially in relation to HIV transmission.Comment: The qualitative part of the study was weak in analysis and did not identify the main themes analysed. Was not clear what themes were ask and analysed and what was the added value to the quantitative study.Response: Thanks for your comments. Our objective for qualitative inquiry is to know their challenges and barriers in seeking health information on HIV/STI. It's aimed to supplement the quantitative information by asking their reasons and perception. Now, we've revised in our manuscript adding the themes we've discussed.Comment: Methodology needs to be more elaborated for the mix method studies using specific guidelines for qualitative and quantitative research methods.Response: We've now added more information in the methodology section.Comment: Clearly defining the unmet need for health information seeking is required.Response: Now, we had added the operational definition in the manuscript.Specific observationAbstractComment: Sample size for the qualitative and quantitative study should be moved to the methodology paragraph.Response: We've already revised in current version.Comment: Abbreviations should be avoided if possible in the abstract. If need to please specify before using it.Response: Already revised.Methods:Comment: There are three main outcome variables that seem to measure the health seeking behaviour. There is a missing variable on obtaining health information. Please clarify.Response: We've added the information in current version.Comment: Rationale for use of purposive sampling needs to be specified and more details of the sampling and data collection procedures are required. Recruitment of participants for the qualitative study was not mentioned and the sample size for KIIs and IDIs was not observed. Using note takers in the interviews for populations like MSMs and FSWs, how was the breach of confidentiality ensured.Response: Regarding rationale of purposive sampling and sample size for KII & IDI were added in current version of the manuscript. For using note takers in the interviews, we tried our best to ensure confidentiality. Our two note takers are well trained and have experienced in dealing with key population. Aim and objectives of the study were also thoroughly explained to YKAP.Comment: Need to specify the main themes analysed for qualitative thematic analysis.Response: we've revised it.Results:Comment: There is lack of results on health information seeking. Not clear how many percent try to seek health information and what are the main barriers? Table 4 describes the qualitative analysis but the linkage is missing with the quantitative findings.Response: In table 2, we've described health seeking for STI, HIV testing and visit to DIC. In this regards, visit to DIC was a proxy measure for health information seeking. Health information seeking was also discussed during the qualitative interview.Discussion:Comment: Discussions need a conclusion paragraph to draw the main findings and a way forward.Response: We've added a conclusion paragraph in current version." } ] } ]	1	https://f1000research.com/articles/7-1515
https://f1000research.com/articles/7-1783/v1	12 Nov 18	{ "type": "Research Article", "title": "Continuous surgical multi-level extrapleural block for video-assisted thoracoscopic surgery: a retrospective study assessing its efficacy as pain relief following lobectomy and wedge resection", "authors": [ "Mark Larsson", "Anders Öwall", "Ulrik Sartipy", "Anders Franco-Cereceda", "Barbro Johansson", "Jan G. Jakobsson", "Anders Öwall", "Ulrik Sartipy", "Anders Franco-Cereceda", "Barbro Johansson", "Jan G. Jakobsson" ], "abstract": "Background: Video-assisted thoracoscopic surgery (VATS) causes less postoperative pain than thoracotomy; however, adequate analgesia remains vital. As part of a multi-modal postoperative analgesia, a continuous surgeon-placed extrapleural block catheter is an option. The aim of this retrospective study was to evaluate the analgesic efficacy of a continuous extrapleural block as part of a multimodal analgesic regimen after VATS in general, and VATS lobectomy and wedge resection in particular. Methods: Case records for patients having undergone VATS surgery and been provided a multi-level continuous extrapleural block with an elastomeric pump infusing levobupivacaine 2.7 mg/ml at a rate of 5 ml/h during 2015 and 2016 were reviewed. Pain (Numeric Rating Scale) at rest and mobilisation as well as opioid requirement (daily, postoperative days 0-3, as well as accumulated) were analysed.\n\nResults: In all, 454 records were reviewed: 150 wedge resections, 264 lobectomies and 40 miscellaneous cases. At rest, pain was mild median NRS rated 3-3-1-1 for postoperative day (POD) 0 to 3, during movement, pain was rated moderate during POD 0 and 1 and mild the remaining days (median NRS 4-4-3-3 for POD 0-3). The proportion of patients exhibiting mild pain at rest increased from 55% on POD 0 to 81 % on POD 3. The percentage of patients experiencing severe pain at rest decreased from 15% to 6%. Median oxycodone consumption was 10 mg per day for POD 1-3. Pain after VATS wedge resection was significantly lower at POD 1 and 3 compared to pain after VATS lobectomy. Conclusion: We found a continuous surgeon-placed extrapleural catheter block to be a valuable and seemingly safe addition to our multimodal procedure specific analgesia after VATS. Whether the efficacy of the block can be improved by increasing local anaesthetic and/or adding adjuncts warrants further investigation.", "keywords": [ "postoperative pain", "VATS surgery", "extrapleural block" ], "content": "Introduction\n\nMinimally invasive thoracic surgery by means of video-assisted thoracoscopic surgery (VATS) has been shown to cause less postoperative pain than thoracotomy1–4. Even though it is not as painful, VATS still requires adequate postoperative analgesia. As part of a multimodal analgesic regimen, different techniques using local anaesthetics can be used to block pain impulses from the surgical area. Epidural, paravertebral and intercostal blocks are all possible modalities. Compared to epidural block, using continuous extrapleural, paravertebral or intercostal block is thought to offer a more limited thoracic block, with possibly fewer side-effects.\n\nFor thoracotomy, several prior studies, including a Cochrane systematic review from 2016, all conclude that epidural and paravertebral block offer similar pain relief, with a paravertebral block possibly having fewer complications5–8. The benefit vs. risk for the paravertebral technique has, however, been argued9.\n\nThe optimal pain management approach following elective VATS is still not known. A review by Steinthorsdottir et al.10 could not provide any firm recommendation. No firm conclusion could be drawn assessing available evidence around thoracic epidural, multilevel and single paravertebral, paravertebral catheter, intercostal catheter, interpleural infusion and long thoracic nerve block. The most recent study comparing epidural and percutaneous paravertebral block by Kosiński et al.11 showed that the paravertebral continuous block technique was a feasible and safe alternative to epidural analgesia. Hutchins et al.12 found ultrasound-guided continuous paravertebral catheter to provide prolonged pain control and superior patient satisfaction compared with single-shot intercostal block after VATS.\n\nIn 2014, continuous extrapleural block after VATS was implemented as a standard technique for pain management following VATS surgery at Karolinska University Hospital.\n\nThe aim of the study was to assess the quality of pain treatment after VATS with the routine use of a multi-level continuous extrapleural block as part of a multimodal analgesic strategy.\n\n\nMethods\n\nThis is a retrospective patient chart study. The study protocol was approved by the regional Human Research Ethics Committee, Stockholm, Sweden (Dnr. 2017/500-31). The need for informed consent was waived by the Ethics Committee. All patients who underwent VATS and received a continuous extrapleural block were included. Records for patients who received an extrapleural block between January 2015 and December 2016 were reviewed. Patient demographics, type of surgery, occurrence of post-operative nausea and vomiting (PONV) or other adverse events, and the postoperative pain course were reviewed and compiled. Data on other analgesics including opioids were also collected. If needed, opioid dose was converted to oral oxycodone equivalents (10 mg oral morphine considered equivalent to 5 mg oral oxycodone).\n\nDuring surgery, a multiply perforated 19-cm extrapleural catheter was inserted by the surgeon, under thoracoscopic control, in a posterolateral position parallel to the spine, thereby covering multiple intercostal spaces. An initial bolus of 75 mg levobupivacaine was followed by an additional bolus of mepivacaine 100 mg at the end of surgery. An elastomeric pump infusing levobupivacaine 2.7 mg/ml at a rate of 5 ml/h (13.5 mg/h) was connected to the catheter and subsequently continued for up to 3 days.\n\nPatients received oral slow-release paracetamol (665 mg, 2 tablets three times daily) and NSAID (naproxen, 250-500 mg twice daily) if tolerated. Patients received oral slow-release oxycodone twice daily with rescue short-acting oxycodone if needed.\n\nPain was assessed by Numeric Rating Scale (NRS)13,14. From the charts, we extracted the NRS score describing the most intensive pain at rest and during ambulation for each day from the day of operation (POD 0) to postoperative day 3 POD3.\n\nFor analysis, surgeries were categorised into two groups. In this report, we considered single and multiple wedge resections procedures causing similar trauma and postoperative pain. Similarly, lobectomy, bi-lobectomy and segment resection were considered as comparable procedures. Consequently, surgical procedures are reported as either a) wedge resection or b) lobectomy.\n\nDemographics are presented as median and range. Pain is presented as median and range and further classified into categories; mild pain as NRS <4, moderate pain as NRS 4-7 and severe pain as NRS >7. For statistical analysis categorized pain was compared as mild versus non-mild (moderate and severe). After calculating the median opioid requirement, we categorized postoperative opioid need as either no opioid if none, low opioid dose if lower than the calculated median opioid requirement or high opioid dose if higher than the calculated median opioid requirement. Categorical data is presented as frequency and percent.\n\nChi-square test was used for test of significant differences between pain categories, age groups, sex and procedure. Mann-Whitney U-test was used for test of significant differences in cumulative opioid dose for different age groups and procedure.\n\n\nResults\n\nA total of 510 patients received an extrapleural catheter during the period studied.\n\nA thoracotomy was performed as the primary procedure in 26 patients. The remaining 484 patients were scheduled for VATS procedures. In 30 cases VATS was converted to thoracotomy. Neither patients that underwent primary thoracotomy nor patients converted to thoracotomy are included in our analysis. Of 454 cases analysed, 150 procedures were classified as wedge resection, 264 cases as lobectomy and the remaining 40 cases as miscellaneous (thymus resection, extirpation of a nodulus, lymph node(s), cyst or hamartoma and decortication with or without wedge resection) (Figure 1).\n\nBaseline characteristics are presented in Table 1. All patients were followed in-hospital from day of operation (POD 0) until POD 3, except for two patients discharged on POD 1, and 37 patients discharged on POD 2, reducing the total number of pain assessments. Missing data occurred where we could not retrieve a pain score or opioid dose, either due to patient discharge or lacking data in the charts. We lack NRS scores with increasing frequency from POD 0 to POD 3.\n\n*Data given as median (range). †Percentage of patients where chest drain was removed by postoperative day (POD) 1.\n\n\nPain at rest and during movement after VATS\n\nAt rest, pain was rated as mild, with median NRS rated 3-3-1-1 for POD 0 to 3. Assessed during movement, pain was rated moderate during POD 0 and 1 and mild the remaining days (median NRS 4-4-3-3 for POD 0 to 3). The proportion of patients exhibiting mild pain at rest increased from 55% on POD 0 to 81% on POD 3. The percentage of patients experiencing severe pain at rest decreased from 15% to 6% (Figure 2).\n\nPain at rest (a) and in movement (b). Percentage of patients with NRS scores available. Absolute numbers are shown in the centre of bars. Patient discharge and missing data cause diminishing absolute numbers during progression of postoperative course.\n\nPain at rest after VATS wedge resection was reported as median NRS 2-2-1-1 compared to pain after lobectomy with median NRS 3-3-2-2 for POD 0 to 3. At rest, more patients were experiencing mild pain and fewer patients were experiencing severe pain after wedge resection than lobectomy all postoperative days except POD 2 where an equal percentage of patients were experiencing mild, moderate and severe pain (Figure 3a, b). The pain pattern, mild pain versus non-mild pain, showed a difference in pain at rest for wedge resection vs. lobectomy at POD 1 (p = 0.03) and POD 3 (p = 0.006), respectively.\n\nPercentage of patients with NRS scores available. Absolute numbers are shown in the centre of bars.\n\nFor POD 0 to 3 median NRS reported in movement was 3-4-3-2 after wedge resection and NRS 5-5-3-3 after lobectomy (p = 0.05 for POD 0).\n\nMedian oxycodone consumption was 10 mg per day for POD 1 to 3 (range: 0 to 210 mg, one patient with preoperative high dose opioid treatment). Median cumulative oxycodone consumption for POD 1 to 3 was 35 mg (range: 0 to 600 mg).\n\nAfter VATS, opioid treatment was not required by 25% of patients at POD 1, 28% at POD 2 and 20% at POD 3. During the first three postoperative days in total, 15% of patients did not require oxycodone at all (Figure 4).\n\nAbsolute number shown inside bars. Categorized data, for definition of low and high opioid dose, see the “Statistics” section of the Methods. Cumulative opioid dose is calculated only where data is complete for all days POD 1 to 3.\n\nThe percentage of patients not requiring opioid POD 1 to 3 was minimally higher after VATS wedge resection. Conversely, the percentage of patients with a high opioid dose was larger for all postoperative days after lobectomy (Figure 5a, b). Patients not needing any opioid during any of POD 1–3 was 14% for both wedge resection and lobectomy. We did not find a significant difference in oxycodone dose after wedge resection and lobectomy for POD 1 to 3 or cumulative for POD 1 to 3.\n\nAbsolute number inside bars. Categorized data, for definition of low and high opioid dose see text under “statistics”. Cumulative opioid dose is calculated only where data is complete for all days POD 1 to 3.\n\nOlder patients (>65 years of age) reported significantly less pain (mild pain vs more than mild pain) at rest (p = 0.01) and in movement (p = 0.002) at the day of surgery. Patients older than 65 had a significantly lower cumulative opioid dose during POD 1 to 3 than younger patients (median 30 vs 45 mg, p < 0.001).\n\nPain at rest after VATS overall was similar for women and men. POD 1, women experienced more often more than mild pain in movement after VATS (p = 0.04) and at rest after wedge resection (p = 0.04). We found no difference between sexes in cumulative oxycodone dose for any procedure.\n\nAdverse effects were rarely documented in the patient records. For 16% of the patients, symptoms of PONV were reported. No major complications related to the procedure (extrapleural catheter) were reported during the 2-year period this audit covers.\n\n\nDiscussion\n\nIn this retrospective study we found that a surgeon placing an extrapleural catheter with continuous infusion of levobupivacaine 2.7 mg/ml at a rate of 5 ml/h was a valuable part of our multimodal pain treatment. Pain was well controlled, pain was at rest overall reported as mild and rated moderate initially and mild from POD 2 during movement. We found VATS lobectomy to be more painful than VATS wedge resection.\n\nTo achieve this level of analgesia, a median dose of only 10 mg oxycodone per day was needed. However, only 15% of our patients did not need any oxycodone at all and 47% needed more than a low dose of oxycodone during the postoperative course. This indicates that there is a potential to further improve the non-opioid part of our multimodal analgesic strategy. We did, however, see a relatively low incidence of opioid-related adverse effects, PONV was experienced in only 16% of patients.\n\nIt should be acknowledged that we used a fixed continuous infusion of 13.5 mg levobupivacaine per hour after an initial bolus of 75 mg regardless of age, body weight and surgery. To our knowledge, the optimal dose for a continuous extrapleural (or paravertebral) block is not known. A higher dose of local anaesthetic, either through higher concentration or higher rate of infusion, might offer superior analgesia. Simultaneously, the risk for adverse events due to toxicity might increase. This dose was chosen taking recommended maximal daily doses into account. Single-shot paravertebral with ropivacaine and adrenaline has been shown to decrease the systemic uptake of ropivacaine15. Addition of adrenaline might allow for more local anaesthetics to be infused. Kosiński et al.11 observed that better analgesia was achieved using a mixture of bupivacaine and adrenaline when infused through a continuous paravertebral block rather than continuous epidural, even implying an additive effect of adrenaline itself. Future trials should address systemic absorption and toxicity for continuous blocks, measuring plasma concentrations of the local anaesthetic used.\n\nAnother possibility to increase the analgesic effect of our extrapleural catheter could be to add adjuncts other than adrenaline to the local anaesthetic infused. In a prospective study, Xu et al.16 showed that adding dexmedetomidine to a single-shot paravertebral block with ropivacaine resulted in better analgesia from 8 to 48 hours after VATS. For patients with unilateral multiple rib fractures, Mohta et al.17 showed equal pain relief with a lower dose of ropivacaine when a continuous paravertebral block contained a low dose of fentanyl in addition to ropivacaine and adrenaline. At the same time, Bauer et al.18 did not show an additional analgesic effect of sufentanil added to ropivacaine in a continuous paravertebral block. Whether a surgical extrapleural continuous block in VATS combining local anaesthetics, adrenaline and a short-acting opioid is superior to a block with only local anaesthetics is, to our knowledge, unknown.\n\nWe used levobupivacaine for our block. It is possible that alternative local anaesthetics offer better analgesia. Lidocaine may be an alternative having both a local block effect as well as systemic anti-inflammatory and possibly analgesic properties19–21.\n\nThis study shows also that different procedures with a difference in trauma applied cause a varying intensity in pain. However, we did not find a significant difference in opioid requirement for different procedures, even though fewer patients needed a high oxycodone dose after VATS wedge resection. Still, tailored analgesia with different types of regional analgesia or different contents infused may be the way to offer patients the optimal balance between analgesia and side-effects, which may improve postoperative recovery.\n\nThere are weaknesses in our study. This is a retrospective patient-record-based study, and all patients operated during the period were included. The wide range of oxycodone dose (0-210 mg/day) may be a result of our retrospective approach. Patients with a preoperative history of chronic pain and/or ongoing opioid treatment were not specifically identified. We used 5-mg increments in oral oxycodone dose. Using a patient-controlled analgesia regimen might have shown a more accurate opioid requirement. It should also be acknowledged that several patients were discharged before POD 3. It is not unlikely that patients discharged earlier and lost to follow-up could have experienced the least pain and least need for opioid treatment. This might skew our results towards higher median pain and higher opioid requirement.\n\n\nConclusion\n\nWe found that a continuous surgeon-placed extrapleural catheter block to be a valuable and seemingly safe addition to our multimodal procedure specific analgesia after VATS. Whether the efficacy of the block can be improved by increasing local anaesthetic and/or adding adjuncts warrants further investigation.\n\n\nData availability\n\nDataset 1. Raw patient data assessed in the present study. Data include pain scores at postoperative days (POD) 0–3 at rest and with movement, alongside basic demographic information. DOI: https://doi.org/10.5256/f1000research.16857.d22436422.", "appendix": "Grant information\n\nU.S. was funded by the Swedish Heart-Lung Foundation [grant numbers 20160522, 20160525] and the Mats Kleberg Foundation. This study was supported by the Heart and Vascular Theme as well as the Function Perioperative Medicine and Intensive Care, Section for Cardiothoracic Anaesthesia and Intensive Care, both at Karolinska University Hospital, Stockholm.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nBendixen M, Jørgensen OD, Kronborg C, et al.: Postoperative pain and quality of life after lobectomy via video-assisted thoracoscopic surgery or anterolateral thoracotomy for early stage lung cancer: a randomised controlled trial. Lancet Oncol. 2016; 17(6): 836–44. PubMed Abstract \| Publisher Full Text\n\nLandreneau RJ, Hazelrigg SR, Mack MJ, et al.: Postoperative pain-related morbidity: video-assisted thoracic surgery versus thoracotomy. Ann Thorac Surg. 1993; 56(6): 1285–9. PubMed Abstract \| Publisher Full Text\n\nWildgaard K, Ringsted TK, Hansen HJ, et al.: Persistent postsurgical pain after video-assisted thoracic surgery--an observational study. Acta Anaesthesiol Scand. 2016; 60(5): 650–8. PubMed Abstract \| Publisher Full Text\n\nTsubokawa N, Harada H, Takenaka C, et al.: Comparison of Postoperative Pain after Different Thoracic Surgery Approaches as Measured by Electrical Stimulation. Thorac Cardiovasc Surg. 2015; 63(6): 519–25. PubMed Abstract \| Publisher Full Text\n\nDavies RG, Myles PS, Graham JM: A comparison of the analgesic efficacy and side-effects of paravertebral vs epidural blockade for thoracotomy--a systematic review and meta-analysis of randomized trials. Br J Anaesth. 2006; 96(4): 418–26. PubMed Abstract \| Publisher Full Text\n\nScarci M, Joshi A, Attia R: In patients undergoing thoracic surgery is paravertebral block as effective as epidural analgesia for pain management? Interact Cardiovasc Thorac Surg. 2010; 10(1): 92–6. PubMed Abstract \| Publisher Full Text\n\nBaidya DK, Khanna P, Maitra S: Analgesic efficacy and safety of thoracic paravertebral and epidural analgesia for thoracic surgery: a systematic review and meta-analysis. Interact Cardiovasc Thorac Surg. 2014; 18(5): 626–35. PubMed Abstract \| Publisher Full Text\n\nYeung JH, Gates S, Naidu BV, et al.: Paravertebral block versus thoracic epidural for patients undergoing thoracotomy. Cochrane Database Syst Rev. 2016; 2: CD009121. PubMed Abstract \| Publisher Full Text\n\nNorum HM, Breivik H: Learning from the past for the present: paravertebral blocks for thoracic surgery are not without risk. Eur J Anaesthesiol. 2011; 28(7): 544–5. PubMed Abstract \| Publisher Full Text\n\nSteinthorsdottir KJ, Wildgaard L, Hansen HJ, et al.: Regional analgesia for video-assisted thoracic surgery: a systematic review. Eur J Cardiothorac Surg. 2014; 45(6): 959–66. PubMed Abstract \| Publisher Full Text\n\nKosiński S, Fryźlewicz E, Wiłkojć M, et al.: Comparison of continuous epidural block and continuous paravertebral block in postoperative analgaesia after video-assisted thoracoscopic surgery lobectomy: a randomised, non-inferiority trial. Anaesthesiol Intensive Ther. 2016; 48(5): 280–7. PubMed Abstract \| Publisher Full Text\n\nHutchins J, Sanchez J, Andrade R, et al.: Ultrasound-Guided Paravertebral Catheter Versus Intercostal Blocks for Postoperative Pain Control in Video-Assisted Thoracoscopic Surgery: A Prospective Randomized Trial. J Cardiothorac Vasc Anesth. 2017; 31(2): 458–63. PubMed Abstract \| Publisher Full Text\n\nBreivik H, Borchgrevink PC, Allen SM, et al.: Assessment of pain. Br J Anaesth. 2008; 101(1): 17–24. PubMed Abstract \| Publisher Full Text\n\nGerbershagen HJ, Rothaug J, Kalkman CJ, et al.: Determination of moderate-to-severe postoperative pain on the numeric rating scale: a cut-off point analysis applying four different methods. Br J Anaesth. 2011; 107(4): 619–26. PubMed Abstract \| Publisher Full Text\n\nKarmakar MK, Ho AM, Law BK, et al.: Arterial and venous pharmacokinetics of ropivacaine with and without epinephrine after thoracic paravertebral block. Anesthesiology. 2005; 103(4): 704–11. PubMed Abstract \| Publisher Full Text\n\nXu J, Yang X, Hu X, et al.: Multilevel Thoracic Paravertebral Block Using Ropivacaine With/Without Dexmedetomidine in Video-Assisted Thoracoscopic Surgery. J Cardiothorac Vasc Anesth. 2018; 32(1): 318–24. PubMed Abstract \| Publisher Full Text\n\nMohta M, Ophrii EL, Sethi AK, et al.: Continuous paravertebral infusion of ropivacaine with or without fentanyl for pain relief in unilateral multiple fractured ribs. Indian J Anaesth. 2013; 57(6): 555–61. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBauer C, Pavlakovic I, Mercier C, et al.: Adding sufentanil to ropivacaine in continuous thoracic paravertebral block fails to improve analgesia after video-assisted thoracic surgery: A randomised controlled trial. Eur J Anaesthesiol. 2018; 35(10): 766–73. PubMed Abstract \| Publisher Full Text\n\nHollmann MW, Durieux ME: Local anesthetics and the inflammatory response: a new therapeutic indication? Anesthesiology. 2000; 93(3): 858–75. PubMed Abstract\n\nMcCarthy GC, Megalla SA, Habib AS: Impact of intravenous lidocaine infusion on postoperative analgesia and recovery from surgery: a systematic review of randomized controlled trials. Drugs. 2010; 70(9): 1149–63. PubMed Abstract \| Publisher Full Text\n\nWeibel S, Jelting Y, Pace NL, et al.: Continuous intravenous perioperative lidocaine infusion for postoperative pain and recovery in adults. Cochrane Database Syst Rev. 2018; 6: CD009642. PubMed Abstract \| Publisher Full Text\n\nJakobsson JG, Larsson M, Sartipy U, et al.: Dataset 1 in: Continuous surgical multi-level extrapleural block for video-assisted thoracoscopic surgery: a retrospective study assessing its efficacy as pain relief following lobectomy and wedge resection. F1000Research. 2018. https://www.doi.org/10.5256/f1000research.16857.d224364" }	[ { "id": "43646", "date": "13 Feb 2019", "name": "Colin F. Royse", "expertise": [ "Reviewer Expertise cardiovascular anaesthesia", "quality of recovery", "echocardiography" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors have conducted a retrospective review of a large sample of patients undergoing VATS with surgeon placed extrapleural catheters. The study is well done and well reported. They found that most patients had mild pain only, and only 15% had severe pain, which diminished to 6% by day 3 – suggesting good analgesic efficacy of the technique.\nI have no major problems with the methods of the paper, and it is well written.\nThey did find a difference between VATS lobectomy compared to other VATS, but the authors have not offered any potential reason for this. It is possible that the surgeons place more intercostal catheters in lobectomy patients, which could increase stimulation beyond the analgesic boundary of the extrapleural catheter?\nFurther in the VATS group, how many patients were included who had a pleurodesis? It is possible that the inflammation produced by pleurodesis could cause stimulation outside of the extrapleural block, and therefore requires additional pain relief.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "45938", "date": "15 Apr 2019", "name": "Waldemar Gozdzik", "expertise": [ "Reviewer Expertise Cardiothoracic anesthesia", "critical care." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this retrospective patient-record-based study, Larsson and colleagues present the analysis of pain treatment quality in 454 patients who received a continuous surgical multi-level extrapleural block for video-assisted thoracoscopic surgery.\nAll patients who underwent VATS and received a continuous extrapleural block were reviewed (150 wedge resections, 264 lobectomies, and 40 miscellaneous cases). The subject taken up by the authors is undoubtedly interesting. This is an important voice in terms of quality and ideal management of postoperative pain after thoracic surgery. The authors have shown the use of an analgesia method with a potentially lower risk of complications than EDA and that the paravertebral block is effective, and satisfactory results could be obtained in this group of patients regardless of the scope of VATS operation.\n\nThe method of analgesia used by the authors is undoubtedly safe and satisfactory, which is confirmed by the results presented, but for obvious reasons, it is not perfect. The described method of anesthesia in VATS operations is, according to the authors' opinion, the standard conduct in their center since the 2014 year.\n\nIn the authors’ study, an initial bolus of 75 mg levobupivacaine was followed by an additional bolus of mepivacaine 100 mg at the end of surgery. An elastomeric pump infusing levobupivacaine 2.7 mg/ml at a rate of 5 ml / h (13.5 mg / h) was connected to the catheter and continued for up to 3 days. This dose was chosen for the maximal daily checks.\nReading the text raises some questions, for example, whether there have been attempts to modify this method of analgesia in their center. Also interesting was the comment on the quality of this method of anesthesia in comparison with the methods used earlier (if it was used, for example, TEDA). Because patients underwent major thoracic surgery, chest tube removal timing seems very short. It would be interesting to know which criteria the authors followed to safely remove the chest tubes, as the authors did not specify them in their otherwise very good paper.\n\nThe authors addressed the limitations of their study well, including a wide range of oxycodone dose, and no PCA method used, lost to follow-up in patients discharged earlier before the third post-operative day. As mentioned by the authors, it is difficult to compare their study to those available in literature because of differences in technique and study design.\n\nIn conclusion: the paper is well written, and the presented data are conclusive. This is an important and valuable work dealing with difficult issues of optimal postoperative analgesia in patients after major thoracic surgery.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1783
https://f1000research.com/articles/5-918/v1	19 May 16	{ "type": "Review", "title": "Hypnotic drug risks of mortality, infection, depression, and cancer: but lack of benefit", "authors": [ "Daniel F. Kripke" ], "abstract": "This is a review of hypnotic drug risks and benefits, reassessing and updating advice presented to the Commissioner of the Food and Drug Administration (United States FDA). Almost every month, new information appears about the risks of hypnotics (sleeping pills). This review includes new information on the growing USA overdose epidemic, eight new epidemiologic studies of hypnotics’ mortality not available for previous compilations, and new emphasis on risks of short-term hypnotic prescription. The most important risks of hypnotics include excess mortality, especially overdose deaths, quiet deaths at night, infections, cancer, depression and suicide, automobile crashes, falls, and other accidents, and hypnotic-withdrawal insomnia. The short-term use of one-two prescriptions is associated with greater risk per dose than long-term use. Hypnotics are usually prescribed without approved indication, most often with specific contraindications, but even when indicated, there is little or no benefit. The recommended doses objectively increase sleep little if at all, daytime performance is often made worse, not better, and the lack of general health benefits is commonly misrepresented in advertising. Treatments such as the cognitive behavioral treatment of insomnia and bright light treatment of circadian rhythm disorders might offer safer and more effective alternative approaches to insomnia.", "keywords": [ "hypnotics and sedatives", "mortality", "cancer", "infection", "depression", "insomnia", "sleep" ], "content": "Introduction\n\nThis is a reassessment of hypnotic drug risks and benefits, updating and expanding information presented October 26, 2015 to the Commissioner of the Food and Drug Administration (United States FDA) as part B of Petition FDA-2015-P-3959, accessible in April, 2016 at http://www.regulations.gov/#!docketDetail;D=FDA-2015-P-3959 along with Comments responding to that Petition. Almost every month, new information about the risks of hypnotics (sleeping pills) appears. This review includes new information on the growing overdose epidemic, eight new epidemiologic studies of hypnotics’ mortality risk not available for previous analyses, and new emphasis on risks of short-term hypnotic usage.\n\n\nRisks of hypnotic drugs\n\nUse of hypnotic drugs is associated prospectively with a greatly increased risk of all-cause mortality. Some of this mortality has been documented as deaths caused by hypnotics by Medical Examiners, attributed to respiratory arrests resulting from “overdose.” However, it is likely that many deaths from respiratory depression occur among patients never seen by coroners, especially when the death is caused by a combination of hypnotics with other contributing factors, so that the lethal hypnotic dosage may by itself have been within currently-customary dosage ranges. In addition to respiratory depression, hypnotics appear to be causally related to serious illnesses and premature deaths from cancer, serious infections, mood disorders, accidental injuries, suicides and homicides.\n\nOnly a small fraction of U.S. deaths has been examined by coroners and Medical Examiners. Thus, it is commonly assumed that overdose deaths have been grossly under-reported. Despite such underestimation, U.S. drug and opioid overdose deaths reported in 2014 reached 47,055, a 137% increase since 20001. Likewise, U.S. use of hypnotics has dramatically increased over about the same interval2. Available death certificate reports seem unclear, but perhaps one third of death certificates listing overdose with an opiate as a cause of death also lists a benzodiazepine, Z hypnotic, or barbiturate as a cause of death (retrieved from CDC Wonder). An opioid prescription is more likely to lead to overdose when a hypnotic is prescribed3. There have also been several thousand yearly reported overdoses involving a hypnotic in which an opiate was not involved. So great is the recent increase in overdose deaths that it has lowered overall life expectancy in much of the U.S. adult population4,5. Overdoses kill more Americans than automobile accidents or murders. Aside from overdose deaths, benzodiazepines were involved in a comparably increasing number of emergency room visits over a similar time interval, and over half of these also involved opioids, alcohol, or all three in combination6. Combined overdoses of opiates and benzodiazepine agonists had more severe outcomes. A recent JAMA Psychiatry Viewpoint gave an even larger estimate of self-injury deaths often due to drug intoxication at 68,298 for 2013, and that did not include the role of suicidally-motivated dangerous driving7. Suicides from all causes per capita have been increasing, particularly among women, and particularly in recent years at about the time that generic zolpidem became available8. Hypnotics are a factor in more than half of intoxication and dangerous driving deaths9.\n\nIn 40 epidemiologic studies that provided comparable risk ratios for mortality associated with hypnotics, 39 showed that hypnotics were associated with excess mortality, as listed in the Appendix. One exception was a small study by Merlo et al. that nevertheless found hypnotics associated with cancer deaths10. A partial exception was a study from Taiwan that found benzodiazepine hypnotics associated with significant excess mortality, but found zolpidem 10 mg associated with significantly reduced mortality in adjusted models, despite a significantly-increased unadjusted mortality risk for zolpidem and a significantly-increased adjusted risk of cancer mortality for zolpidem11. In a comment to this report appearing with it on the internet, I have questioned the statistical methods of adjusting zolpidem risks11.\n\nOnly this one of the 40 epidemiologic studies of hypnotic drugs reported any association with improved patient survival, and that only after questionable statistical adjustments11. None of the other 39 studies found hypnotic drug risk ratios significantly less than 1.0. That is, in 39 studies there was no evidence that hypnotics ever benefit patient survival. To find 39 of 40 studies showing a positive risk ratio was very highly significant, P<0.000001. Also, the evidence of association satisfied all nine Bradford Hill criteria for inferring causality12, though skepticism despite meeting these criteria may be warranted13. There remain questions concerning the magnitude of the causality. The randomized placebo-controlled trials I have suggested would help clarify this causality magnitude12.\n\nOf the 40 epidemiologic studies, 31 individual studies reported statistically significant mortality odds ratios, risk ratios, or hazard ratios exceeding 1.0. All 18 studies reporting on samples of >14,000 people found significant mortality risks, but nine of 22 smaller studies found positive trends that were not significant. Most of the non-significant reports were among the earliest 15 published before 2006. Of studies analyzing follow-ups of 8 years or less, 22 of 26 studies reported a significant association, but of studies with longer follow-ups, only 10 of 14 studies observed significant mortality risks. This may suggest that during long prospective follow-ups, many patients initially taking hypnotics will discontinue hypnotic usage, whereas many controls not using hypnotics at prospective baseline may have begun using hypnotics during a long follow-up, so that the longer the follow-up, the more mixing of hypnotic-consuming and control groups becomes likely. Mixing weakens the risk-ratio contrasts observed. In long follow-ups, one is also studying the selected survivors of the more marked short-term risks that have been recently described14,15.\n\nMost of the 40 studies reported mortality risk ratios of less than 1.5, but some of the highest quality studies reported among the highest risk ratios. Four of the most recent studies were particularly persuasive, as presented below.\n\nFrom electronic records of the Geisinger Health System in Eastern Pennsylvania, a sample of 34,205 patients was drawn with carefully controlled 2:1 matching of hypnotic users with non-user controls for age, gender, smoking, and various comorbidities. Compared to a reference hazard ratio of 1.0 for non-users of hypnotics, the fully-adjusted mortality hazard ratio for use of 0.4–18 hypnotic doses per year was 3.60 (2.92–4.44, 95% CI), for those using 18–132 doses per year, the hazard ratio was 4.43 (3.67–5.36), and for >132 doses per year, the hazard ratio was 5.32 (4.50–6.30)16. Each of these associations was significant with P<0.001. Sensitivity studies showed that little of the hypnotic-associated mortality could be explained by known confounders or use of hypnotics before commencement of the study. In this sample, prescriptions for each of the following drugs were found to significantly predict increased mortality with statistical significance: zolpidem, temazepam, eszopiclone, zaleplon, triazolam, flurazepam or quazepam, and barbiturates prescribed to induce sleep. This review is principally concerned with these popular hypnotics for which drug-specific mortality data are available. Barbiturates prescribed at night for sleep considered as a group had about the same empirical hazard ratios as the benzodiazepines and zolpidem, but the observed hazard ratio for eszopiclone was significantly higher than that of barbiturates, possibly biased by the shorter average follow-up intervals for this more-recently introduced drug16.\n\nIn a sample of over 100,000 hypnotic users and matched controls from the representative British General Practice Research Database17, users of 1–30 defined daily doses (DDD) of hypnotics and anxiolytics within a year had fully adjusted dose-responsive mortality hazard ratios of 2.55 (2.42–2.69, 95% CI) for 1–30 DDD (defined daily doses in the first year); 3.78 (3.54–4.04) for 31–60 DDD, 4.19 (3.84–4.58) for DDD 61–90, and 4.51 (4.22–4.82) for DDD >90. Extensive full adjustment for potential confounders resulted in only very small and inconsistent decreases in the estimated hazard ratios, and many methodological details were focused on minimizing possibilities of confounding. Use of benzodiazepine hypnotics only was associated with higher hazard ratios than use of “Z” hypnotics only. These hazard ratios were remarkably similar to those from the Geisinger Health System, considering the many differences in drug characteristics, samples, design, confounder controls, and analyses. Note that as in the Geisinger Health System study, much of the mortality was associated with early deaths after limited doses of hypnotics, perhaps as little as one-two prescriptions filled or refilled.\n\nA recent representative study of the Norwegian Pharmacy Database found that benzodiazepine-receptor-agonist use was associated with a mortality odds ratio of 2.30 (2.20–2.40)18. The authors argued that terminal illness caused an upturn in benzodiazepine-receptor agonist use shortly before death (which might be appropriate for hospice care), and therefore they argued that the increased benzodiazepine-agonist use among those who would die was demonstrated as a confound of terminal illness. To the contrary, their data demonstrated an excess of benzodiazepine use even among those who would not die until 22 months or later, so the benzodiazepine use of this population was elevated before the terminal upturn in hypnotic usage that the authors had demonstrated. Also, the upturn in death-associated hypnotic use 6–10 months before subsequent death might be consistent with a causal lethal hazard resulting from only a few short months' exposures to hypnotics. The Norwegian Pharmacy Database did not enable this study to identify terminal illnesses, to analyze comorbidities or to control for other confounders.\n\nIn this large study, both French and British case-control samples were drawn from reasonably-representative national samples14. Results had many similarities to those of Weich et al.17 despite numerous differences in statistical design. Substantially lower overall hazard ratios were found in the French sample (not all significant after adjustment), perhaps because a large number of occasional users were included. An important finding was the much higher hazard ratios associated with the initial 3–6 months of hypnotics-benzodiazepines use, as high as 11.12 (95% CI, 9.91–12.47) for the 3-month analysis of the British sample. This sharpened the evidence, also noted in the three previous studies discussed, that although dose-response is observed over several years, much of the hypnotic-associated hazard is observed during the early months of usage after as little as one or two prescriptions.\n\nIt should be noted that Weich et al.17 and Palmaro et al.14 found significant hazard ratios associated with diazepam and other benzodiazepines that are not considered hypnotics (though tranquilizer benzodiazepines may often be used for sleep). These more modern data with better drug identification and measurements of prescriptions during follow-up must be considered more reliable, but neither “Valium” nor “Librium” had been associated with excess mortality in the previous large U.S. CPSII study19. One might argue that if diazepam has a different hazard from temazepam, for example, this specificity tends to bolster the evidence for causality with temazepam. On the other hand, it would not be clear if the specificity is in the drug’s pharmaceutical effects, in its absorption and half-life, in its usual time of administration, in other aspects of frequency and dosage of administration, or in various associated confounders.\n\nNote that these epidemiologic studies had many limitations12. However, the limitations that would tend to bias the results towards underestimating the associations of hypnotics and mortality appeared more influential than those that would bias towards overestimation of the risks. In particular, studies with the most careful efforts to control for confounders found that such control made little difference in the estimated risk ratios, but the hazard ratios in these carefully-controlled studies tended to be higher. The risk ratios derived, like the studies themselves, were extremely heterogeneous, probably due to differences in the size, age, gender, and ethnicity of samples and their health status, the nature of the hypnotics studied, the accuracy with which the drugs involved and their dosages were known, the control variables available, and the duration of follow-up observations to ascertain mortality. Meta-analyses attempting to group 40 such heterogeneous studies would not be clarifying.\n\nData provided by Palmaro et al.14 and Chung et al.15 expanded the hints in the other large epidemiologic studies16–18 that short-term hypnotic usage has surprisingly high risks: apparently short-term hypnotic use has higher risks than long-term usage on a per dose or per-unit-time basis. It is logical that for a patient with an “overdose” of common contributory factors such as aging, obesity, sleep apnea, alcohol overuse, and opiate use, even a single hypnotic dose could be lethal on the first night of consumption. Depending on the drug and the patient’s metabolic capacity, the hypnotic drug concentration in blood could increase for the first few consecutive nights of dosage, but eventually, developing tolerance might make each dose less risky among those who had survived the initial doses. There would continue to be deaths at a lower rate after tolerance develops because of hypnotic dose-escalation in response to tolerance, addition of other sedatives or opiates, especially-heavy pre-sleep alcohol consumption, body position, altitude, upper respiratory infections, and other contributing factors that could suddenly produce hypnotic lethality even after several years of steady consumption. In addition, new consumption of non-sedative drugs that impair liver drug metabolism and even foods such as grapefruit can suddenly make a patient more vulnerable to a customary dose. Understanding these considerations, limitation of hypnotic prescribing to a small number of doses or a single prescription cannot be considered safe.\n\nSeveral reports carefully examined insomnia and depression as potential confounders of the association of hypnotics with mortality, finding that insomnia and depression could explain little if any of this association17,20–22. Note also that the evidence does not permit us to assume that causality between insomnia, depression, and hypnotic usage is one-way when contemplating confounder control, as there is reverse causality23,24.\n\nAltogether, the epidemiologic literature is conclusive that hypnotic use is associated with excess mortality. The better studies tend to show very high dose-response risk ratios suggesting association with a very large number of deaths. A supplement to the Geisinger Health System data showed that the risk ratios demonstrated lead to estimated U.S. deaths associated with hypnotic usage of the same order of magnitude as those associated with cigarette use, around 300,000–500,000 per year16. Evidence has been presented from several independent studies that most of these deaths cannot be attributed to known forms of confounding, and indeed, adjustment for the major confounders such as smoking and comorbidities produced little change in the estimated associations in most of these studies. Authors acknowledge that their estimates of adjusted association of hypnotics and mortality could be influenced by inadequate ascertainment of confounding factors or lack of control for a very large number of potential confounds with small or rare effects. It is because skeptics may question whether the strong associations of hypnotics with mortality are causal, despite data fulfilling the Bradford Hill criteria for inferring causality13, that large post-marketing controlled trials of vulnerable patients may still be needed12.\n\nDespite its well-known risks of lethality, pentobarbital was nevertheless for decades a preferred hypnotic routinely prescribed for patients seeking sleep aids. In the U.S. today, the most notable human application of pentobarbital is in implementing the death sentence. Although it has been believed that the more modern benzodiazepine and benzodiazepine-receptor-agonist hypnotics that replaced barbiturates have higher acute margins of safety and therefore lower risks than pentobarbital, death certificate and epidemiologic data do not confirm that the newer drugs are significantly safer than barbiturates in routine use12,25.\n\nIn the first Cancer Prevention Study, the percentage of deaths at night were found to be increased by 15.6% among those taking hypnotics (P=0.01), presumably due to respiratory suppression26. In that study, the higher percentage of excess deaths at night associated with taking hypnotics accounted for about one third of total excess mortality associated with hypnotics. These nocturnal deaths were attributed to other causes, even though quiet respiratory suppression as a cause would explain the higher percentage of nocturnal deaths observed among those taking hypnotics than among controls.\n\nThe mechanisms of dangerous hypnotic respiratory depression are well-understood. The common hypnotics including barbiturates, benzodiazepines, the “Z” drugs and other benzodiazepine-receptor agonists bind to gamma-aminobutyric acid (GABA) receptors. These ligands-agonists alter the configuration of the receptors to allow negative chloride ions to more readily enter the neurons, where the chloride negatively hyperpolarizes the membranes and inhibits the neurons from firing. When they depress neural respiratory center firing, such drugs can acutely suppress respiration and in large enough dosage, or when individuals are particularly sensitive, may effectively arrest respiration, which leads rapidly to cardiac arrest and consequent death25,27,28. Respiratory depression is accordingly, and accurately, listed among zolpidem’s warnings and precautions29. The barbiturates and alcohol bind to different locations on GABA receptors, where they exert additive or perhaps synergistic respiratory depression effects which may add to benzodiazepine-agonist effects. An antihistamine, diphenhydramine, also binds to GABAA receptors, but it does not seem known whether the actions of diphenhydramine on GABA receptors are similar to benzodiazepines. Opiates bind to mu (μ) opioid receptors on respiratory neurons, where they hyperpolarize neural membranes by opening potassium channels30. Thus opiates, benzodiazepine agonists, and alcohol have additive or synergistic effects inhibiting respiratory neurons.\n\n\nHypnotics can cause serious and potentially lethal infections\n\nA meta-analysis of available placebo-controlled randomized clinical trials showed that hypnotics cause infections (p<0.00001)31. Because these clinical trials randomized hypnotics versus placebos, the 44% higher infection rate among participants who were given hypnotics was proven to be caused by the hypnotics. Moreover, the lead manufacturer of zolpidem has acknowledged that zolpidem induces infections, based on that manufacturer’s own clinical trials data32. The FDA also found dozens of reports of zolpidem-related severe infections among post-marketing reports29.\n\nExtensive epidemiologic data demonstrated that hypnotics are associated with increased pneumonia including fatal pneumonia33. This finding was not confirmed by one Taiwanese study34, but another Taiwanese study focusing on patients with sleep disturbances found that use of zolpidem was associated with 62%–91% increased hospitalizations for serious infections35. A Taiwan study of patients with chronic obstructive pulmonary disease found highly significant odds ratios associated with benzodiazepine use of 9.3 for pneumonia, 10.4 for acute chronic obstructive pulmonary disease (COPD) exacerbation, 45.0 for acute respiratory failure, and 18.6 for cardiopulmonary arrest; whereas the odds ratios for \"Z\" drugs such as zolpidem were of almost similar magnitude3. In confirmation, note in the Geisinger Health Study supplement, Table 716, mortality hazard ratios were likewise specifically elevated among hypnotics users with COPD. Another Taiwanese study observed that use of zolpidem was associated with increased risk of pyogenic liver abscess36. British data showed that use of benzodiazepines and use of the hypnotic zopiclone (containing 50% eszopiclone as the active ingredient) were significantly related to asthma exacerbation and to all-cause mortality following exacerbation37. This asthma study described some of the benzodiazepine-agonist-mediated impairments of immune surveillance37. Perhaps as a consequence of post-hospital continuation of benzodiazepines and resultant infection, use of benzodiazepines was associated with 23% increased hospital readmission in North Carolina38. In summary, epidemiologic evidence indicates that hypnotics not only cause the mild upper-respiratory infections most commonly reported in available controlled clinical trials31, but also more severe and life-threatening infections. Since such infections demonstrably impair survival, infection is shown to be an additional mechanism by which hypnotics covertly increase mortality. The death certificate would be likely to list the infection as a cause of death but not the hypnotic which may have caused that infection.\n\nAnimal studies confirm that hypnotics can cause infections. A controlled trial demonstrated in mice that diazepam exacerbated Streptococcus pneumoniae infection through GABAA receptors, partly explaining the underlying immune mechanisms39. In mice, diazepam also exacerbated cowpox, a viral infection40. Midazolam impaired equine immune responses, attributable to effects on macrophage peripheral benzodiazepine receptors (now called TSPO)41. Evidence for involvement of the peripheral benzodiazepine receptor TSPO in immune impairment also came from specific test compounds in mice42. Thus, hypnotic drugs cause increased risk of potentially lethal infections in controlled laboratory experiments.\n\n\nHypnotics are associated with increased cancer\n\nA compilation of randomized controlled trials of hypnotics showed 12 cancers or tumors of uncertain malignancy reported among participants randomized to a hypnotic, but none (zero) among those randomized to placebo (P=0.032, two-tailed Fisher Exact Test)43. When the FDA repeated this audit of their controlled trials data, they counted 13 cancers among those randomized to hypnotics versus none (zero) from placebo43.\n\nThe controlled-trials compilations described above did not include indiplon, an unlicensed zaleplon-like benzodiazepine agonist and hypnotic, for which studies published subsequently indicated three incident cancers in the indiplon groups and none in the randomized control groups44,45. The compilations did include cancers associated with the marketed hypnotic ramelteon that admittedly has a very different molecular mode of action from the benzodiazepine agonists.\n\nThe FDA was not persuaded that these human controlled-trials data required regulatory action, because most of the definite cancers were only minor skin cancers, because of heterogeneities in the data, and because the cancers were recognized after such short randomization periods. Nevertheless, the controlled trials data suggested more than skin cancer. There were cancers of organs apart from skin noted among those treated with hypnotics but none among those randomized to placebo. Reconsideration of FDA's deferral of action is now encouraged by new animal testing and new epidemiologic findings: over half of the research referenced in this manuscript appeared after that FDA deferral of action.\n\nBecause hypnotics seem to cause cancers to be suddenly recognized during short clinical trials, e.g., from one month to one year, the short-term effects are likely to arise more from hypnotics promoting progression of tiny pre-existing cancers rather than from effects upon microscopic cancer initiation. Such progression may cause a cancer death, whether or not the hypnotics initiated the cancer.\n\nThe animal data in the FDA files for zolpidem indicated that increasing doses of zolpidem fed to rats resulted in increasing numbers of renal liposarcomas and lipomas combined (statistically significant). These data also showed increased thyroid follicular adenomas and carcinomas combined, and increased testicular interstitial cell adenomas, but the latter findings did not reach statistical significance46. There were no such tumors – that is, zero tumors – in the placebo groups. These studies were too small, however, to have substantial power for these neoplasms. Expert FDA pharmacy examiners interpreted the data as suggesting an unknown degree of cancer risk for humans.\n\nThese experiments, which showed tumors resulting from feeding zolpidem to rats and suggested a dose-dependent relationship, apparently were never extended, clarified, published, or otherwise followed up.\n\nSimilarly, the animal data used for eszopiclone evaluation relied largely on old zopiclone data, since eszopiclone is roughly 50% of zopiclone, and eszopiclone is thought to be the active isomer. Along with other issues, the animal evidence that zopiclone caused animal cancers was of great enough concern to FDA’s scientists, that at least five FDA scientists and medical officers recommended against approval of eszopiclone47. Tumors of the lung in rodents were of special concern; these findings also anticipated the human-specific association of hypnotics with lung and esophageal cancers, as will be described below. Eszopiclone was nevertheless approved as a hypnotic.\n\nSince zolpidem and eszopiclone were evaluated, much additional evidence has appeared relating hypnotics to cancer. Amerio et al. systematically surveyed FDA records including much animal data not included in the earlier compilation of hypnotics trials and concluded that hypnotics and sedatives had among the most elevated cancer hazards among psychotropic drugs48.\n\nZopiclone, zaleplon, and ramelteon are clastogenic47,49,50. These hypnotics damage chromosomes. Clastogens are potentially mutagenic agents that induce disruption or breakages of chromosomes. This process can lead to carcinogenesis. Cells that are not killed by the clastogenic effect may become transformed to cancer51. One of the several formulations of zolpidem was said from in vitro studies not to be clastogenic52. Other than the four drugs mentioned, no information has been located that other hypnotic drugs found to be associated with cancer have ever been adequately tested for clastogenicity.\n\nClastogenicity is only one mechanism by which hypnotics are likely to be carcinogenic, through either initiating cancers or promoting progression through additional mutations of cancer cells, or both. The alterations of immune surveillance produced by benzodiazepine agonists, discussed in relation to infection above, suggest additional mechanisms by which cancer initiation and progression might be facilitated or disinhibited53. Hypnotic-initiated increases in infections and consequent inflammation is another potential carcinogenic mechanism. These animal-demonstrated and in-vitro mechanisms for carcinogenicity of hypnotics, that have been widely ignored, support evidence that hypnotics cause human cancer.\n\nA 2008 paper43 listed three prior epidemiologic studies reporting associations of hypnotics with cancer deaths10,54,55. Analysis of CPSII data found that the elevation in deaths associated with hypnotics was comparable to that associated with cigarettes55, though not entirely due to cancer. The report of Merlo et al. was unique in that the small study showed significantly increased cancer deaths among hypnotic users, but the increase in overall deaths associated with hypnotics was not significant10. Mallon, Broman, and Hetta found a much higher cancer adjusted hazard ratio for habitual sleeping pill use of 5.3 (95% C.I. 1.8–15.4) than for smoking among males; none of the specific causes of death were individually significant among females54. A similar result was shown in a later paper for males, but the simple significant mortality elevation of regular hypnotic use among females was lost after multivariate adjustment in the second study21. More recently, a number of new studies have appeared reporting that hypnotic usage is related to cancer incidence and mortality. Hartz and Ross found a significant association of hypnotic use with melanoma and close-to-significant associations for lung and breast cancers56. Kao et al. found a remarkable 6.24 (4.13–9.43, 95% CI) hazard ratio for cancer incidence among those using at least 300 mg of zolpidem per year without other-benzodiazepine consumption (this would correspond to slightly more than one 5 mg dose per week)57. In this Taiwanese national study, smoking and body mass index (BMI) were not controlled, but the overall cancer hazard ratios for zolpidem users were almost identical among men and women, despite an almost 11-fold greater prevalence of smoking among adult men compared to Taiwanese women at the time58. BMI was not controlled, but at that time in Taiwan, although being overweight was more common among women, obesity was more common among men59. In a complementary study of benzodiazepines in Taiwan, benzodiazepines were associated with a 1.19 (1.08–1.32 95% CI) cancer incidence hazard ratio, with over twice the benzodiazepine-associated hazard among men as among women60. Similarly, a brief analysis of the national data from Taiwan found a significant cancer adjusted odds ratio for two of three benzodiazepine hypnotics61.\n\nIn the Geisinger Health study using electronic medical records, Kripke et al. found a hazard ratio for cancer incidence of 1.35 (1.18–1.55 95% CI) associated with use of >132 hypnotic doses per year, with specific hazard ratios of 1.28 (1.03–1.59) for high-dose zolpidem and 1.99 (1.57–2.52) for high-dose temazepam16. There was a significant dose-response. This study was carefully controlled for age, gender, smoking, BMI, and by matching comorbidities among cases and controls. Jiao et al. found no excess of colorectal cancer among those reporting sleeping pill usage <3 times per week versus ≥3 times per week in the Women’s Health Initiative data set62, a result consistent with the Hartz and Ross report on the same data set56, but since the contrast of frequencies of usage was weak and the type and quantity of hypnotic consumption were not determined objectively, the negative observation was not very persuasive. We would not expect hypnotics to promote all cancers equally. Indeed, selective specificity among cancer types would be anticipated if the mechanisms are causal. Pottegard et al. and Sivertsen et al. found small but significant associations of hypnotic usage with cancer, especially lung cancer63,64, but since they had not controlled for cigarette smoking, both groups thought their result might have arisen from confounding, albeit confounding was not conclusively demonstrated65. That investigators failed to control for important confounders is not proof that confounding explains the significant hazard. Several U.S. and European groups63,64 and also Kao et al.57 found high hazard ratios for lung and esophageal tumors, and the two San Diego studies had carefully controlled for smoking16,55. We had proposed that effects of hypnotics on weakening the gastro-esophageal sphincter and permitting more gastro-esophageal regurgitation66 might account for the high cancer-specific rates of esophageal and lung tumors16. These multiple studies finding hypnotics associated with human lung cancer were consistent with concerns of FDA scientists about lung cancers found in animal studies of zopiclone. The lung cancer specificity supports causality.\n\nThere was one pair of studies that was neither clearly confirmatory nor negative. A large-scale survey screening many drugs with a questionable scheme for reusing controls for multiple tests and incorporating a questionable 2-year drug-to-cancer lag remarked no significant association of cancer with temazepam or zolpidem but did find significant associations with oxazepam and perhaps lorazepam, using P<0.01 and relative risk >1.50 as criteria67. In that study, it was not always possible to control for smoking, and control for other confounders was crude and not well-standardized. A similar study added a possible association for phenobarbital68.\n\nTo summarize the cancer epidemiology, the great majority of relevant studies have noted either small or large hazards for new cancers associated with hypnotics. Epidemiologic studies of hypnotics and cancer incidence have had many limitations, and some of these studies were not well-controlled for important confounders such as smoking and obesity, but the preponderance of evidence suggested a causal elevation of cancer incidence among those who take hypnotics.\n\nThe available clastogenicity data, animal data, randomized placebo-controlled clinical trials, and human epidemiology studies consistently, if not always conclusively, suggested that hypnotics likely cause human cancers and cancer deaths.\n\n\nHypnotics increase incidence of clinical depression\n\nIn combined clinical trials, participants randomized to hypnotics suffered 2.1 times as many incident (new) depressions as those randomized to placebo (P<0.002)23. These were not exacerbations of pre-existing depressions. These were depressions caused by the hypnotics. There are other data demonstrating worsening of depression with a wider variety of popular benzodiazepine and GABA agonists9. Treatment of insomnia by hypnotics causing comorbid depression stands in marked contrast to cognitive-behavioral treatment of insomnia, that has been shown to decrease comorbid depression69.\n\nSome studies have appeared designed to show that a hypnotic reduced depression scores among patients given an antidepressant known to cause insomnia70,71. In the first of these studies, the benefit of the hypnotic for depression was not significant at week 4 after the investigators removed the rating scale items related to insomnia, whereas the week 8 benefit was only significant at the P=0.04 level not correcting for multiple comparisons. In other words, using rigorous Bonferroni correction for multiple comparisons, the alleged benefit of hypnotic for depression symptoms was not significant. In the second study the authors more readily conceded that the hypnotic had no significant benefit for depression. These studies failed to rebut the evidence that hypnotics cause new depressions.\n\nDepression is the major cause of suicide. Panic attacks are another risk factor for suicide9. Short-acting benzodiazepine agonists such as triazolam and zolpidem may cause withdrawal anxiety and even panic attacks during the daytime72. Suicide has been recently described as the 8th or 10th leading cause of death in the United States7,73. Hypnotic use is associated with high rates of suicide25,55,74. Indeed, comprehensive toxicological studies have found intoxicating abusable substances (mainly sedative-hypnotics) in a majority of suicides, often combined with alcohol in 30–40%9. Suicides due to overdoses have increased dramatically from 1999 to 2010 in the U.S.75, but there have been an even larger number of deaths of undetermined manner in which suicide through overdose must be suspected76. A very recent report estimated that in 2013 there were 7,000 overdose deaths related to anxiety and sleep medications73, but this did not include all suicides in which the most rigorous toxicology shows a sedative or anxiolytic often mixed with alcohol to be present9. The adjusted odds rate for suicide was 4.2 among hypnotic users as compared to nonusers in one study of elderly people, whereas the odds were not elevated among anti-depressant users (tending to exclude depression and other comorbidities as confounders77.) Prescription sleeping pill use was a stronger predictor of suicide attempts than insomnia symptoms in the National Comorbidity Survey Replication78. In a large study from Taiwan, the adjusted suicide hazard ratio for “needing sleeping pills” was 11.1, whereas the hazard ratio for those reporting sleeping only 0–4 hours adjusted for sleeping pill use was only 3.5, and none of the hazard ratios for insomnia symptoms exceeded 2.079. Another national Taiwan study found increased suicides and attempts associated with zolpidem74. The findings indicate that the association of suicides with hypnotic use cannot be entirely attributed to confounders with reverse causality, since the association of hypnotic usage with depression is known to be largely caused by the hypnotics23.\n\nZolpidem specifically has been implicated as a causal agent in a number of suicides, some of which involved kinds of dissociative behavior often attributed to zolpidem or to combined use of zolpidem with other drugs or alcohol80. Impairments of cognition and judgment that may be caused by sleeping pills81 as well as hallucinations82, irrational behaviors75,83–85, and behavioral disinhibition9 may all contribute to suicides, violence, and accidents, even among people who are not severely depressed.\n\n\nAutomobile crashes, falls, and other accidents are associated with hypnotics\n\nHypnotic drugs impair next-day alertness, motor skills, reasoning, and overall performance. Accidents of all sorts are associated with use of benzodiazepines and benzodiazepine agonists such as zolpidem86–88. Most hypnotics impair automobile driving, as indicated by on-the-road controlled performance testing89. This impairment in some instances exceeds the impairment produced by a blood alcohol concentration of 0.05%90. Drivers’ ability to predict their own impairment is poor91. The use of hypnotics and other sedatives is strongly associated with on-the-road driver-at-fault crashes92–96. In addition to accidents attributable to impaired coordination, impaired motor skills and loss of alertness, hypnotics may also lead to fatal crashes due to drug induced suicidal thinking, impaired judgment, or recklessness on the part of intoxicated drivers7. Some of these crashes result in deaths of passengers and other-vehicle occupants not themselves using hypnotics, but the other-vehicle deaths are not attributed to the hypnotics on death certificates.\n\nOne study showed that use of benzodiazepines and \"Z\" hypnotics was increased among victims of homicide as well as among the homicide perpetrators97. Thus, both through bad driving and homicides, hypnotics result in deaths that have not been accounted directly as deaths from these hypnotic drugs.\n\nIt is well known that falls and accidental injuries are strongly associated with hypnotic usage, in particular hip fractures among aging patients98–106. Hip fracture is a sometimes-lethal injury. The preponderance of studies indicates a true association of the use of hypnotics and falls, that is thought to be due to the properties of benzodiazepine agonists in inhibiting psychomotor skills and in causing weakness, slowed reflexes, and impaired judgment, especially less than 8 hours after ingestion. After taking a hypnotic at bedtime, older people may get up during the night, e.g., to visit the bathroom, when the pharmacologic impairment from a hypnotic is near-maximum and is combined with impairments from sleepiness.\n\nA nursing-home study challenged these conclusions, arguing that it was insomnia, not hypnotics, that was associated with falls. This study did not appear to control for confounding sleep apnea, Alzheimer’s disease, or cognitive-behavioral disorders107. It should be conceded that confounders are likely have some influence on risk ratios associating hypnotics with accidental injuries, but the scientific consensus suggests that the association is nevertheless partly causal, based in part on controlled trials showing hypnotic impairments of driving and other forms of psychomotor performance. A causal element is inferred by the majority of authorities.\n\n\nSafe doses of hypnotics for target populations are unknown\n\nAnimal studies indicate that some individuals in an animal research sample may succumb to a lethal hypnotic-drug effect at doses as low as one-fifth that which is universally lethal108. Variations in susceptibility in a human population varying in age, gender, genetics, and health status is likely to be greater than that in a sample of laboratory animals. The minimum lethal dose of hypnotic drugs in humans is unknown, that is, the dose that might produce fatal respiratory arrest in one person out of 1000 in a representative population or one in 10,000. So many billions of hypnotic doses are prescribed yearly in the U.S. that one death per 10,000 doses would yield over 100,000 deaths per year. Moreover, there are no human dose-response data and very little animal data concerning what doses of hypnotics may be lethal in the presence of opiates, other sedatives, alcohol, aging, obesity, COPD, and other comorbidities. Yet most recognized hypnotic-related deaths are observed in the presence of such additional factors. More study is needed to establish safe doses of hypnotics (if any) when taken with other medications and in the presence of potential comorbidities. As for aging, the consensus of the American Geriatrics Society is that hypnotics should not be given to elderly patients in any dose109.\n\n\nContributory factors combined with hypnotics could cause covert deaths\n\nThere is a vast discrepancy between the hundreds of thousands of yearly hypnotic-associated deaths implied by the high epidemiologic hazard ratios and the mere thousands of yearly death certificates in which a hypnotic is listed among the causes of death. Below are presented some of the possible explanations for this discrepancy.\n\nObesity and aging are perhaps the two most important risk factors for sleep apneas, that is, brief cessations of breathing during sleep110. Sleep apneas occurs at least a few times per hour in the majority of adults over age 40 years and in a great majority of those over age 65110,111. If the duration of a sleep apnea before arousal becomes excessively prolonged, e.g., by a hypnotic, death could result. Even apart from apneas caused by obesity, mechanical problems produced by obesity may impair ventilation. Thus, hypnotic-related hazard ratios are higher among obese patients (see Geisinger Health study supplement Tables 2 and 716.) Since there is no evidence that the huge increase in hypnotic hazards among obese patients can be attributed overdoses alone, it appears that obesity predisposes to covert hypnotic-related deaths, probably by prolonging apneas. The contributing role of combining hypnotics with opiates, other sedatives, and alcohol has already been discussed and is further discussed below. It is plausible that among susceptible patients, combinations of aging, obesity, sleep apnea, hypnotics, opiates, other sedatives, and alcohol could produce quiet respiratory cessations followed by cardiac cessation and death even without any ingested doses above common medical practice being taken. In effect, the “overdose” comes from the overdose of contributory factors along with a hypnotic, not from an excessive dosage of a hypnotic considered by itself. Recall that the American Geriatrics Society does not approve of administering hypnotics to elderly patients in any dose109, though a substantial portion of all hypnotic prescriptions go to elderly patients.\n\nThe use of opiates has become increasingly common in recent years112. Opiates are respiratory suppressants that (like pentobarbital) in overdose can produce respiratory arrest and cardiac arrest. Among patients taking both benzodiazepines and opiates, a remarkable 75% were found to have sleep apnea, and causality was suggested by significant dose-response correlations both for the opiates and for the benzodiazepines113. In some patients, this combination of benzodiazepine and opiate causes hypoxemia (low oxygen)114. Our sleep clinic has recorded polysomnographic data from patients who suffered profound almost continuous apnea with severe hypoxemia due to combinations of hypnotics and opiates. Recall that it is understood on a molecular level how benzodiazepine agonists and opiates combine to suppress firing of respiratory neurons that are necessary to breathe. Patients receiving a combination of benzodiazepines and opiates have increased mortality115–117. The combination of opiates and benzodiazepines has caused a growing overdose problem in emergency rooms112. Moreover, the most serious overdose problems are seen when opiates and benzodiazepines are combined with alcohol in older patients, reflecting combined effects of opiate, benzodiazepine, alcohol, and aging6,117.\n\nIt may be relevant that close to 70% of hospice patients were taking an opiate and an anxiolytic or hypnotic in the last week of life118. This is not evidence by itself whether this combination influences the survival of hospice patients, nor is the author commenting on the ethics of combining such drugs in a genuine hospice situation. However, most patients given hypnotics and opiates have not consented to hospice management.\n\nIn many combined-sedative deaths, the individual drug concentrations present in blood may appear within customary therapeutic ranges. Even if a patient is undergoing cardio-respiratory monitoring at the time when respiratory cessation followed by cardiac cessation occurs, there is usually no way of determining whether the fatal respiratory cessation was due to hypnotic drugs in combination with various contributory factors. Especially when death occurs quietly at night (for example, death of an elderly obese patient known to have various comorbidities,) there usually is no autopsy. Physicians signing the death certificates may be tempted to list a cardiac event or a stroke or some long-standing comorbidity as the cause of death without recognizing when hypnotic-induced respiratory suppression was the precipitant.\n\nThe press described a highly distinguished example of how the cause-of-death is often declared after U.S. Supreme Court Justice Scalia died unexpectedly at night. According to numerous news reports and sheriff’s documents, Justice Scalia’s appearance was that of a person who had peacefully stopped breathing at night. There was no sign of agitation due to cardiac pain, nor had Justice Scalia complained of cardiac symptoms before going to bed. Justice Scalia might have been taking hypnotics and opiates for the jet lag and pain he was known to be suffering when he arrived at a hunting lodge that routinely gives each guest a free bottle of wine. By the bedside, his CPAP machine for control of sleep apnea was not connected, perhaps because he was too excessively sedated to connect his CPAP device or because of problems related to the 4400 feet altitude. Without ever viewing the deceased or his bedroom, much less determining what hypnotics, opiates for pain, and alcohol Justice Scalia had consumed, a local official was encouraged by Justice Scalia's physician (thousands of miles away) to declare heart attack as the cause of death. Without an autopsy, we will never know if this death was precipitated by hypnotics or opiates and alcohol or if there was a heart attack. Even if a physician suspects that a hypnotic had a role, the physician has little motivation to list the hypnotic as a cause of death when it would be hard to prove and may reflect negatively on the physician prescribing that hypnotic.\n\nAlong the same lines, when hypnotics cause infection, cancer, depression, falls or other accidents, or murder, hypnotics are rarely listed among the causes of death. These patterns along with quiet respiratory deaths may explain why epidemiology shows much higher risks of death associated with hypnotics than the death certificates document. Nevertheless even the numbers documented in death certificates are much too high to be acceptable.\n\nZolpidem, reportedly the most commonly-prescribed hypnotic in the U.S., with an estimated 40 million outpatient prescriptions in 2013119, ranked first for emergency department visits among psychotropic drugs according to CDC data119,120. According to the Agency for Healthcare Research and Quality (AHRQ) data, 68% of zolpidem patients were sustained users (three or more prescriptions), and of those 22% were also sustained users of opioids119. Note that recent CDC guidelines recommend against use of benzodiazepine agonists with opiates121. Although the FDA had recommended that women use only 5 mg or 6.25 mg of zolpidem, only 5% of women and 10% of elderly were prescribed these low doses119. Moreover, 23% of patients with sustained use took another drug targeting the same receptors. A high percentage were depressed, as indicated by 34% of sustained zolpidem users also receiving antidepressants119. Recall also that the American Geriatrics Society recommended avoidance of any use of hypnotics for elderly patients109, though many receiving hypnotics are elderly\n\n\nHypnotics cause withdrawal insomnia, anxiety, panic, and epilepsy\n\nIt has been well known since they came into use over a century ago that hypnotics and similar sedatives are addicting drugs, frequently eliciting tolerance, physical dependence, and withdrawal reactions. Most of the benzodiazepine agonist hypnotics and even suvorexant are controlled like addicting drugs by the U.S. Drug Enforcement Agency (DEA). Withdrawal from benzodiazepine agonists can cause insomnia, anxiety, agitation, confusion, and panic and even more severe somatic symptoms such as seizures and death in extreme cases24,72,121,122. In addition, some of the short-acting sedatives such as triazolam and zolpidem may sometimes cause anxiety or agitation during the day following administration before the previous bedtime. Dr. Kripke has seen two patients taking triazolam who developed daytime panic attacks that remitted upon triazolam withdrawal and recurred upon re-challenge. There is also evidence that prolonged use of hypnotics may lead to lasting insomnia, as a consequence causing patients who withdrew from hypnotics to sleep worse than patients who had been randomized in parallel clinical trials to placeboes24. How long this withdrawal insomnia might persist has never been adequately defined.\n\nIn another example of sedative withdrawal leading to hyperexcitability, there is a report that benzodiazepine use and withdrawal may result in lasting increased epilepsy123.\n\n\nRelationship of hypnotics to insomnia, long sleep, and short sleep\n\nA pioneering large epidemiological study by the American Cancer Society study conducted over 50 years ago showed an increased risk of death following hypnotic use. The Cancer Prevention Study I (CPSI) obtained questionnaires in 1958 from over 1,000,000 participants and reliably followed up their death or survival for 6 years124. The data showed that both long and short sleep predicted elevated mortality (with 7 hours associated with minimal mortality for each age group). This study (often replicated) raised scientific doubt whether there is medical value to increasing reported sleep duration of an adult beyond 7 hours, though it also demonstrated that many adults reporting more than 7 hours of sleep were taking sleeping pills. Sleep durations below the population median are partly attributable to inherited traits, so whether there would be any health benefit in sedating people with short sleep durations to sleep longer remains to be demonstrated. A small objective study of sleep duration recorded by wrist activity suggested increased mortality above 390 minutes of actual sleep (which is greater than the current median sleep of American adults studied with similar technology125.) In the CPSI data, self-reported insomnia had little or no additional mortality effect beyond hours of sleep, although insomnia was moderately associated with short sleep. In contrast, reported sleeping pill use was associated with about 50% increased mortality after controlling for age, gender, reported sleep duration, and reported insomnia126. This was statistically a highly significant result in a million participants, but uncertainty about what participants meant by taking “sleeping pills” “Often” in terms of drug type and frequency demanded more study. The American Cancer Society performed a second Cancer Prevention Study (CPSII) with participants completing over 1.1 million questionnaires in the fall of 1982. CPSII used more explicit questions about sleep duration, insomnia, and “prescription sleeping pills.” After controlling simultaneously for 32 covariates and confounders such as insomnia and sleep duration in Cox Proportional Hazards models, results again showed that use of hypnotics was associated with elevated mortality not attributable to major confounders such as cigarette smoking. Indeed, the mortality risk associated with taking “prescription sleeping pills” was surprisingly comparable to that associated with smoking a pack of cigarettes a day55.\n\n\nBenefits of hypnotics: minimal\n\nIn an authoritative National Institutes of Health (NIH)-sponsored meta-analysis of controlled trials including unpublished trials127, Buscemi and colleagues found that although non-benzodiazepine zolpidem-like drugs [“Z-drugs”] shortened sleep onset latency by an average of 12 minutes (9–17 min, 95% CI), according to objective polysomnograms, these hypnotics increased total sleep time by only 11 minutes (-1 to 23 min, 95% CI, NS). That is, these “Z” drugs produced no substantial statistically-reliable increase in total sleep, even at doses higher than currently recommended. Most of the meta-analyzed studies of zolpidem used doses of 10 mg or more (as high as 30 mg)127, and most of the studies of zopiclone used 7.5 mg doses or more (containing more eszopiclone than any dose approved in the U.S.). The FDA-approved recommended initial zolpidem dosage for most patients is now 5 mg (6.25 mg for the sustained-release form128.) Zolpidem and zolpidem-like drugs constitute the bulk of the current U.S. hypnotics market. Based on all available clinical studies, these lower doses would objectively increase sleep little if at all119. Indeed, the primary zolpidem manufacturer advised the FDA that the 5–6.25 mg dosages were generally ineffective32. The newly-recommended 1-mg dosage of eszopiclone is similarly ineffective129,130. Patients typically report more increase in sleep than is measured objectively, but even this self-reported “improvement” at above-recommended doses (which is not supported by objective measurement) is a mere 32 minutes. (26–38 minutes, 95% CI)127. The discrepancies between objective and patient-subjective data may be attributable to the amnesic properties of hypnotics, erasing patients’ memories of how much time they are awake in bed. In conclusion, the FDA-recommended doses of the most popular benzodiazepine agonists are virtually ineffective for objectively increasing sleep. Older benzodiazepines are not much more effective, if at all.\n\nA new Comparative Effectiveness Review sponsored by the U.S. AHRQ has recently examined the Management of Insomnia Disorder, largely referring to chronic insomnia131. As a prepublication Peer Reviewer of this report, I was and still remain very critical of its limitation to mainly-subjective data that are known to give a rosier evaluation of hypnotic effects than objective evaluations, its focus on published reports that are known to be commonly biased towards reporting favorable drug results132,133, and the AHRQ report's incomplete attention to adverse effects. Nevertheless, it is striking that the AHRQ study found that the strongest evidence for treatment efficacy was with the cognitive-behavioral treatment of insomnia. The evidence for short-term efficacy of zolpidem and eszopiclone in high doses was considered less sufficient, and evidence for efficacy of other hypnotics was judged to be almost entirely insufficient. Moreover, by its clinical trial selection criteria, this Review found essentially no evidence for efficacy of the very low doses of zolpidem and eszopiclone currently recommended by the FDA for most patients, because higher doses appeared unsafe to FDA. In short, the AHRQ study presented no reason why hypnotics are needed, since cognitive-behavioral treatment of insomnia is better and is available even on the internet.\n\nMoreover, the AHRQ Review found evidence for increased adverse effects with hypnotics compared to placebo, including hypnotic adverse effects of concern (their selection of studies highlighted fractures and dementia.) The Review found mention of adverse effects virtually absent for the cognitive-behavioral treatment studies131. Although the Comparative Effectiveness Review found insufficient studies to estimate the comparative effectiveness of hypnotics versus cognitive-behavioral treatments, when it reviewed potential harms, there was no contest. Moreover, controlled trials reviewed above prove that hypnotics cause comorbidities such as infection and depression and driving impairments, whereas cognitive-behavioral treatment has been found to decrease medical comorbidities such as depression69.\n\nBased on manufacturers’ advertising, patients expect that a hypnotic will improve their function and performance the following day. In fact, the truth is just the opposite. In 1982, two sleep experts received partial support from a hypnotics’ manufacturer to survey daytime performance literature about hypnotics and found, “Drug-related improvement in performance was not found, and, in comparing active drug to placebo, it is clear that all hypnotics, at some doses, produce decrements in performance the next day81.” Since 1982, the author has been looking for objective evidence that hypnotics improve the performance of insomnia patients. Decades later, no evidence that GABA-agonist hypnotics improve objective daytime performance in treating insomnia has been located. If there is a proven significant effect, it is to make performance worse89,134. Incidentally, the AHRQ Comparative Effectiveness Review mentioned no objective evidence of functional benefits from hypnotic drugs131. On average, most hypnotics make patients more sleepy the next day, not more fully awake.\n\nAfter 34 years, the author is still looking for evidence of objective functional benefit. In a recent letter to Sleep Medicine, readers were asked to inform us if “any U.S.-licensed hypnotic ever objectively improved any aspect of insomnia patients’ daytime function or any aspect of general health135.” So far, nobody has informed me of any such evidence.\n\nAccording to the U.S. National Ambulatory Medical Care Survey, insomnia is a stated reason for a patient’s visit in less than one quarter of office visits where a hypnotic is prescribed2, but for most of these drugs, insomnia could have been the only approved indication. Moreover, no diagnosis of any sleep disorder at all is made on 35% of office visits when a hypnotic is prescribed, and of the 65% of such patients who are diagnosed with a sleep disorder (such as hypersomnia and most forms of sleep apnea), often a hypnotic would be contraindicated2. Other data have likewise shown that hypnotics are commonly prescribed for patients who have no diagnosis or complaint of insomnia126,136,137. Thus, hypnotics are routinely being prescribed without any apparent valid indication in as much as three quarters of the cases. Moreover, it appears from the data reviewed above that in most cases or at least a great many, hypnotics are prescribed despite a specific contraindication. For example, in the 2015 Beers criteria of the American Geriatrics Society, the hypnotics of concern in this presentation are all listed as drugs to avoid109. It would be fanciful thinking to imagine that addicting hypnotics could be generally beneficial as usually prescribed: without indication and despite specific contraindications.\n\nAn instructive example is a 2006 advertisement representing that “[eszopiclone] provides a full night of sleep (7 to 8 hours).” An equivalent claim was made in a 2007 eszopiclone-hypnotic print advertisement titled “Sleep the night and seize the day. . . A better tomorrow begins tonight.” In the scientific study cited by both advertisements as evidence138, the average sleep of patients receiving eszopiclone 2 mg was 382 minutes (6 hours, 22 min) and for 3 mg, it was 412 minutes (6 hours, 52 min). The clinical results cited did not support the manufacturer’s claims to “a full 7 to 8 hours of sleep,” even though the 2 mg and 3 mg doses then studied were greater than the currently-recommended starting doses.\n\nAs for the manufacturer’s advertised benefits of “seizing the day,” and a “better tomorrow,” the eszopiclone manufacturer’s study demonstrated no significant objective improvement in measured next-day daytime performance or accomplishment. Specifically, an objective morning performance test did not demonstrate significantly better performance with eszopiclone than with placebo138.\n\nIt is not my intention to imply that misrepresentation in consumer advertising has come only from a single manufacturer. There have been many examples with other hypnotics.\n\n\nSummary of benefits and risks\n\nThe evidence is clear: the most popular hypnotics offer little to no benefit to patients in recommended doses. This statement applies most specifically to zolpidem, temazepam, eszopiclone, zaleplon, triazolam, flurazepam, quazepam, and barbiturates used for sleep. The most recent American Academy of Sleep Medicine's Clinical Guideline for Management of Chronic Insomnia139 stated that the primary goals of treatment of insomnia should be to increase sleep quantity and to enhance daytime function. To the contrary, popular hypnotics do not increase objective sleep substantially (if at all,) and for many patients, hypnotics cause substantial objective next-day functional impairment. In confirmation, FDA records show that the lead manufacturer of zolpidem (the most commonly-prescribed hypnotic) stated that the FDA-recommended dosage was ineffective32. The National Ambulatory Medical Care Survey indicated that over three quarters of hypnotic prescriptions are given to patients who do not even come to the physician for insomnia2. Yet the specified hypnotics do not substantially improve objective sleep or objective daytime performance and have no known objective benefits for any aspect of general health.\n\nContrasting with the questionable benefits, the popular benzodiazepine agonists in the U.S. are associated with increased mortality hazards, comparable to the hazards of barbiturates. Medical examiner data document that over 10,000 deaths every year are directly caused by and attributed to hypnotic drugs, and there is substantial evidence that hypnotics cause additional covert respiratory depression, suicides, infection, cancer, accidents, and other disorders that lead to a far larger number of deaths as well as to non-fatal morbidities and suffering. Use of hypnotics kills large numbers of Americans yearly; however, the exact number of deaths caused by hypnotics cannot be estimated from medical examiner data alone112, because most of the deaths produced by hypnotics are covert or indirect due to hypnotic-induced or hypnotic-exacerbated morbidities such as respiratory depression, infection, accidents, etc.\n\nThis presentation focused primarily on zolpidem, temazepam, eszopiclone, zaleplon, triazolam, flurazepam, quazepam, and barbiturates used for sleep (such as pentobarbital, amobarbital, and secobarbital). These specified drugs have little or no benefit for insomnia, and are commonly not even prescribed for presenting insomnia or other approved indications. These drugs were the focus because each has been shown epidemiologically to be associated with high mortality hazards16. There are other hypnotics approved for treating insomnia in the U.S.: specifically, diphenhydramine, ramelteon, doxepin, and suvorexant. Moreover, other drugs very commonly prescribed for sleep include trazodone (off label) and melatonin (unregulated). This presentation was not focused on these other drugs used as hypnotics, either because the epidemiologic and controlled-trials data have not been sufficient to assess their risks as hypnotics or because these drugs are approved and may be effective for indications other than insomnia.\n\nIn the supplement to the Geisinger Health System study, the best-estimate extrapolation from the data suggested that 300,000 – 500,000 deaths each year in the U.S. were associated with hypnotic usage16. The risks of under-estimation in this study were thought to be as great as the risks of over-estimation. Since this estimate was derived from risk ratios fully-adjusted for confounders, it is plausible that most of the association was causal.\n\nThis hypnotic mortality risk is almost comparable to that of cigarette smoking and many-fold greater than the risk to Americans of violent death.\n\nHypnotic drugs 300,000–500,000 U.S. deaths per year16\n\nCigarettes 560,000 U.S. deaths per year140\n\nMurders 14,196 U.S. deaths in 2013\n\nThe number of hypnotic-associated deaths may be almost comparable to the number of deaths attributed to cigarette smoking (560,000), cancer (585,000), or heart disease (611,000).\n\nApart from hypnotics, the preferred treatment for insomnia is the cognitive-behavioral treatment of insomnia, which appears to be more effective in the long run, better for comorbidities, and safer131. Cognitive-behavioral therapy can be effectively provided through written materials, internet training programs, and brief group therapies. It has been argued that cognitive-behavioral treatment saves money, compared to hypnotics141.\n\nCircadian rhythm timing disorders can cause the biologic propensity for sleep to be either delayed (causing trouble falling asleep and trouble waking in the morning) or too advanced (causing evening sleepiness and early awakening). It is unclear how often the circadian rhythm timing disorders have a more important role in insomnia than the cognitive-behavioral elements, but one estimate suggests that “eveningness” may be associated with trouble falling asleep in as much as one quarter of the adult population142. When circadian timing issues are important, properly timed bright light treatment can be a safe, effective, and inexpensive non-drug treatment that also has benefits for comorbidities such as depression, but more clinical trials are needed to better define the applicability of bright light treatment for insomnia.\n\nA very recent Clinical Practice Guideline from the American College of Physicians has recommended that cognitive behavioral therapy be the initial treatment for chronic insomnia disorder. Because of only very weak evidence for benefits and more impressive evidence for occasional serious harms, the Guideline left uncertain whether hypnotics were ever wise as a secondary treatment for short-term use, and the Guideline did not recommend long-term hypnotic prescribing at all143.", "appendix": "Competing interests\n\n\n\nSince 1979 publication of hypnotics’ epidemiology from the American Cancer Society CPSI study, the author has been a frequent critic of hypnotics risks and benefits, especially through his non-profit internet web site, www.DarkSideOfSleepingPills.com, that offers readers additional information and references about hypnotics. Dr. Kripke's family owns non-controlling stock and options in a large conglomerate that in turn invested a tiny percentage of its capital in Sanofi-Aventis and Johnson and Johnson stock. The author has advised the USA Food and Drug Administration to take certain actions regarding hypnotics (Petition available at http://www.regulations.gov/#!docketDetail;D=FDA-2015-P-3959), and related litigation may arise. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, other stock ownership or options, expert testimony, grants or patents received or pending or royalties. No writing assistance was utilized in the production of this manuscript.\n\n\nGrant information\n\nThe author declared that no grants were involved in supporting this work.\n\n\nAppendix A: Epidemiologic Studies of the Mortality Risks of Hypnotic Drugs\n\n1) Sun, Y., Lin, C. C., Lu, C. J., Hsu, C. Y., and Kao, C. H. Association Between Zolpidem and Suicide: A Nationwide Population-Based Case-Control Study. Mayo Clin Proc. 2016;91(3):308-315.\n\n2) Lan, T. Y., Zeng, Y. F., Tang, G. J., Kao, H. C., Chiu, H. J., Lan, T. H., and Ho, H. F. The use of hypnotics and mortality - A population-based retrospective cohort study. PLoS One. 10(12), e0145271. 2015.\n\n3) Palmaro A, Dupouy J, Lapeyre-Mestre M. Benzodiazepines and risk of death: Results from two large cohort studies in France and UK. Eur Neuropsychopharmacol 2015;25(10), 1566-1577.\n\n4) Chung, W. S., Lai, C. Y., Lin, C. L., and Kao, C. H. Adverse respiratory events associated with hypnotics use in patients of chronic obstructive pulmonary disease: A population-based case-control Study. Medicine (Baltimore) 94(27), e1110. 2015.\n\n5) Kriegbaum, M., Hendriksen, C. Vass, M., Mortensen, E. L., Osler, M. Hypnotics and mortality—partial confounding by disease, substance abuse and socioeconomic factors? Pharmacoepidemiol Drug Saf 2015;24(7):779-783.\n\n6) Pinot J, Herr M, Robine JM, Aegerter P, Arvieu JJ, Ankri J. Does the Prescription of Anxiolytic and Hypnotic Drugs Increase Mortality in Older Adults? J Am Geriatr Soc 2015;63(6):1263-5.\n\n7) Weisberg DF, Gordon KS, Barry DT, Becker WC, Crystal S, Edelman EJ, Gaither J, Gordon AJ, Goulet J, Kerns RD, Moore BA, Tate J, Justice AC, Fiellin DA. Long-term Prescription of Opioids and/or Benzodiazepines and Mortality Among HIV-Infected and Uninfected Patients. J Acquir Immune Defic Syndr 2015;69(2):223-33.\n\n8) Nakafero G, Sanders RD, Nguyen-Van-Tam JS, Myles PR. Association between benzodiazepine use and exacerbations and mortality in patients with asthma: a matched case-control and survival analysis using the United Kingdom Clinical Practice Research Datalink. Pharmacoepidemiol Drug Saf 2015;24(8):793-802.\n\n9) Neutel CI, Johansen HL. Association between hypnotics use and increased mortality: causation or confounding? Eur J Clin Pharmacol 2015;71(5):637-42.\n\n10) Frandsen R, Baandrup L, Kjellberg J, Ibsen R, Jennum P. Increased all-cause mortality with psychotropic medication in Parkinson’s disease and controls: a national register-based study. Parkinsonism Relat Disord 2014;20(11):1124-8.\n\n11) Weich S, Pearce HL, Croft P, Singh S, Crome I, Bashford J, Frisher M. Effect of anxiolytic and hypnotic drug prescriptions on mortality hazards: retrospective cohort study. BMJ 2014;348:g1996.\n\n12) Chen H-C, Su T-P, Chou P. A 9-year Follow-up Study of Sleep Patterns and Mortality in Community-Dwelling Older Adults in Taiwan. Sleep 2013;36(8):1187-98.\n\n13) Gunnell D, Chang SS, Tsai MK, Tsao CK, Wen CP. Sleep and suicide: an analysis of a cohort of 394,000 Taiwanese adults. Soc Psychiatry Psychiatr Epidemiol. 2013 Apr 2;48:1457-65.\n\n14) Jaussent I, Ancelin ML, Berr C, Peres K, Scali J, Besset A, Ritchie K, Dauvilliers Y. Hypnotics and mortality in an elderly general population: a 12-year prospective study. BMC Med 2013;11(1):212.\n\n15) Obiora E, Hubbard R, Sanders RD, Myles PR. The impact of benzodiazepines on occurrence of pneumonia and mortality from pneumonia: a nested case-control and survival analysis in a population-based cohort. Thorax 2012;68(2):163-70.\n\n16) Hartz A, Ross JJ. Cohort study of the association of hypnotic use with mortality in postmenopausal women. BMJ Open 2012;2:pii: e001413. doi: 10.1136/bmjopen-2012-001413.\n\n17) Kripke DF, Langer RD, Kline LE. Hypnotics’ association with mortality or cancer: a matched cohort study. BMJ Open 2012;2(1):e000850.\n\n18) Gisev N, Hartikainen S, Chen TF, Korhonen M, Bell JS. Mortality associated with benzodiazepines and benzodiazepine-related drugs among community-dwelling older people in Finland: a population-based retrospective cohort study. Can J Psychiatry 2011;56(6):377-81.\n\n19) Rod NH, Vahtera J, Westerlund H, Kivimaki M, Zins M, Goldberg M, Lange T. Sleep Disturbances and Cause-Specific Mortality: Results From the GAZEL Cohort Study. Am J Epidemiol 2010;173(3):300-9.\n\n20) Belleville G. Mortality hazard associated with anxiolytic and hypnotic drug use in the national population health survey. Can J Psychiatry 2010;55(9):558-67.\n\n21) Mallon L, Broman JE, Hetta J. Is usage of hypnotics associated with mortality? Sleep Med 2009;10(3):279-86.\n\n22) Winkelmayer WC, Mehta J, Wang PS. Benzodiazepine use and mortality of incident dialysis patients in the United States. Kidney Int 2007;72(11):1388-93.\n\n23) Hublin C, Partinen M, Koskenvuo M, Kaprio J. Sleep and mortality: a population-based 22-year follow-up study. Sleep 2007;30(10):1245-53.\n\n24) Hoffmann VP, Dossenbach M, West TM, Lowry AJ. Mortality in a cohort of outpatients with schizophrenia: 3-year outcomes from the Intercontinental Outpatient Health Outcomes Study (IC-SOHO). Biol Psychiatry 61(8S):163S-164S. Accessed 2007.\n\n25) Hausken AM, Skurtveit S, Tverdal A. Use of anxiolytic or hypnotic drugs and total mortality in a general middle-aged population. Pharmacoepidemiol Drug Saf 2007;16(8):913-8.\n\n26) Fukuhara S, Green J, Albert J, Mihara H, Pisoni R, Yamazaki S, Akiba T, Akizawa T, Asano Y, Saito A, Port F, Held P, Kurokawa K. Symptoms of depression, prescription of benzodiazepines, and the risk of death in hemodialysis patients in Japan. Kidney Int 2006;70(10):1866-72.\n\n27) Lack LC, Prior K, Luszcz M. 708. Does insomnia kill the elderly? Sleep 29[Abstract Supplement], A240. Accessed 2006.\n\n28) Phillips B, Mannino DM. Does insomnia kill? Sleep 2005;28(8):965-71.\n\n29) Ahmad R, Bath PA. Identification of risk factors for 15-year mortality among community-dwelling older people using Cox regression and a genetic algorithm. J Gerontol A Biol Sci Med Sci 2005;60A:1052-8.\n\n30) Mallon L, Broman J-E, Hetta J. Sleep complaints predict coronary artery disease mortality in males: a 12-year follow-up study of a middle-aged Swedish population. J Int Med 2002;251:207-16.\n\n31) Hedner J, Caidahl K, Sjoland H, Karlsson T, Herlitz J. Sleep habits and their association with mortality during 5-year follow-up after coronary artery bypass surgery. Acta Cardiol 2002;57(5):341-8.\n\n32) Kripke DF, Garfinkel L, Wingard DL, Klauber MR, Marler MR. Mortality associated with sleep duration and insomnia. Arch Gen Psychiatry 2002;59(2):131-6.\n\n33) Kripke DF, Klauber MR, Wingard DL, Fell RL, Assmus JD, Garfinkel L. Mortality hazard associated with prescription hypnotics. Biol Psychiatry 1998;43(9):687-93.\n\n34) Merlo J, Ostergren PO, Mansson NO, Hanson BS, Ranstam J, Blennow G, Isacsson SO, Melander A. Mortality in elderly men with low psychosocial coping resources using anxiolytic-hypnotic drugs. Scand J Public Health 2000;28(4):294-7.\n\n35) Sundquist J, Ekedahl A, Johansson S-E. Sales of tranquillizers, hypnotics/sedatives and antidepressants and their relationship with underprivileged area score and mortality and suicide rates. Eur J Clin Pharmacol 1996;51:105-9.\n\n36) Hays JC, Blazer DG, Foley DJ. Risk of napping: excessive daytime sleepiness and mortality in an older community population. J Am Geriatr Soc 1996;44:693-8.\n\n37) Merlo J, Hedblad B, Ogren M, Ranstam J, Ostergren PO, Ekedahl A, Hanson BS, Isacsson SO, Liedholm H, Melander A. Increased risk of ischaemic heart disease mortality in elderly men using anxiolytics-hypnotics and analgesics. Eur J Clin Pharmacol 1996;49:261-5.\n\n38) Brabbins CJ, Dewey ME, Copeland RM, Davidson IA, McWilliam C, Saunders P, Sharma VK, Sullivan C. Insomnia in the elderly: Prevalence, gender differences and relationships with morbidity and mortality. Int J Ger Psych 1993;8:473-80.\n\n39) Thorogood M, Cowen P, Mann J, Murphy M, Vessey M. Fatal myocardial infarction and use of psychotropic drugs in young women. Lancet 1992;340:1067-8.\n\n40) Isacson D, Carsjo K, Bergman U, Blackburn JL. Long-term use of benzodiazepines in a Swedish community: an eight-year follow-up. J Clin Epidemiol 1992 Apr;45(4):429-36.\n\n41) Rumble R, Morgan K. Hypnotics, sleep, and mortality in elderly people. J Am Geriatr Soc 1992;40:787-91.\n\n42) Kripke DF, Simons RN, Garfinkel L, Hammond EC. Short and long sleep and sleeping pills: Is increased mortality associated? Arch Gen Psychiatry 1979;36(1):103-16.\n\n\nReferences\n\nRudd RA, Aleshire N, Zibbell JE, et al.: Increases in Drug and Opioid Overdose Deaths--United States, 2000–2014. MMWR Morb Mortal Wkly Rep. 2016; 64(50–51): 1378–82. PubMed Abstract \| Publisher Full Text\n\nFord ES, Wheaton AG, Cunningham TJ, et al.: Trends in Outpatient Visits for Insomnia, Sleep Apnea, and Prescriptions for Sleep Medications among US Adults: Findings from the National Ambulatory Medical Care Survey 1999–2010. Sleep. 2014; 37(8): 1283–93. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLarochelle MR, Zhang F, Ross-Degnan D, et al.: Trends in opioid prescribing and co-prescribing of sedative hypnotics for acute and chronic musculoskeletal pain: 2001–2010. Pharmacoepidemiol Drug Saf. 2015; 24(8): 885–92. PubMed Abstract \| Publisher Full Text\n\nKolata G, Cohen S: Drug overdoses propel rise in mortality rates of young whites. International NY Times. 2016. Reference Source\n\nCase A, Deaton A: Rising morbidity and mortality in midlife among white non-Hispanic Americans in the 21st century. Proc Natl Acad Sci U S A. 2015; 112(49): 15078–83. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSubstance Abuse and Mental Health Services Administration CfBHSaQ: The DAWN Report: Benzodiazepines in combination with opioid pain relievers or alcohol: Greater risk of more serious ED visit outcomes. Dawn. 2014; 192: 1–6. Reference Source\n\nRockett IR, Caine ED: Self-injury is the eighth leading cause of death in the United States: It is time to pay attention. JAMA Psychiatry. 2015; 72(11): 1069–70. PubMed Abstract \| Publisher Full Text\n\nCurtin SC, Warner M, Hedegaard H: Increase in Suicide in the United States, 1999–2014. Hyattsville, M.D., National Center for Health Statistics. NCHS Data Brief. 2016; (241): 1–8. PubMed Abstract\n\nYoussef NA, Rich CL: Does acute treatment with sedatives/hypnotics for anxiety in depressed patients affect suicide risk? A literature review. Ann Clin Psychiatry. 2008; 20(3): 157–69. PubMed Abstract \| Publisher Full Text\n\nMerlo J, Hedblad B, Ogren M, et al.: Increased risk of ischaemic heart disease mortality in elderly men using anxiolytics-hypnotics and analgesics. Results of the 10-year follow-up of the prospective population study \"Men born in 1914\", Malmo, Sweden. Eur J Clin Pharmacol. 1996; 49(4): 261–5. PubMed Abstract \| Publisher Full Text\n\nLan TY, Zeng YF, Tang GJ, et al.: The Use of Hypnotics and Mortality--A Population-Based Retrospective Cohort Study. PLoS One. 2015; 10(12): e0145271. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKripke DF: Mortality Risk of Hypnotics: Strengths and Limits of Evidence. Drug Saf. 2016; 39(2): 93–107. PubMed Abstract \| Publisher Full Text\n\nIoannidis JP: Exposure-wide epidemiology: revisiting Bradford Hill. Stat Med. 2016; 35(11): 1749–62. PubMed Abstract \| Publisher Full Text\n\nPalmaro A, Dupouy J, Lapeyre-Mestre M: Benzodiazepines and risk of death: Results from two large cohort studies in France and UK. Eur Neuropsychopharmacol. 2015; 25(10): 1566–77. PubMed Abstract \| Publisher Full Text\n\nChung WS, Lai CY, Lin CL, et al.: Adverse Respiratory Events Associated With Hypnotics Use in Patients of Chronic Obstructive Pulmonary Disease: A Population-Based Case-Control Study. Medicine (Baltimore). 2015; 94(27): e1110. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKripke DF, Langer RD, Kline LE: Hypnotics' association with mortality or cancer: a matched cohort study. BMJ Open. 2012; 2(1): e000850. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWeich S, Pearce HL, Croft P, et al.: Effect of anxiolytic and hypnotic drug prescriptions on mortality hazards: retrospective cohort study. BMJ. 2014; 348: g1996. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNeutel CI, Johansen HL: Association between hypnotics use and increased mortality: causation or confounding? Eur J Clin Pharmacol. 2015; 71(5): 637–42. PubMed Abstract \| Publisher Full Text\n\nKripke DF, Garfinkel L, Wingard DL, et al.: Mortality associated with sleep duration and insomnia. Arch Gen Psychiatry. 2002; 59(2): 131–6. PubMed Abstract \| Publisher Full Text\n\nChen HC, Su TP, Chou P: A nine-year follow-up study of sleep patterns and mortality in community-dwelling older adults in Taiwan. Sleep. 2013; 36(8): 1187–98. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMallon L, Broman JE, Hetta J: Is usage of hypnotics associated with mortality? Sleep Med. 2009; 10(3): 279–86. PubMed Abstract \| Publisher Full Text\n\nBelleville G: Mortality hazard associated with anxiolytic and hypnotic drug use in the national population health survey. Can J Psychiatry. 2010; 55(9): 558–67. PubMed Abstract\n\nKripke DF: Greater incidence of depression with hypnotic use than with placebo. BMC Psychiatry. 2007; 7: 42. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKripke DF: Hypnotics cause insomnia: evidence from clinical trials. Sleep Med. 2014; 15(9): 1168–9. PubMed Abstract \| Publisher Full Text\n\nCooper JR: Sedative-Hypnotic Drugs: Risks and Benefits. Rockville, MD: U.S. Department of HEW, National Inst. on Drug Abuse; 1977. Reference Source\n\nKripke DF, Garfinkel L: Excess nocturnal deaths related to sleeping pill and tranquilliser use. Lancet. 1984; 1(8368): 99. PubMed Abstract \| Publisher Full Text\n\nLeary S; , Members of Panel: AVMA Guidelines for the Euthanasia of Animals: 2013. Edition 2013.0.1. Schaumburg, IL: American Veterinary Medical Association; 2013. Reference Source\n\nWikibooks contributors: Pentobarbital. Accessed 8-26-2015. Wikibooks, The Free Textbook Project. Wikibooks. Reference Source\n\nFarkas RH, Katz R, Illoh K, et al.: Application Number 204569Orig1s000: Medical Review(s). Accessed 6-25-2013. Reference Source\n\nLalley PM: Opioidergic and dopaminergic modulation of respiration. Respir Physiol Neurobiol. 2008; 164(1–2): 160–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJoya FL, Kripke DF, Loving RT, et al.: Meta-Analyses of Hypnotics and Infections: eszopiclone, ramelteon, zaleplon, and zolpidem. J Clin Sleep Med. 2009; 5(4): 377–83. PubMed Abstract \| Free Full Text\n\nFarkas R; Center for Drug Evaluation and Research: Approval Package for: Application Number: 019908Orig1s032s034 021774Orig1s013s015. Accessed 2013. Silver Spring, MD, FDA. 2013. Reference Source\n\nObiora E, Hubbard R, Sanders RD, et al.: The impact of benzodiazepines on occurrence of pneumonia and mortality from pneumonia: a nested case-control and survival analysis in a population-based cohort. Thorax. 2013; 68(2): 163–70. PubMed Abstract \| Publisher Full Text\n\nIqbal U, Syed-Abdul S, Nguyen PA, et al.: The impact of benzodiazepines on occurrence of pneumonia and mortality from pneumonia: a nested case-control and survival analysis in a population-based cohort. Thorax. 2013; 68(6): 591–2. PubMed Abstract \| Publisher Full Text\n\nHuang CY, Chou FH, Huang YS, et al.: The association between zolpidem and infection in patients with sleep disturbance. J Psychiatr Res. 2014; 54(7): 116–20. PubMed Abstract \| Publisher Full Text\n\nLiao KF, Lin CL, Lai SW, et al.: Zolpidem Use Associated With Increased Risk of Pyogenic Liver Abscess: A Case-Control Study in Taiwan. Medicine (Baltimore). 2015; 94(32): e1302. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNakafero G, Sanders RD, Nguyen-Van-Tam JS, et al.: Association between benzodiazepine use and exacerbations and mortality in patients with asthma: a matched case-control and survival analysis using the United Kingdom Clinical Practice Research Datalink. Pharmacoepidemiol Drug Saf. 2015; 24(8): 793–802. PubMed Abstract \| Publisher Full Text\n\nPavon JM, Zhao Y, McConnell E, et al.: Identifying risk of readmission in hospitalized elderly adults through inpatient medication exposure. J Am Geriatr Soc. 2014; 62(6): 1116–21. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSanders RD, Godlee A, Fujimori T, et al.: Benzodiazepine augmented γ-amino-butyric acid signaling increases mortality from pneumonia in mice. Crit Care Med. 2013; 41(7): 1627–36. PubMed Abstract \| Publisher Full Text\n\nHuemer HP, Lassnig C, Nowotny N, et al.: Diazepam leads to enhanced severity of orthopoxvirus infection and immune suppression. Vaccine. 2010; 28(38): 6152–8. PubMed Abstract \| Publisher Full Text\n\nMassoco C, Palermo-Neto J: Effects of midazolam on equine innate immune response: a flow cytometric study. Vet Immunol Immunopathol. 2003; 95(1–2): 11–9. PubMed Abstract \| Publisher Full Text\n\nTorres SR, Fröde TS, Nardi GM, et al.: Anti-inflammatory effects of peripheral benzodiazepine receptor ligands in two mouse models of inflammation. Eur J Pharmacol. 2000; 408(2): 199–211. PubMed Abstract \| Publisher Full Text\n\nKripke DF: Possibility that certain hypnotics might cause cancer in skin. J Sleep Res. 2008; 17(3): 245–50. PubMed Abstract \| Publisher Full Text\n\nRoth T, Zammit GK, Scharf MB, et al.: Efficacy and safety of as-needed, post bedtime dosing with indiplon in insomnia patients with chronic difficulty maintaining sleep. Sleep. 2007; 30(12): 1731–8. PubMed Abstract \| Free Full Text\n\nScharf MB, Black J, Hull S, et al.: Long-term nightly treatment with indiplon in adults with primary insomnia: results of a double-blind, placebo-controlled, 3-month study. Sleep. 2007; 30(6): 743–52. PubMed Abstract \| Free Full Text\n\nWeissinger J: NDA 19-908 Ambien Pharmacology Memos & Exclusivity Summary. 1991. Reference Source\n\nAndreason PJ, Brugge K, Katz R: Center for Drug Evaluation and Research. Approval Package for: Application Number 21-476: Medical Review(s). 2004. Reference Source\n\nAmerio A, Gálvez JF, Odone A, et al.: Carcinogenicity of psychotropic drugs: A systematic review of US Food and Drug Administration-required preclinical in vivo studies. Aust N Z J Psychiatry. 2015; 49(8): 686–96. PubMed Abstract \| Publisher Full Text\n\nRoca R, McNeil DE; Center for Drug Evaluation and Research: Application Number 21-782: Medical Review(s). 2005; 1–315. Reference Source\n\nFitzgerald GG; Center for Drug Evaluation and Research: Application Number 020859: Pharmacology Review(s). FDA. 1998. Reference Source\n\nWikipedia contributors: Clastogen. Wikipedia. Accessed 6-20-2014. Reference Source\n\nWasserman A, Mellon RD; Center for Drug Evaluation and Research: Application number 21-774: Pharmacology Reviews(s). FDA. 2005. Reference Source\n\nGrivennikov SI, Greten FR, Karin M: Immunity, inflammation, and cancer. Cell. 2010; 140(6): 883–99. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMallon L, Broman JE, Hetta J: Sleep complaints predict coronary artery disease mortality in males: a 12-year follow-up study of a middle-aged Swedish population. J Int Med. 2002; 251(3): 207–16. PubMed Abstract \| Publisher Full Text\n\nKripke DF, Klauber MR, Wingard DL, et al.: Mortality hazard associated with prescription hypnotics. Biol Psychiatry. 1998; 43(9): 687–93. PubMed Abstract \| Publisher Full Text\n\nHartz A, Ross JJ: Cohort study of the association of hypnotic use with mortality in postmenopausal women. BMJ Open. 2012; 2(5): pii: e001413. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKao CH, Sun LM, Liang JA, et al.: Relationship of zolpidem and cancer risk: a Taiwanese population-based cohort study. Mayo Clin Proc. 2012; 87(5): 430–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWen CP, Levy DT, Cheng TY, et al.: Smoking behaviour in Taiwan, 2001. Tob Control. 2005; 14(Suppl 1): i51–i55. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChu NF: Prevalence of obesity in Taiwan. Obes Rev. 2005; 6(4): 271–4. PubMed Abstract \| Publisher Full Text\n\nKao CH, Sun LM, Su KP, et al.: Benzodiazepine use possibly increases cancer risk: a population-based retrospective cohort study in Taiwan. J Clin Psychiatry. 2012; 73(4): e555–e560. PubMed Abstract \| Publisher Full Text\n\nIqbal U, Jian WS, Huang CW, et al.: Do all hypnotic and sedatives have risk for cancer? Sleep Med. 2015; pii: S1389-9457(15)00867-9, in press. PubMed Abstract \| Publisher Full Text\n\nJiao L, Duan Z, Sangi-Haghpeykar H, et al.: Sleep duration and incidence of colorectal cancer in postmenopausal women. Br J Cancer. 2013; 108(1): 213–21. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPottegård A, Friis S, Andersen M, et al.: Use of benzodiazepines or benzodiazepine related drugs and the risk of cancer: a population-based case-control study. Br J Clin Pharmacol. 2013; 75(5): 1356–64. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSivertsen B, Salo P, Pentti J, et al.: Use of sleep medications and risk of cancer: a matched case-control study. Sleep Med. 2015; 16(12): 1552–5. PubMed Abstract \| Publisher Full Text\n\nKripke DF, Langer RD: Evidence for harm, comment on 'Use of benzodiazepines or benzodiazepine related drugs and the risk of cancer: a population-based case-control study'. Br J Clin Pharmacol. 2014; 78(1): 186–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGagliardi GS, Shah AP, Goldstein M, et al.: Effect of zolpidem on the sleep arousal response to nocturnal esophageal acid exposure. Clin Gastroenterol Hepatol. 2009; 7(9): 948–52. PubMed Abstract \| Publisher Full Text\n\nFriedman GD, Udaltsova N, Chan J, et al.: Screening pharmaceuticals for possible carcinogenic effects: initial positive results for drugs not previously screened. Cancer Causes Control. 2009; 20(10): 1821–35. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFriedman GD, Jiang SF, Udaltsova N, et al.: Epidemiologic evaluation of pharmaceuticals with limited evidence of carcinogenicity. Int J Cancer. 2009; 125(9): 2173–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWu JQ, Appleman ER, Salazar RD, et al.: Cognitive Behavioral Therapy for Insomnia Comorbid With Psychiatric and Medical Conditions: A Meta-analysis. JAMA Intern Med. 2015; 175(9): 1461–72. PubMed Abstract \| Publisher Full Text\n\nFava M, McCall WV, Krystal A, et al.: Eszopiclone co-administered with fluoxetine in patients with insomnia coexisting with major depressive disorder. Biol Psychiatry. 2006; 59(11): 1052–60. PubMed Abstract \| Publisher Full Text\n\nFava M, Asnis GM, Shrivastava RK, et al.: Improved insomnia symptoms and sleep-related next-day functioning in patients with comorbid major depressive disorder and insomnia following concomitant zolpidem extended-release 12.5 mg and escitalopram treatment: a randomized controlled trial. J Clin Psychiatry. 2011; 72(7): 914–28. PubMed Abstract \| Publisher Full Text\n\nTan TL, Bixler EO, Kales A, et al.: Early morning insomnia, daytime anxiety, and organic mental disorder associated with triazolam. J Fam Pract. 1985; 20(6): 592–4. PubMed Abstract\n\nLevi J, Segal LM, Martin A: The Facts Hurt: A state-by-state injury prevention policy report. Washington, D.C.: Trust for America's Health; 2015. Reference Source\n\nSun Y, Lin CC, Lu CJ, et al.: Association Between Zolpidem and Suicide: A Nationwide Population-Based Case-Control Study. Mayo Clin Proc. 2016; 91(3): 308–15. PubMed Abstract \| Publisher Full Text\n\nPressman MR: Sleep driving: sleepwalking variant or misuse of z-drugs? Sleep Med Rev. 2011; 15(5): 285–92. PubMed Abstract \| Publisher Full Text\n\nBreiding MJ, Wiersema B: Variability of undetermined manner of death classification in the US. Inj Prev. 2006; 12(Suppl 2): ii49–ii54. PubMed Abstract \| Publisher Full Text\n\nCarlsten A, Waern M: Are sedatives and hypnotics associated with increased suicide risk of suicide in the elderly? BMC Geriatr. 2009; 9(1): 20. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBrower KJ, McCammon RJ, Wojnar M, et al.: Prescription sleeping pills, insomnia, and suicidality in the National Comorbidity Survey Replication. J Clin Psychiatry. 2011; 72(4): 515–21. PubMed Abstract \| Publisher Full Text\n\nGunnell D, Chang SS, Tsai MK, et al.: Sleep and suicide: an analysis of a cohort of 394,000 Taiwanese adults. Soc Psychiatry Psychiatr Epidemiol. 2013; 48(9): 1457–65. PubMed Abstract \| Publisher Full Text\n\nDarke S, Deady M, Duflou J: Toxicology and characteristics of deaths involving zolpidem in New South Wales, Australia 2001–2010. J Forensic Sci. 2012; 57(5): 1259–62. PubMed Abstract \| Publisher Full Text\n\nJohnson LC, Chernik DA: Sedative-hypnotics and human performance. Psychopharmacology (Berl). 1982; 76(2): 101–13. PubMed Abstract \| Publisher Full Text\n\nDrover D, Lemmens H, Naidu S, et al.: Pharmacokinetics, pharmacodynamics, and relative pharmacokinetic/pharmacodynamic profiles of zaleplon and zolpidem. Clin Ther. 2000; 22(12): 1443–61. PubMed Abstract \| Publisher Full Text\n\nPoceta JS: Zolpidem ingestion, automatisms, and sleep driving: a clinical and legal case series. J Clin Sleep Med. 2011; 7(6): 632–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTsai JH, Yang P, Chen CC, et al.: Zolpidem-induced amnesia and somnambulism: rare occurrences? Eur Neuropsychopharmacol. 2009; 19(1): 74–6. PubMed Abstract \| Publisher Full Text\n\nMorgenthaler TI, Silber MH: Amnestic sleep-related eating disorder associated with zolpidem. Sleep Med. 2002; 3(4): 323–7. PubMed Abstract \| Publisher Full Text\n\nOster G, Huse DM, Adams SF, et al.: Benzodiazepine tranquilizers and the risk of accidental injury. Am J Public Health. 1990; 80(12): 1467–70. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLai MM, Lin CC, Lin CC, et al.: Long-term use of zolpidem increases the risk of major injury: a population-based cohort study. Mayo Clin Proc. 2014; 89(5): 589–94. PubMed Abstract \| Publisher Full Text\n\nChung SD, Lin CC, Wang LH, et al.: Zolpidem Use and the Risk of Injury: A Population-Based Follow-Up Study. PLoS One. 2013; 8(6): e67459. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVerster JC, Veldhuijzen DS, Patat A, et al.: Hypnotics and driving safety: meta-analyses of randomized controlled trials applying the on-the-road driving test. Curr Drug Saf. 2006; 1(1): 63–71. PubMed Abstract \| Publisher Full Text\n\nVerster JC, Spence DW, Shahid A, et al.: Zopiclone as positive control in studies examining the residual effects of hypnotic drugs on driving ability. Curr Drug Saf. 2011; 6(4): 209–18. PubMed Abstract \| Publisher Full Text\n\nVerster JC, Roth T: Drivers can poorly predict their own driving impairment: a comparison between measurements of subjective and objective driving quality. Psychopharmacology (Berl). 2012; 219(3): 775–81. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOrriols L, Philip P, Moore N, et al.: Benzodiazepine-like hypnotics and the associated risk of road traffic accidents. Clin Pharmacol Ther. 2011; 89(4): 595–601. PubMed Abstract \| Publisher Full Text\n\nPhilip P, Chaufton C, Orriols L, et al.: Complaints of Poor Sleep and Risk of Traffic Accidents: A Population-Based Case-Control Study. PLoS One. 2014; 9(12): e114102. PubMed Abstract \| Publisher Full Text\n\nHansen RN, Boudreau DM, Ebel BE, et al.: Sedative Hypnotic Medication Use and the Risk of Motor Vehicle Crash. Am J Public Health. 2015; 105(8): e64–e69. PubMed Abstract \| Publisher Full Text\n\nHemmelgarn B, Suissa S, Huang A, et al.: Benzodiazepine use and the risk of motor vehicle crash in the elderly. JAMA. 1997; 278(1): 27–31. PubMed Abstract \| Publisher Full Text\n\nGustavsen I, Bramness JG, Skurtveit S, et al.: Road traffic accident risk related to prescriptions of the hypnotics zopiclone, zolpidem, flunitrazepam and nitrazepam. Sleep Med. 2008; 9(8): 818–22. PubMed Abstract \| Publisher Full Text\n\nHedlund J, Ahlner J, Kristiansson M, et al.: A population-based study on toxicological findings in Swedish homicide victims and offenders from 2007 to 2009. Forensic Sci Int. 2014; 244: 25–9. PubMed Abstract \| Publisher Full Text\n\nTinetti ME, Speechley M, Ginter SF: Risk factors for falls among elderly persons living in the community. N Engl J Med. 1988; 319(26): 1701–7. PubMed Abstract \| Publisher Full Text\n\nWang PS, Bohn RL, Glynn RJ, et al.: Hazardous benzodiazepine regimens in the elderly: Effects of half-life, dosage, and duration on risk of hip fracture. Am J Psychiatry. 2001; 158(6): 892–8. PubMed Abstract \| Publisher Full Text\n\nWang PS, Bohn RL, Glynn RJ, et al.: Zolpidem use and hip fractures in older people. J Am Geriatr Soc. 2001; 49(12): 1685–90. PubMed Abstract \| Publisher Full Text\n\nCumming RG, Le Couteur DG: Benzodiazepines and risk of hip fractures in older people: a review of the evidence. CNS Drugs. 2003; 17(11): 825–37. PubMed Abstract \| Publisher Full Text\n\nKang DY, Park S, Rhee CW, et al.: Zolpidem use and risk of fracture in elderly insomnia patients. J Prev Med Public Health. 2012; 45(4): 219–26. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBerry SD, Lee Y, Cai S, et al.: Nonbenzodiazepine Sleep Medication Use and Hip Fractures in Nursing Home Residents. JAMA Intern Med. 2013; 173(9): 754–61. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKolla BP, Lovely JK, Mansukhani MP, et al.: Zolpidem is independently associated with increased risk of inpatient falls. J Hosp Med. 2013; 8(1): 1–6. PubMed Abstract \| Publisher Full Text\n\nDiem SJ, Ewing SK, Stone KL, et al.: Use of non-benzodiazepine sedative hypnotics and risk of falls in older men. J Gerontol Geriatr Res. 2014; 3(3): 158. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPark SM, Ryu J, Lee DR, et al.: Zolpidem use and risk of fractures: a systematic review and meta-analysis. Osteoporos Int. In press, 2016; 1–10. PubMed Abstract \| Publisher Full Text\n\nAvidan AY, Fries BE, James ML, et al.: Insomnia and hypnotic use, recorded in the minimum data set, as predictors of falls and hip fractures in Michigan nursing homes. J Am Geriatr Soc. 2005; 53(6): 955–62. PubMed Abstract \| Publisher Full Text\n\nOkamoto M, Rao SN, Aaronson LM, et al.: Ethanol drug interaction with chlordiazepoxide and pentobarbital. Alcohol Clin Exp Res. 1985; 9(6): 516–21. PubMed Abstract \| Publisher Full Text\n\nBy the American Geriatrics Society 2015 Beers Criteria Update Expert Panel: American Geriatrics Society 2015 Updated Beers Criteria for Potentially Inappropriate Medication Use in Older Adults. J Am Geriatr Soc. 2015; 63(11): 2227–46. PubMed Abstract \| Publisher Full Text\n\nKripke DF, Ancoli-Israel S, Klauber MR, et al.: Prevalence of sleep-disordered breathing in ages 40-64 years: A population-based survey. Sleep. 1997; 20(1): 65–76. PubMed Abstract \| Free Full Text\n\nAncoli-Israel S, Kripke DF, Klauber MR, et al.: Sleep-disordered breathing in community-dwelling elderly. Sleep. 1991; 14(6): 486–95. PubMed Abstract \| Free Full Text\n\nJann M, Kennedy WK, Lopez G: Benzodiazepines: a major component in unintentional prescription drug overdoses with opioid analgesics. J Pharm Pract. 2014; 27(1): 5–16. PubMed Abstract \| Publisher Full Text\n\nWebster LR, Choi Y, Desai H, et al.: Sleep-disordered breathing and chronic opioid therapy. Pain Med. 2008; 9(4): 425–32. PubMed Abstract \| Publisher Full Text\n\nMogri M, Desai H, Webster L, et al.: Hypoxemia in patients on chronic opiate therapy with and without sleep apnea. Sleep Breath. 2009; 13(1): 49–57. PubMed Abstract \| Publisher Full Text\n\nPark TW, Saitz R, Ganoczy D, et al.: Benzodiazepine prescribing patterns and deaths from drug overdose among US veterans receiving opioid analgesics: case-cohort study. BMJ. 2015; 350: h2698. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWeisberg DF, Gordon KS, Barry DT, et al.: Long-term Prescription of Opioids and/or Benzodiazepines and Mortality Among HIV-Infected and Uninfected Patients. J Acquir Immune Defic Syndr. 2015; 69(2): 223–33. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiang Y, Goros MW, Turner BJ: Drug Overdose: Differing Risk Models for Women and Men among Opioid Users with Non-Cancer Pain. Pain Med. 2016; pii: pnw071, in press. PubMed Abstract \| Publisher Full Text\n\nDwyer LL, Lau DT, Shega JW: Medications That Older Adults in Hospice Care in the United States Take, 2007. J Am Geriatr Soc. 2015; 63(11): 2282–9. PubMed Abstract \| Publisher Full Text\n\nMoore TJ: ISMP Quarter Watch: Monitoring FDA MedWatch Reports. Accessed 5-06-2015. Philadelphia, PA, ISMP. Quarter Watch. Reference Source\n\nHampton LM, Daubresse M, Chang HY, et al.: Emergency Department Visits by Adults for Psychiatric Medication Adverse Events. JAMA Psychiatry. 2014; 71(9): 1006–14. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDowell D, Haegerich TM, Chou R: CDC Guideline for Prescribing Opioids for Chronic Pain--United States, 2016. JAMA. 2016; 315(15): 1624–45. PubMed Abstract \| Publisher Full Text\n\nInstitute of Medicine (US) Committee on Halcion: An Assessment of Data Adequacy and Confidence: Halcion: An Independent Assessment of Safety and Efficacy Data. Washington, D.C.: National Academy Press; 1997. PubMed Abstract \| Publisher Full Text\n\nHarnod T, Wang YC, Kao CH: Association Between Benzodiazepine Use and Epilepsy Occurrence: A Nationwide Population-Based Case-Control Study. Medicine (Baltimore). 2015; 94(37): e1571. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHammond EC: Some preliminary findings on physical complaints from a prospective study of 1,064,004 men and women. Am J Public Health Nations Health. 1964; 54(1): 11–23. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKripke DF, Langer RD, Elliott JA, et al.: Mortality related to actigraphic long and short sleep. Sleep Med. 2011; 12(1): 28–33. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKripke DF, Simons RN, Garfinkel L, et al.: Short and long sleep and sleeping pills: Is increased mortality associated? Arch Gen Psychiatry. 1979; 36(1): 103–16. PubMed Abstract \| Publisher Full Text\n\nBuscemi N, Vandermeer B, Friesen C, et al.: The efficacy and safety of drug treatments for chronic insomnia in adults: a meta-analysis of RCTs. J Gen Intern Med. 2007; 22(9): 1335–50. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFarkas RH, Unger EF, Temple R: Zolpidem and driving impairment--identifying persons at risk. N Engl J Med. 2013; 369(8): 689–91. PubMed Abstract \| Publisher Full Text\n\nRosenberg R, Caron J, Roth T, et al.: An assessment of the efficacy and safety of eszopiclone in the treatment of transient insomnia in healthy adults. Sleep Med. 2005; 6(1): 15–22. PubMed Abstract \| Publisher Full Text\n\nScharf M, Erman M, Rosenberg R, et al.: A 2-week efficacy and safety study of eszopiclone in elderly patients with primary insomnia. Sleep. 2005; 28(6): 720–7. PubMed Abstract\n\nBrasure M, MacDonald R, Fuchs E, et al.: Management of Insomnia Disorder [Internet]. AHRQ Publication No. 15(16)-EHC027-EF, Accessed 12-30-2015. Rockville, MD, USA, Agency for Healthcare Research and Quality. Comparative Effectiveness Review. 1–288. PubMed Abstract\n\nGiles J: Drug trials: stacking the deck. Nature. 2006; 440(7082): 270–2. PubMed Abstract \| Publisher Full Text\n\nMattila T, Stoyanova V, Elferink A, et al.: Insomnia medication: do published studies reflect the complete picture of efficacy and safety? Eur Neuropsychopharmacol. 2011; 21(7): 500–7. PubMed Abstract \| Publisher Full Text\n\nBoyle J, Groeger JA, Paska W, et al.: A method to assess the dissipation of the [corrected] residual effects of [corrected] hypnotics: eszopiclone versus zopiclone. J Clin Psychopharmacol. 2012; 32(5): 704–9. PubMed Abstract \| Publisher Full Text\n\nKripke DF: I petitioned the FDA to restrict hypnotics: here is why. Sleep Med. in press, 2015. pii: S1389-9457(15)02088-2. PubMed Abstract \| Publisher Full Text\n\nMellinger GD, Balter MB, Uhlenhuth EH: Insomnia and its treatment. Prevalence and correlates. Arch Gen Psychiatry. 1985; 42(3): 225–32. PubMed Abstract \| Publisher Full Text\n\nBertisch SM, Herzig SJ, Winkelman JW, et al.: National use of prescription medications for insomnia: NHANES 1999-2010. Sleep. 2014; 37(2): 343–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZammit GK, McNabb LJ, Caron J, et al.: Efficacy and safety of eszopiclone across 6-weeks of treatment for primary insomnia. Curr Med Res Opin. 2004; 20(12): 1979–91. PubMed Abstract \| Publisher Full Text\n\nSchutte-Rodin S, Broch L, Buysse D, et al.: Clinical guideline for the evaluation and management of chronic insomnia in adults. J Clin Sleep Med. 2008; 4(5): 487–504. PubMed Abstract \| Free Full Text\n\nCarter BD, Abnet CC, Feskanich D, et al.: Smoking and mortality--beyond established causes. N Engl J Med. 2015; 372(7): 631–40. PubMed Abstract \| Publisher Full Text\n\nTannenbaum C, Diaby V, Singh D, et al.: Sedative-hypnotic medicines and falls in community-dwelling older adults: a cost-effectiveness (decision-tree) analysis from a US Medicare perspective. Drugs Aging. 2015; 32(4): 305–14. PubMed Abstract \| Publisher Full Text\n\nKripke DF: When our body clocks run late: Does it make us depressed? Ann Transl Med. In press, 2016. Reference Source\n\nQaseem A, Kansagara D, Forciea MA, et al.: Management of Chronic Insomnia Disorder in Adults: A Clinical Practice Guideline From the American College of Physicians. Ann Intern Med. in press, 2016; 165. PubMed Abstract \| Publisher Full Text" }	[ { "id": "13914", "date": "23 May 2016", "name": "Jerome M. Siegel", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis paper pulls together evidence from Dr. Kripke’s own work and subsequent work, which indicate that the use of benzodiazepines and perhaps other sleeping pills is causing thousands, perhaps hundreds of thousands of deaths annually in the United States. He reviews the complete lack of evidence for any positive health effect of the use of these drugs. This is especially striking because drug companies sponsor a considerable amount of research on their sleeping pills and would undoubtedly publicize any data indicating positive health or lifespan effects – but there do not appear to be any. Kripke also points out the effectiveness of cognitive behavioral therapy for insomnia. This well studied treatment is less expensive, without any known deleterious effects on lifespan or health and produces a long-lasting reduction in insomnia. The effectiveness of cognitive behavioral therapy in the treatment of insomnia contrasts with the miniscule (0-20 min) increase in sleep time produced by sleeping pills, followed by a considerable withdrawal effect if the patient stops taking the pills.\n\nMinor suggestions include the following:\nI would delete the discussion of Judge Scalia’s death. Although it gets one’s attention, without knowing what Scalia was taking and without any documentation of the cause of death, it does more harm than good to the impact of the paper.\n\nOn page 10, I would delete the paragraph on prescriptions without valid clinical indication. I guess that in many cases the prescribing physician would just say he forgot to document the need. I do not doubt that Dr. Kripke is bringing attention to an important issue, but it is not persuasively presented, in contrast to the rest of his argument.\n\nSmall typo under “Obesity and aging exacerbate hypnotic risks:” “can be attributed overdoses “", "responses": [ { "c_id": "2001", "date": "31 May 2016", "name": "Daniel F. Kripke", "role": "Author Response", "response": "The kind reviews and useful contributions from Dr. Siegel and Dr. Phillips are much appreciated. Regarding the paragraph about the death certificate of Justice Scalia, debate about his cause of death and concern about the lack of autopsy received considerable press attention in major media in the United States. The process by which a rural judge decided what cause of death to record on the death certificate was uniquely well documented by the press. This illustrated how a death that could have been due to an overdose might not be explored and the overdose possibility might not be recorded. Knowing what the patient’s primary doctor had or had not prescribed would not resolve the issue of what drugs were or were not taken. This paragraph was intended to exemplify how we may indeed lack adequate documentation of the real cause of death when the plausible possibility of death caused by a hypnotic is not acknowledged on a death certificate. Regarding hypnotic prescriptions without a recorded diagnosis of insomnia, indeed the prescribing physician might just say that forgetting to document the insomnia was an oversight, but it is implausible that oversight is the explanation for such a large percentage of total hypnotic prescriptions. If lack of indication is usually an oversight, where is the proof? When I was a medical student in the 1960’s, I was trained that a hypnotic drug should be part of preprinted routine admission orders, and I have verified that routine admission orders for hypnotics are still preprinted in distinguished academic training hospitals in 2016. If we are training young doctors to prescribe a hypnotic without asking the patient whether that patient is experiencing trouble sleeping and without weighing the benefits and risks for the individual, it is plausible that habit persists in primary care. My impression is that prescribing doctors often do not ascertain that the patient has diagnostic criteria for insomnia, and in many cases, physicians know that the patient has no trouble sleeping. The physician might be trying to treat depression or to supplement opiates, but both uses are contraindicated. The physician might be treating some condition further afield such as hypertension or might be intentionally trying to achieve a placebo effect. There were studies documenting such practices several decades ago (references 126 and 136), but I know of no adequate study of 21st century U.S. outpatient hypnotic prescribing intentions. The manufacturers of both zolpidem and suvorexant have informed the FDA that the currently-FDA-mandated recommended doses are ineffective. It is tempting to infer that the FDA countenances the use of ineffective hypnotic doses as placebo implements of the bedside manner, without evidence that benefits outweigh risks of potentially addicting or lethal placeboes." } ] }, { "id": "14033", "date": "27 May 2016", "name": "Barbara A. Phillips", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis report is important and over due, and most likely would not be published in a journal that accepts advertising from pharmaceutical companies. The author is careful NOT to confuse association with causation. References are complete and up-to-date.", "responses": [] } ]	1	https://f1000research.com/articles/5-918
https://f1000research.com/articles/7-1770/v1	09 Nov 18	{ "type": "Research Article", "title": "Analysis of survival for patients in relation to central venous catheter and nosocomial blood stream infections: A case study of Aga Khan University Hospital, Nairobi", "authors": [ "Francis Maina Kiroro", "Majid Twahir", "Majid Twahir" ], "abstract": "Background: This study was focussed on survival rates of patients admitted to acute care units who utilized medical devices known as central venous catheters (CVC). CVCs are useful devices in clinical care; however some infections such as central line associated bloodstream infections (CLABSI) may occur, which are associated with increased lengths of stay and costs as well as higher morbidity and mortality rates. The overall objective of the present study was to determine survival probabilities and hazard rates for patients who used CVC devices and compare the subgroups by infection status. Methods: The study was focused on all patients who were admitted to Critical Care Units between 8th December 2012 and 31st March 2016 and utilized CVC devices. It was a retrospective study. Survival analysis techniques, test of equality of proportions, Man-Whitney test and Chi–square test of independence were used. Results: A total of 363 out of 1089 patients included in the study died during hospitalization. 47 patients developed nosocomial CLABSI. The average duration was 18.19 days and median of 12 days for hospitalized patients who did not develop a nosocomial CLABSI compared to an average of 56.79 days and a median of 51 days for those who did. There was a significantly higher proportion of mortality of those who developed nosocomial CLABSI compared to those that didn’t (p-value=0.01379). The results indicate that there was a significant association between infection status and discharge status, and significant difference to the survival rates of the patients based on infection status. Conclusions: There is a significant impact on mortality and morbidity of patients who develop nosocomial CLABSI. The duration of hospitalization by patients who developed CLABSI was significantly higher compared to patients who did not. Increased length of stay leads to higher cost of hospitalization.", "keywords": [ "Critical care unit", "central venous catheter", "length of stay", "nosocomial infection", "survival time", "survival analysis", "survival function", "central line associated blood stream infection", "CLABSI" ], "content": "Introduction\n\nCentral line associated bloodstream infection (CLABSI) is a type of infection that affects patients who utilize central venous catheter (CVC) during their hospitalization. CVC refers to any central venous access device inserted into the internal jugular, subclavian or femoral vein that terminates in the inferior vena cava or right atrium1. CVCs are commonly used in wards and critical care units, such as the intensive care unit (ICU), and they are also referred to as central lines. In accordance with The Joint Commission (2012), CVCs are essential in health care provision as they facilitate administration of medications, intravenous fluids, and hemodialysis among other functions. CVCs are used both in the in-patient and out-patient clinical care management2.\n\nAccording to The Joint Commission (2012), there are a variety of CVCs available in various sizes as well as different catheter materials, for example CVCs can be single or multi-lumen (double, triple or quadruple lumen). Another categorization by design classifies them as tunnelled catheters, non-tunnelled catheters, peripherally inserted central catheters and implantable ports2. The choice of catheter is as a result of defined need and preferences of the clinical care giver or the patient. Every catheter device carries with it some risk of infection, however, the extent of risk depends on the type of catheter used3.\n\nHealthcare associated infection (HAI) refers to infections that occur in the course of healthcare management in any setting, more specifically, the infections acquired during hospitalization are referred to as nosocomial infections2,4. US Centers for Disease Control and Prevention (2016) observed that there was ~46% decrease in CLABSIs in hospitals across the U.S.A from year 2008 to 2013, whilst about 30,100 still occurred in critical care units and wards5. There was a declaration by the European Union that policy on HAI prevention be prioritized in 20086.\n\nSeveral studies have linked CLABSIs to prolonged hospitalization and increased health management cost, as well as increased morbidity and mortality2,7–12 and 13. Morgan et al. (2010) conducted a five year study, in which the authors established that HAIs, especially CLABSIs, contributed to about one third of unexpected in-hospital mortality13. Soufir et al. (1999) conducted a prospective, matched cohort study from 1st January 1990 to 31st December 1995, in two ICUs in Paris, France and deduced that for ICU survival rates, the risk of death was significantly increased in patients with catheter related septicaemia as compared to control patients who did not develop septicaemia, with a relative risk of 2.06 (95% CI, 1.16-3.68; p-value <0.05)8. Garnacho-Montero et al. (2008) found that the median time from insertion of the catheter to the development of bacteraemia was 10 days. Further, 18 out of 66 patients died in the ICU (27.3%)10.\n\nHowever, various studies face a myriad of challenges which may be attributed to difficulties in estimating the mortality resulting from blood stream infections8,14 and 15. De Angelis et al. (2010) observed that when the duration of hospitalization of patients is increased, chances of utilizing invasive catheter devices increases leading to likelihood of infection14. Umscheid et al. (2011) observed that majority of researchers have not been able to relate CLABSIs autonomously with increase in mortality due to multiple causes of patient mortality such that exclusive impact of an infection may not be explicit16. Carrico and Ramírez (2007) observed that it may not be easy to determine the patients who die “with” CLABSI compared to those who die “because of” CLABSI15.\n\nVery few studies have been carried out on CVC utilization or its relationship with CLABSI, in particular in developing countries. Methodologies utilised in some published studies are varied, with some facing validity problems, necessitating the need for more studies. Hospitals follow a distinct culture of patient safety programmes, which among other factors affect outcomes of hospitalized patients in different ways. Globally, the risk associated with nosocomial CLABSI is dire, and even though there has been application of various strategies to reduce the impact on patients, it remains a crucial problem.\n\nThis study was intended to provide results that explain survival for patients who used CVC devices and developed CLABSI during their duration of hospitalization and compare with those who do not develop the infection in Kenya. The conceptual framework as adopted from De Angelis et al., 201014 can be viewed in Figure 1.\n\nThe first state: hospital critical care admission refers to the date when patients who utilized CVC devices during hospitalization were admitted. The second state: nosocomial infection, captures the date the nosocomial CLABSI was detected. Discharge/transfer and death, refers to the date the admitted patients exited the hospital14.\n\nThe overall aim was to determine the survival probabilities and hazard rates for patients who used CVC and establish factors affecting their survival and assess how they differ with the subset of the population that develops CLABSI. The specific objectives were as follows: i) to assess the survival probabilities for persons who utilized CVC during hospitalization; ii) to compare the survival time of patients’ group with CLABSI and the group not infected; ii) to determine whether there is an association between infection status and mortality; iii) to determine whether mortality is significantly different between the two groups and to determine factors that affect survival of patients with CLABSI.\n\n\nMethods\n\nThe study was conducted at critical care units of Aga Khan University Hospital, Nairobi, Kenya namely 11-bed medical-surgical ICU, 6-bed coronary care unit, 4-bed cardiothoracic intensity care unit and 16-bed high dependency unit. The focus was on acute care admitted patients, who utilized CVC access devices between 8th December, 2012 and 31st March, 2016. The data was obtained retrospectively and included the following variables: patient’s age, date of admission and discharge, discharge status, count of devices utilized, type of CVC device used, gender, infection status, event status and survival time in days. Data were obtained from the hospital’s electronic database and infection control surveillance datasheets. Infection control team of specialized nurses were involved in data collection. A CLABSI was considered nosocomial if a recognized pathogen was isolated from one or more percutaneous blood cultures after 48hrs of vascular catheterization and was confirmed to be unrelated to an infection from a different site. Determination of the source of infection was determined by both intensivists and microbiologists through clinical evaluation.\n\nData were extracted and exported into MS Excel 2010 where aggregation and organization was initially carried out. Data were then exported into Statistical Package for the Social Sciences (SPSS v.24), R GUI (R-3.1.1) and Stata (SE 11.1) for further data management, cleaning and analysis. Each statistical program was utilized to run suitable analyses as per the objectives of the study as well as for attainment of desired outputs.\n\nBoth exploratory and inferential analyses were utilized. Kaplan Meier curves were used to assess the survival probabilities of exposed versus un-exposed. Log rank, Breslow, Tarone- Ware, Peto and Flemington- Harrington tests helped in testing whether the survival curves between groups differed significantly. Additional statistical tests utilized were Chi-Square test of association and test of equality of proportions, as well as Man-Whitney test.\n\nSurvival analysis is a group of statistical techniques used in analysis whereby outcome variable of interest is time taken until an event occurs. In this study the event of interest was death, which occurred among hospitalized patients who had utilized central line devices. In survival analysis time variable is referred to as survival time, in this study, the length of stay. Censoring was done when a patient did not die by the end of hospitalization, meaning he/she was either discharged alive or transferred to another facility (lost for follow-up). Censoring assumptions provided that there was independence, randomness and non-informative censoring17. In particular, right censoring was utilized.\n\nLet T denote the random variable representing survival time where T≥0. The distribution of survival times is characterized by any of three functions: the survival function (S(t)), the probability density (f(t)) or the hazard function (h(t)). Equation (1) shows an expression of a survival function\n\n\n\nThe hazard function, h(t), presents instantaneous rate upon which events occur, provided no such previous events. The hazard function h(t) is portrayed in Equation (2), it can also be referred to as a conditional failure rate.\n\n\n\nHazard rate is used in providing information regarding conditional failures, in model identification and as a basis of expressing survival analysis math models.\n\nIts cumulative function (Equation (3)) produces accumulated risk up to time t,\n\n\n\nKaplan Meier function enables approximation of survival probabilities by use of the product limit formula. Kaplan Meier product limit formula is given by Equation (4)\n\n\n\nThe Log Rank Test is used to compare two or more Kaplan Meier Curves. The test is approximately Chi-Square particularly for large samples with G-1 degrees of freedom (df), where G is the number of groups involved as provided in Equation (5).\n\n\n\nLog rank is used to test the null hypothesis that there is no significant difference between two survival curves. Similarly the alternative tests to Log rank are Breslow (Wilcoxon) and the Tarone-Ware; these tests apply different weights to ith failure time.\n\nBreslow (Wilcoxon) test weights the observed minus expected score at time ti by the number at risk ni over all the groups at time ti. In this test the weights subjected at the earlier failure times are higher as compared to later failure times17.\n\nThe test statistic is as shown in Equation (6)\n\n\n\nTarone-Ware test statistic applies more weight to the early failure times by weighting the observed minus expected score at time t(i) by ni , where ni refers to the number at risk.\n\nFlemington-Harrington test uses the KM survival estimates S^(t) for entire data to determine its weights for the ith failure time. Different choices of values for p and q can be done. Weights are computed as shown in Equation (7) whereas the test statistic is similar to Equation (6).\n\n\n\nWe consider the difference in the two proportions to be given by d=x1n1−x2n2 , which is approximately normally distributed with mean zero and\n\n\n\nif the counts are binomially distributed with the same parameter. The null hypothesis is such that P1= P2, the common estimate for the proportion p⌢=x1+x2n1+n2\n\nChi Square test of independence is used to determine whether there is a significant association between two categorical variables. Chi square test statistic is as portrayed in Equation (9)\n\n\n\n\nResults\n\nA total of 1086 patients were included in the study; males 648 (59.7%). 363 patients experienced the event of interest (death). 47 patients developed nosocomial CLABSI during their hospitalization. Table 1 summarised the participant’s demographics.\n\nTable 2 demonstrates survival time summary statistics disaggregated by infection status. Time was measured in terms of days. The average duration of 18.19 days (standard error (s.e) 0.611) was taken by patients who did not develop a nosocomial CLABSI compared to an average of 56.79 days (s.e of 5.171) taken by those who got an infection. Median days taken by the infected group were 51 days whereas those taken by non-infected group were 12 days.\n\nTest of association between infection status and event (discharge) status depict a χ2 1=6.868, p-value=0.009, which is significant at 5% level; therefore, there is a significant statistical association between the infection status and the event (discharge) status. Test of association between gender and event (discharge) status depict a χ2 1=0.705, p-value=0.401, which is not significant at 5% level; hence there is no significant association between the gender of the patient and the discharge (event) status at 5% level (Table 3). Test of association between the gender of the patient and the infection status depict a χ21= 1.446, p-value=0.229, which is not significant at 5% level; hence there is no significant association between the patient’s gender and the infection status at 5% level (Table 3).\n\ndf, degrees of freedom.\n\nTest of equality of proportions of death between CLABSI infected and group not infected by CLABSI produced a P1=0.511 and P2=0.326 with χ21=6.0647, p-value=0.014, the 95% CI [0.02752, 0.34121]. This indicates that the proportions between the two groups are significantly different at 5% level. Hence, nosocomial CLABSI subjects a patient to a higher mortality rate as compared to a patient who does not get the infection during hospitalization.\n\nSurvival probabilities were assessed graphically as well by use of survival tables. Graphic assessment was done by use of Kaplan Meier (KM) curves for all subjects (see Figure 2), in the initial period of about 50 days after admission, survival probabilities decline at relatively more close ranges as compared to the period after 50 days. The overall mean time estimate was 70.72 days (95% CI; 60.362, 81.084) and the overall median time estimate was 44 days (95% CI; 36.49, 51.51). Further, graphical analysis by use of the KM curves was done for each group (infected vs not infected). As shown in Figure 3, the initial stages indicate higher probabilities of survival among the patients with CLABSI as compared to the group of patients with no CLABSI. However, this trend changes after about 113 days where the survival probabilities of the group not infected are higher. At about 140 days of admission, the survival probabilities of infected group decline sharply. The small sample of the patients with nosocomial CLABSI compared to the rest (with no CLABSI) may have contributed to this pattern. The median number of days estimates for the group which did not have nosocomial CLABSI were 43 days (95% CI; 34.911, 51.089) whereas for the group whose patients developed nosocomial CLABSI was 76 days (95% CI; 36.836, 115.164) (see Figure 2). On the other hand, the average number of days estimates were 76.376 days (95% CI; 63.610, 89.141) and 83.807days (95% CI; 67.382, 100.232) for non-infected and infected groups respectively.\n\nTests on the survival curves Figure 2 were Log Rank (Mantel-Cox), Breslow (Generalized Wilcoxon), Tarone-Ware, Peto and Fleming-Harrington. Results from all the five tests show that the survival curves are significantly different at 5% level as provided in Table 4.\n\nTest on survival time in relation to the infection status was conducted using a two sample Wilcoxon rank sum test (Mann–Whitney test) which is a distribution free/non parametric method. The value of test statistic was 6112.5 and a corresponding p-value < 2.2e-16, which indicates that it is highly significant at 5% level. We can hence deduce that there is a significant difference between the length of stay by the patients who develop nosocomial CLABSI compared to the patients who do not develop the infection. Thus the patients who develop nosocomial CLABSI stay longer in hospital.\n\nStratification by gender yielded the following survival curves for male and female, portrayed in Figure 3. The survival probabilities for the female subgroup were higher at the initial phase of about 40 days after admission after which the male’s survival rates remained higher until the end. The overall average duration by male patients was 73.69 days (95% CI; 60.970, 86.958), whereas that of female patients was 59.9 (95% CI, 60.970, 86.958). The average duration taken by the male patients who were infected by CLABSI was 94.324 days (95% CI; 73.6, 115.0), whereas that of female patients was 59.904 days (95% CI; 48.7, 71.1). Median was 46 days (95% CI; 34.8, 57.2) for male patients and 42 days (95% CI; 33.5, 50.5) for female patients.\n\nTests on survival curves by infection status after adjusting for gender using the three tests displayed on Table 5 indicate that the survival curves were all significantly different at 5% level.\n\nTests on the survival curves with respect to gender adjusted for infection status; initially both male and female patients indicate similar survival rates from the beginning of the admission period up to about the tenth day of admission. After approximately the tenth day the female subgroup seem to portray higher survival rates up to about 45th day of admission. Stratification by patients who developed nosocomial CLABSI indicates that initially both male and female experienced similar survival rates up to approximately 6 days where the survival rates for male remain higher (see Figure 3). Three pair wise tests were conducted on survival curves with respect to gender after adjusting for infection status. Results indicate that all the three tests were not significant at 5% level.\n\nThis section portrays the survival tables for all the patients studied as well as for specific subgroups. The survival table for all the patients who were admitted into the critical care units and utilized CVCs, as provided in Table 6.\n\nNote: survivor function is calculated over full data and evaluated at indicated times; it is not calculated from aggregates shown at left.\n\nSurvival table for the group who were infected by CLABSI is provided in Table 7.\n\nNote: survivor function is calculated over full data and evaluated at indicated times; it is not calculated from aggregates shown at left.\n\nThe survival table for the patient group who did not get infected by CLABSI is shown in Table 8.\n\nNote: survivor function is calculated over full data and evaluated at indicated times; it is not calculated from aggregates shown at left.\n\nAssessment of the proportional hazards (PH) assumptions done using log minus log plots, observed versus expected and goodness of fit test. Assessment of the log minus log curves by infection status indicated that they were not parallel hence violating the PH assumption. Secondly, using Observed versus Expected plots; evaluated by comparing observed (KM survival estimates) versus expected (Cox adjusted) survival curve estimates plotted on the same graph with an anticipation that they would be as close to each other as possible. Results depicted a divergence between the two curves, hence violating the assumptions of PH. Lastly, using the goodness of fit test, the results depict that only the infection status doesn’t violate the PH assumption. Age, gender and number of the devices used had p-values < 0.05 indicating that violation of the PH assumption (Table 9).\n\n\nDiscussion\n\nBased on the results, we deduce that there is a significant difference between the lengths of stay by the patients who develop nosocomial CLABSI compared to that of patients who do not develop the infection. The patients who develop nosocomial CLABSI recorded a longer duration of stay in the hospital this in agreement with other studies conducted by14,2,7–10 and 11.\n\nThe proportion of patients who died after developing the nosocomial CLABSI compared to the proportion of those who died having not developed the infection revealed that there is a significant difference in the two proportions at 5% level (χ2 1=6.0647, p-value=0.01379). Consequently, indicating that mortality as well as morbidity is significantly increased when a patient develops the nosocomial CLABSI, similar studies by 2,7–1014 and 11 made a similar conclusion.\n\nIn the index study, the survival curves between those who developed an infection and those that didn’t were different. This difference was not due to gender or the discharge status (alive/dead) of the patient. To the best of our knowledge no such findings have been documented in similar studies.\n\nThe data having two subgroups that were considered in the study were disproportionate; the group (with the nosocomial CLABSI) had a very small proportion of 4.3% of the total number patients in the study compared to the proportion of the group that did not have the CLABSI infection during hospitalization. These disproportionate groups may not be very useful in survival model fitting. Searching for published journals from studies carried out in developing countries regarding survival analysis in relation to the use of CVC devices were not available.\n\n\nConclusions\n\nThe results reveal that there was a significant association between the nosocomial CLABSI infection status and the discharge status as well as the patient’s length of stay. In addition, there is a difference between the survival rates of the patients who developed nosocomial CLABSI compared to those who did not. Mortality was higher among patients who developed nosocomial CLABSI compared to those who did not during their hospitalization. The duration of hospitalization (length of stay) by the patients who developed CLABSI was significantly higher compared with the duration taken (length of stay) by patients who did not develop CLABSI, which has a great impact on the financial burden the patients are subjected to due to added hospital bed days as well as medication administered.\n\nSince CLABSI infections have been found to elongate patient’s length of stay, appropriate strategies such as implementation of and adherence to the CVC insertion and maintenance bundles would be ideal in order to reduce the infection rates. Further matched studies in relation to CVC utilization along specific age groups and in specific diagnoses is recommended. More survival analysis research is needed in developing countries in regard to CLABSI as well as in utilization of the CVCs.\n\nAuthority to conduct the study was issued by the Ethics and Research Committee (ERC) of the participating hospital. Ethical review was conducted by the Kenyatta National Hospital/ University of Nairobi ERC, who provided approval (Ref: KNH-ERC/A/381) and an affiliation letter. Participant consent was waived since it was a retrospective study. Privacy and confidentiality was maintained by ensuring the information gathered was not relayed to anyone, but used for this study only. Patients’ names were not included in the data collected and only identification numbers utilized. No risks were subjected to the patients. Direct benefit was not intended for the study population; however the results are useful in terms of adding knowledge to the existing research.\n\nRaw data were gathered electronically and the computer used in data analysis was personalized with limited access through password protection. No public computer was utilized in data aggregation or analysis.\n\n\nData availability\n\nF1000Research: Dataset 1. All raw data obtained in the present study. Columns A-K correspond with information contained in the sheet titled Variable_Info., https://dx.doi.org/10.5256/f1000research.16819.d22354744", "appendix": "Grant information\n\nThe authors declare that no grant(s) were involved in supporting this work.\n\n\nReferences\n\nChopra V, O'Horo JC, Rogers MA, et al.: The risk of bloodstream infection associated with peripherally inserted central catheters compared with central venous catheters in adults: a systematic review and meta-analysis. Infect Control Hosp Epidemiol. 2013; 34(9): 908–918. PubMed Abstract \| Publisher Full Text\n\nThe Joint Commission: Preventing Central Line–Associated Bloodstream Infections: A Global Challenge, a Global Perspective. 2012; [Accessed 01 December 2015]. Reference Source\n\nMaki DG, Kluger DM, Crnich CJ: The risk of bloodstream infection in adults with different intravascular devices: a systematic review of 200 published prospective studies. Mayo Clinic Proc. 2006; 81(9): 1159–1171. PubMed Abstract \| Publisher Full Text\n\nSiegel JD, Rhinehart E, Jackson M, et al.: 2007 Guideline for Isolation Precautions. 2007; [Accessed 01 December 2015]. Reference Source\n\nN. H. S. N. US Centers for Disease Control and Prevention: Device-Associated (DA) Module Protocol and Instructions: Central Line–Associated Bloodstream Infection (CLABSI) Event. CDC, 2016; [Accessed 01 January 2016]. Reference Source\n\nEuropean Commission: European Comission. 2008.\n\nUS Centers for Disease Control and Prevention: The Direct Medical Costs of Healthcare-Associated Infections in US Hospitals and the Benefits of Prevention. 2009; [Accessed 31 December 2015]. Reference Source\n\nSoufir L, Timsit JF, Mahe C, et al.: Attributable morbidity and mortality of catheter-related septicemia in critically ill patients: a matched, risk-adjusted, cohort study. Infect Control Hosp Epidemiol. 1999; 20(6): 396–401. PubMed Abstract \| Publisher Full Text\n\nBlomberg B, Manji KP, Urassa WK, et al.: Antimicrobial resistance predicts death in Tanzanian children with bloodstream infections: a prospective cohort study. BMC Infect Dis. 2007; 7: 43. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGarnacho-Montero J, Gutiérrez-Pizarraya A, Escoresca-Ortega A, et al.: Adequate antibiotic therapy prior to ICU admission in patients with severe sepsis and septic shock reduces hospital mortality. Crit Care. 2015; 19(1): 302. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPittet D, Tarara D, Wenzel RP: Nosocomial bloodstream infection in critically ill patients. Excess length of stay, extra costs, and attributable mortality. JAMA. 1994; 271(20): 1598–1601. PubMed Abstract \| Publisher Full Text\n\nRasslan O, Seliem ZS, Ghazi IA, et al.: Device-associated infection rates in adult and pediatric intensive care units of hospitals in Egypt. International Nosocomial Infection Control Consortium (INICC) findings. J Infect Public Health. 2012; 5(6): 394–402. PubMed Abstract \| Publisher Full Text\n\nMorgan DJ, Lomotan LL, Agnes K, et al.: Characteristics of healthcare-associated infections contributing to unexpected in-hospital deaths. Infect Control Hosp Epidemiol. 2010; 31(8): 864–866. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDe Angelis G, Murthy A, Beyersmann J, et al.: Estimating the impact of healthcare-associated infections on length of stay and costs. Clin Microbiol Infect. 2010; 16(12): 1729–1735. PubMed Abstract \| Publisher Full Text\n\nCarrico R, Ramírez J: A process for analysis of sentinel events due to health care-associated infection. Am J Infect Control. 2007; 35(8): 501–507. PubMed Abstract \| Publisher Full Text\n\nUmscheid CA, Mitchell MD, Doshi JA, et al.: Estimating the proportion of healthcare-associated infections that are reasonably preventable and the related mortality and costs. Infect Control Hosp Epidemiol. 2011; 32(2): 101–114. PubMed Abstract \| Publisher Full Text\n\nKleinbaum DG, Klein M: Survival Analysis: A Self Learning Text, Third Edition. ed. K. K. J. S. A. T. W. W. M. Gail, Ed., Atlanta: Springer, 2012. Publisher Full Text\n\nIwamoto P: Aseptic technique. APIC Text of Infection Control and Epidemiology. 2009; 20–21.\n\nWenzel RP, Gennings C: Bloodstream infections due to Candida species in the intensive care unit: identifying especially high-risk patients to determine prevention strategies. Clin Infect Dis. 2005; 41 Suppl 6: S389–393. PubMed Abstract \| Publisher Full Text\n\nZingg W, Cartier-Fässler V, Walder B: Central venous catheter-associated infections. Best Pract Res Clin Anaesthesiol. 2008; 22(3): 407–21. PubMed Abstract \| Publisher Full Text\n\nWylie MC, Graham DA, Potter-Bynoe G, et al.: Risk factors for central line-associated bloodstream infection in pediatric intensive care units. Infect Control Hosp Epidemiol. 2010; 31(10): 1049–1056. PubMed Abstract \| Publisher Full Text\n\nKritchevsky SB, Braun BI, Kusek L, et al.: The impact of hospital practice on central venous catheter associated bloodstream infection rates at the patient and unit level: a multicenter study. Am J Med Qual. 2008; 23(1): 24–38. PubMed Abstract \| Publisher Full Text\n\nSafdar N, Kluger DM, Maki DG: A review of risk factors for catheter-related bloodstream infection caused by percutaneously inserted, noncuffed central venous catheters: implications for preventive strategies. Medicine (Baltimore). 2002; 81(6): 466–479. PubMed Abstract\n\nAlmuneef MA, Memish ZA, Balkhy HH, et al.: Rate, risk factors and outcomes of catheter-related bloodstream infection in a paediatric intensive care unit in Saudi Arabia. J Hosp Infect. 2006; 62(2): 207–213. PubMed Abstract \| Publisher Full Text\n\nO'Grady NP, Alexander M, Burns LA, et al.: Guidelines for the prevention of intravascular catheter-related infections. Clin Infect Dis. 2011; 52(9): e162–193. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nParienti JJ, Thirion M, Mégarbane B, et al.: Femoral vs jugular venous catheterization and risk of nosocomial events in adults requiring acute renal replacement therapy: a randomized controlled trial. JAMA. 2008; 299(20): 2413–2422. PubMed Abstract \| Publisher Full Text\n\nDatta P, Rani H, Chauhan R, et al.: Health-care-associated infections: Risk factors and epidemiology from an intensive care unit in Northern India. Indian J Anaesth. 2014; 58(1): 30–35. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGonzales M, Rocher I, Fortin E, et al.: A survey of preventive measures used and their impact on central line-associated bloodstream infections (CLABSI) in intensive care units (SPIN-BACC). BMC Infect Dis. 2013; 13: 562. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTang HJ, Lin HL, Lin YH, et al.: The impact of central line insertion bundle on central line-associated bloodstream infection. BMC Infect Dis. 2014; 14: 356. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFuruya EY, Dick A, Perencevich EN, et al.: Central line bundle implementation in US intensive care units and impact on bloodstream infections. PLoS One. 2011; 6(1): e15452. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWenzel RP, Thompson RL, Landry SM, et al.: Hospital-acquired infections in intensive care unit patients: an overview with emphasis on epidemics. Infect Control. 1983; 4(5): 371–5. PubMed Abstract \| Publisher Full Text\n\nSan Miguel LG, Cobo J, Otheo E, et al.: Secular trends of candidemia in a large tertiary-care hospital from 1988 to 2000: emergence of Candida parapsilosis. Infect Control Hosp Epidemiol. 2005; 26(6): 548–552. PubMed Abstract \| Publisher Full Text\n\nAllegranzi B, Bagheri Nejad S, Combescure C, et al.: Burden of endemic health-care-associated infection in developing countries: systematic review and meta-analysis. Lancet. 2011; 377(9761): 228–241. PubMed Abstract \| Publisher Full Text\n\nLorente L, Henry C, Martín MM, et al.: Central venous catheter-related infection in a prospective and observational study of 2,595 catheters. Crit Care. 2005; 9(6): R631–R635. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAdvani S, Reich NG, Sengupta A, et al.: Central line-associated bloodstream infection in hospitalized children with peripherally inserted central venous catheters: extending risk analyses outside the intensive care unit. Clin Infect Dis. 2011; 52(9): 1108–1115. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBlot SI, Vandewoude KH, Hoste EA, et al.: Effects of nosocomial candidemia on outcomes of critically ill patients. Am J Med. 2002; 113(6): 480–485. PubMed Abstract \| Publisher Full Text\n\nIP: Association for Professionals in Infection Control and Epidemiology. Aseptic technique. In Carrico R. 2009; 3: 20–21.\n\nDemissie MS: Investigating center effects in a multi-center clinical trial study using a parametric proportional hazards meta-analysis model. 2006. Reference Source\n\nAdvani S, Reich NG, Sengupta A, et al.: Central line-associated bloodstream infection in hospitalized children with peripherally inserted central venous catheters: extending risk analyses outside the intensive care unit. Clin Infect Dis. 2011; 52(9): 1108–1115. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nArthur G, Nduba VN, Kariuki SM, et al.: Trends in bloodstream infections among human immunodeficiency virus-infected adults admitted to a hospital in Nairobi, Kenya, during the last decade. Clin Infect Dis. 2001; 33: 248–256. PubMed Abstract \| Publisher Full Text\n\nHaddadi A, Lemdani M, Hubert H, et al.: Risk factors’ identification according to the development of Healthcare Associated Infections and mortality by using competing models at Timone University Hospital's ICU. Eur Sci J. 2014; 10(6): ISSN: 1857–7881. Reference Source\n\nResar R, Griffin FA, Haraden C, et al.: Using Care Bundles to Improve Health Care Quality. I. I. S. w. paper, Ed., Cambridge, Massachusetts: Institute for Healthcare Improvement, 2012. Reference Source\n\nKleinbaum DG, Klein M: Statistics for Biology and Health. 2nd Edition ed., K. K. J. S. A. T. W. W. M. Gail, Ed., New York: Springer Science+Business Media, Inc., 2005. Reference Source\n\nKiroro FM, Twahir M: Dataset 1 in: Analysis of survival for patients in relation to central venous catheter and nosocomial blood stream infections: A case study of Aga Khan University Hospital, Nairobi. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16819.d223547" }	[ { "id": "40519", "date": "23 Jan 2019", "name": "Gerald Mbuthia Mahuro", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe flow of the paper is good with applicable content and references to other literature. The format is well done and readable. Despite the paper informing us of what we know about extended length of stay and high probability of mortality for patients with nasocomial infections, it illuminates some statistics especially in Kenya which can be used as baseline. However, there seem to be four specific objectives for which the author has repeated (ii) twice. Kindly check the numbering and amend accordingly.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4383", "date": "29 Jan 2019", "name": "Francis Kiroro", "role": "Author Response", "response": "Thank you for taking time to review our paper. We will make an amendment to the numbering of the objectives." } ] }, { "id": "56618", "date": "18 Nov 2019", "name": "Jean-François Timsit", "expertise": [ "Reviewer Expertise nosocomial infections in ICU" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThank you for this interesting report about CLABSI in Kenya. Data are rare and this publication should be acknowledged.\nI have many advices to improve the format and the content of the paper:\nCLABSI definition: the definition that you used should be reported. Do you systematically performed blood culture? do you perform catheter tip culture? Recent syst review report very weak relationship between CLABSI and CRBSI and CLABSI is often a mix of CRBSI and primary bacteria with really opposite pathophysiology and impact on outcome...\n\nsee discussion on: de Grooth HJ et al. 20191; see also result from Adrie C et all.2.\n\nWhen catheter tip culture is systematically performed with the definition of CRBSI as reported by IDSA, the over risk of death is not so important. In another chapter you should explain the catheter infection policy used in your hospital, the number of single rooms the nurse to patient ratio the compliance to hands hygiene etc...\n\nDetails on the population and more clinical characteristics should be included: at least age, severity scores, rate of mechanical ventilation, cause of ICU admission. For the catheters: site of insertion duration of catheter insertion in days\n\nDetails on CLABSI: what are the micro-organisms recovered from blood cultures?\n\nDetails on therapy: do you systematically removed the catheter in case of positive BC? do you start antibiotic therapy?\n\nMethods: as far as I understand a cox model is used. (although the figure one revered to a progressive disability model and competing risks? Could you clarify? If it is a cox model the method section should be largely reduced. Number at risk, time perspective? Censors? pH assumptions. If a model for competing risk is used please clearly mention the different event and the way you take into account the competing event (discharge alive?). The use of a log rank test is preferable.\n\nReporting of the results: should include a table with main characteristics (see point 2 and 3); include one or 2 Kaplan Meier curves.. the table 6,7,8 should be removed. A sub analysis according to the microorganism responsible for the infection should be added.\n\nIn the discussion you should discuss your result in view of other data on prognosis of CRBSI and CLABSI.\n\nIs the work clearly and accurately presented and does it cite the current literature? No\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/7-1770
https://f1000research.com/articles/7-1768/v1	08 Nov 18	{ "type": "Research Article", "title": "Evaluation of postoperative pain in infected root canals after using double antibiotic paste versus calcium hydroxide as intra-canal medication: A randomized controlled trial", "authors": [ "Sarah Samir Abouelenien", "Salsaby Mohamed Ibrahim", "Olfat Gamil Shaker", "Geraldine Mohamed Ahmed", "Salsaby Mohamed Ibrahim", "Olfat Gamil Shaker", "Geraldine Mohamed Ahmed" ], "abstract": "Background: Postoperative pain is defined as pain of any degree after initiation of endodontic treatment either intra-appointment or post-obturation and is considered an undesirable occurrence for both patient and dentist. It was suggested that bacterial injury is probably the major cause of pain. Intra-canal medicaments are widely used to kill any bacteria surviving after instrumentation and irrigation. The aim of this study was to assess the ability of double antibiotic paste versus calcium hydroxide used as intra-canal medication in reducing postoperative pain. Methods: 36 patients with single rooted necrotic premolars with apical periodontitis were randomly assigned into two groups according to the intra-canal medication used: calcium hydroxide group (CH) and double antibiotic paste group (DAP). Preoperative pain was recorded using numerical rating scale. After isolation, access cavity was performed followed by chemico-mechanical preparation using rotary Race files with 2.5% sodium hypochlorite irrigation. Subsequently, intra-canal medication was placed and postoperative pain was recorded at 6, 12, 24 and 48 hours postoperatively. Results: There was no statistically significant difference between both groups. Both groups resulted in an increase in median pain value from preoperative to 6 hours postoperative, followed by gradual decrease from 6 hours to 12, 24, 48 hours postoperatively with statistically significant difference. When comparing both groups, DAP group showed lower postoperative pain values than CH group at 12 and 24 hours, but this was not statistically significant. Conclusion: The use of intra-canal medication in necrotic teeth with apical periodontitis was efficient in reducing postoperative pain regardless of type of intra-canal medication used. Trial registration: PACTR201605001482394 (Date: 22nd February 2016).", "keywords": [ "apical periodontitis", "calcium hydroxide", "antibiotic paste", "intra-canal medication", "postoperative pain" ], "content": "Introduction\n\nPain is the main symptom in many medical and dental conditions that can significantly alter the person’s general functioning1. Post-endodontic pain remains a serious problem facing the dental profession. It may occur as a result of several factors, including chemical, mechanical or bacterial irritation to the periapical tissues2. It has been suggested that the presence of bacteria as a result of failure to properly disinfect the root canal is the major cause of pain3. Many studies have found a direct relationship between root canal infection and the level of pain after endodontic treatment4. Thus, endodontic treatment is primarily focused on utmost elimination of these bacteria. Antibacterial intra-canal medicaments have been advocated in efforts to kill any bacteria surviving after chemico-mechanical preparation.\n\nSeveral intra-canal medications have been used, including calcium hydroxide which is the most widely used intra-canal medication. It is characterized by its initial bactericidal, bacteriostatic effect and its high pH5. It has been suggested that calcium hydroxide has pain preventive characteristics due to its antibacterial action or tissue altering effects. Moreover, it controls inflammation and has repair potential6. However, there is no clear evidence of its effect on post-endodontic pain7. Other studies found that calcium hydroxide was not very effective in decreasing post-treatment pain when used alone8,9.\n\nDouble antibiotic paste is a mixture of metronidazole and ciprofloxacin, which is used as intra-canal medication for disinfection of necrotic teeth. It has been reported to be effective at reducing bacterial counts in the infected root canals10. When commonly used medicaments fail in eliminating the symptoms, then antibiotic paste can be used for treatment of teeth with large periapical lesions11.\n\nThe objectives of the present study were to test the null hypothesis of whether the use of double antibiotic paste will be more effective in reduction of pain than calcium hydroxide, or if this paste will have disadvantages and complaints more than the traditional method used.\n\n\nMethods\n\nThe article has been written in concordance with CONSORT guidelines (Supplementary File 1).\n\nThe study design is a parallel randomized controlled trial with allocation ratio 1:1 and a Superiority Framework. The study was conducted in the duration between March 2016 and March 2017. The protocol of this study was approved by the Ethics Committee of the Faculty of Oral and Dental Medicine, Cairo University, Egypt (approval number 15918). The nature of this study, and associated risks were fully explained to the patients and informed consent forms were signed before initial treatment.\n\nBased on a previous study by Ehrmann et al. (2003), a medium effect size of approximately 0.25 is expected. A total sample size of 24 patients will be sufficient to detect an effect size of 0.25, a power of 80%, and a significance level of 5%. This number has been increased to a total sample size of 28, to adjust for using nonparametric tests. The number is again increased to a total sample size of 36 (18 in each of the two groups) to allow for losses of around 25%. Sample size was calculated using G*Power program (University of Düsseldorf, Düsseldorf, Germany). Minimal clinical difference 3.5 and 4.5 at 24 hrs and 48 hrs respectively.\n\nIn total, 36 participants were included in the study with maxillary or mandibular non-vital single rooted premolar teeth showing tenderness to percussion.\n\nExclusion criteria: pregnant women, patients who had received antibiotic treatment during the last 3 months, patients having more than one tooth requiring root canal treatment, teeth with periodontal probing depth >4mm, teeth that couldn't be isolated with a rubber dam, teeth with previous root canal treatment, or teeth with fluctuant facial swelling where emergency management should include incision and drainage.\n\nPatients were randomly assigned into two groups (n=18/group) according to the intra-canal medication used, using a computer generated random number table at the Center of Evidence-based Dentistry (EBD), Cairo University. The assistant supervisor generated the random sequence and assigned the participants to the intervention or control groups.\n\nExperimental group: double antibiotic paste (DAP); control group: calcium hydroxide paste (CH). All clinical procedures were performed by a single endodontist.\n\nAll patients accepted two-visit root canal treatment. Preoperative pain was recorded before any intervention.\n\nFirst visit: Teeth were anaesthetized using nerve block or infiltration technique by local anesthesia (Articaine HCI 4% & Adrenaline 1:100,000) and properly isolated with rubber dam. After access cavity preparation, the working length was checked using apex locator and confirmed by a radiograph. The root canals were prepared using rotary Race files (Fkg Dentaire, Switzerland) along with 2.5% sodium hypochlorite irrigation. The canals were dried and filled with intra-canal medication plugged into the canal by using Lentulo spiral according to the randomly selected group (DAP group - combination of 500 mg ciprofloxacin and 500 mg metronidazole ground then mixed with saline to obtain creamy consistency; CH group). The access cavities were properly filled with glass ionomer to ensure proper sealing with no leakage of any oral fluids inside the root canal, which might disturb the action of the intra-canal medication.\n\nPatients were instructed to record their postoperative pain level (see below) after 6, 12, 24 and 48 hours.\n\nSecond visit: After 7 days, all the patients returned the pain charts and the root canals were obturated using lateral condensation technique and resin-based root canal sealer.\n\nPrimary outcome measure is postoperative pain.\n\nA pain chart using numerical rating scale (NRS) was used to record the patients’ pain level. The NRS (0–10 scale) consists of a line anchored by two extremes “No pain” and “the worst pain”. Patients were asked to choose the mark that represented their level of pain from 0 to 10. Pain level was assigned as follow: 0, “no pain”; 1–3, “mild pain”; 4–6, “moderate pain”; 7-10, “severe pain”.\n\nThe study was double-blinded (the participants and the assessor). Participants did not know which group they were treated with; they chose a folded paper inside an opaque envelope which contained the numbers of each random sequence for both groups. The assessor, who assessed all outcome data, did not know which group the participants were related to.\n\nData was analyzed using IBM SPSS 21 (SPSS Inc., Chicago, IL, USA). The level of pain was described as median and range. Pain score was compared between the two groups using Mann-Whitney U test. Friedman test was performed to test the significance between the 4-time periods within each group. Chi square test was used to compare qualitative pain score. The level of significance was set at P < 0.05. All tests were two tailed.\n\n\nResults\n\nIn total, 36 patients were included in this study; Table 1 shows their baseline characteristics. A CONSORT flow diagram can be found in Supplementary File 2.\n\nns=non-significant.\n\nThe overall highest incidence of pain values in both groups was at 6 hours postoperatively, whereas the lowest incidence of pain values was at 48 hours postoperatively with no statistically significant difference between them. The DAP group showed lower postoperative pain values at 12 and 24 hours compared with the CH group, but with no statistical significance difference (Table 2).\n\nns = not significant.\n\n\nDiscussion\n\nEffective control over the intra-canal microbial load before root canal obturation is a key element that increases the success rate of endodontic treatment12. Necrotic pulp with periapical periodontitis was selected for assessment of postoperative pain, as the highest frequency of postoperative pain occurred in patients with necrotic pulp and apical periodontitis4,13 that harbored large number of bacteria due to its polymicobial nature as reported in previous studies14–18. In the present study, pain was recorded using a numerical rating scale (NRS), which is considered a consensus-based, standardized assessment method and reports better compliance compared to other scales19. The literature shows that the NRS provides sufficient distinctive power for chronic pain patients for describing their pain intensity20. As well, it was preferred by the majority of patients in different regions21,22. In the present study, the Rotary Race system was chosen for root canal preparation, owing to the fact that rotary NiTi instruments result in minimal debris extrusion compared to hand instrumentation23, which in turn would result in less postoperative pain. This was attributed to their rotary action, as well as copious irrigation associated with these instruments24.\n\nIn the present study, in patients with necrotic teeth with apical periodontitis, the use of double antibiotic paste showed no difference regarding postoperative pain compared to CH. DAP showed lesser postoperative pain at 12 and 24 hours than the CH group but this didn’t reach a level of statistical significance. This might be attributed to the combined spectrum of antimicrobial activity and synergetic or additive action of antibiotics found in DAP. In accordance with the results of this study, Pai et al.25 reported no statistically significant difference in postoperative pain between calcium hydroxide and triple antibiotic paste (TAP), although TAP was more effective that calcium hydroxide but this was not statistically insignificant. As reported in other studies7–9,26,27, there was no statistically significant difference in postoperative pain between calcium hydroxide paste and no intra-canal medication. In contrast, some studies counteracted our findings28,29, suggesting that CH had pain-preventive properties because of its antimicrobial as well as its tissue-altering effects. These properties could be referred to the chemical, physical, and antimicrobial effects of calcium hydroxide due to the diffusion of its hydroxyl (OH−) ions produced from ionization in aqueous solution, which leads to a highly alkaline environment. This is not conducive for the survival of microorganisms inside root canals3,30, and most of the microorganisms cannot sustain that high a pH (15.5).\n\n\nConclusion\n\nThe use of intra-canal medication for 7 days in necrotic teeth with apical periodontitis was efficient in reducing postoperative pain regardless of the type of intra-canal medication used.\n\n\nData availability\n\nF1000Research: Dataset 1. Datasheet containing patients age, gender, and pain score for all time points for both the CH and DAP groups., https://doi.org/10.5256/f1000research.16820.d22435931", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nSupplementary material\n\nSupplementary File 1: CONSORT checklist.\n\nClick here to access the data\n\nSupplementary File 2: CONSORT flow diagram.\n\nClick here to access the data\n\nSupplementary File 3: Trial protocol.\n\nClick here to access the data\n\n\nReferences\n\nMemon NA, Memon MR, Ali F: Assessment of the interappointment pain by using two different intracanal medicaments. Pakistan Oral Dent J. 2013; 33(1): 145–50. Reference Source\n\nSiqueira JF Jr, Rôças I, Favieri A, et al.: Incidence of postoperative pain after intracanal procedures based on an antimicrobial strategy. J Endod. 2002; 28(6): 457–60. PubMed Abstract \| Publisher Full Text\n\nSiqueira JF Jr, Lopes HP: Mechanisms of antimicrobial activity of calcium hydroxide: a critical review. Int Endod J. 1999; 32(5): 361–9. PubMed Abstract \| Publisher Full Text\n\nMor C, Rotstein I, Friedman S: Incidence of interappointment emergency associated with endodontic therapy. J Endod. 1992; 18(10): 509–11. PubMed Abstract \| Publisher Full Text\n\nFarhad A, Mohammadi Z: Calcium hydroxide: a review. Int Dent J. 2005; 55(5): 293–301. PubMed Abstract \| Publisher Full Text\n\nFava LR: Acute apical periodontitis: incidence of post-operative pain using two different root canal dressings. Int Endod J. 1998; 31(5): 343–7. PubMed Abstract \| Publisher Full Text\n\nAnjaneyulu K, Nivedhitha MS: Influence of calcium hydroxide on the post-treatment pain in Endodontics: A systematic review. J Conserv Dent. 2014; 17(3): 200–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSingh RD, Khatter R, Bal RK, et al.: Intracanal medications versus placebo in reducing postoperative endodontic pain--a double-blind randomized clinical trial. Braz Dent J. 2013; 24(1): 25–9. PubMed Abstract \| Publisher Full Text\n\nEhrmann EH, Messer HH, Adams GG: The relationship of intracanal medicaments to postoperative pain in endodontics. Int Endod J. 2003; 36(12): 868–75. PubMed Abstract \| Publisher Full Text\n\nMohammadi Z, Abbott PV: On the local applications of antibiotics and antibiotic-based agents in endodontics and dental traumatology. Int Endod J. 2009; 42(7): 555–67. PubMed Abstract \| Publisher Full Text\n\nTaneja S, Kumari M, Parkash H: Nonsurgical healing of large periradicular lesions using a triple antibiotic paste: A case series. Contemp Clin Dent. 2010; 1(1): 31–5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSiqueira JF Jr, Rôças IN: Clinical implications and microbiology of bacterial persistence after treatment procedures. J Endod. 2008; 34(11): 1291–301.e3. PubMed Abstract \| Publisher Full Text\n\nAli Shah S, Maxood A, Imran Shah S: Incidence of endodontic flare-ups Using either calcium hydroxide or creosote as intracanal medicament in symptomatic teeth. Jkcd. 2010; 1(1): 15–9. Reference Source\n\nMartinho FC, Gomes BP: Quantification of endotoxins and cultivable bacteria in root canal infection before and after chemomechanical preparation with 2.5% sodium hypochlorite. J Endod. 2008; 34(3): 268–72. PubMed Abstract \| Publisher Full Text\n\nMartinho FC, Chiesa WM, Marinho AC, et al.: Clinical investigation of the efficacy of chemomechanical preparation with rotary nickel-titanium files for removal of endotoxin from primarily infected root canals. J Endod. 2010; 36(11): 1766–9. PubMed Abstract \| Publisher Full Text\n\nJacinto RC, Gomes BP, Shah HN, et al.: Quantification of endotoxins in necrotic root canals from symptomatic and asymptomatic teeth. J Med Microbiol. 2005; 54(Pt 8): 777–83. PubMed Abstract \| Publisher Full Text\n\nGomes BP, Martinho F, Vianna M: Comparison of 2.5% sodium hypochlorite and 2% chlorhexidine gel on oral bacterial lipopolysaccharide reduction from primarily infected root canals. J Endod. 2009; 35(10): 1350–3. PubMed Abstract \| Publisher Full Text\n\nSousa EL, Martinho FC, Nascimento GG, et al.: Quantification of endotoxins in infected root canals and acute apical abscess exudates: monitoring the effectiveness of root canal procedures in the reduction of endotoxins. J Endod. 2014; 40(2): 177–81. PubMed Abstract \| Publisher Full Text\n\nHjermstad MJ, Fayers PM, Haugen DF, et al.: Studies comparing Numerical Rating Scales, Verbal Rating Scales, and Visual Analogue Scales for assessment of pain intensity in adults: a systematic literature review. J Pain Symptom Manage. 2011; 41(6): 1073–93. PubMed Abstract \| Publisher Full Text\n\nJensen MP, Turner JA, Romano JM: What is the maximum number of levels needed in pain intensity measurement? Pain. 1994; 58(3): 387–92. PubMed Abstract \| Publisher Full Text\n\nPaice JA, Cohen FL: Validity of a verbally administered numeric rating scale to measure cancer pain intensity. Cancer Nurs. 1997; 20(2): 88–93. PubMed Abstract \| Publisher Full Text\n\nCaraceni A, Cherny N, Fainsinger R, et al.: Pain measurement tools and methods in clinical research in palliative care: recommendations of an Expert Working Group of the European Association of Palliative Care. J Pain Symptom Manage. 2002; 23(3): 239–55. PubMed Abstract \| Publisher Full Text\n\nSowmya HK, Subhash TS, Rani Goel B, et al.: Quantitative assessment of apical debris extrusion and intracanal debris in the apical third, using hand instrumentation and three rotary instrumentation systems. J Clin Diagn Res. 2014; 8(2): 206–10. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShahi S, Asghari V, Rahimi S, et al.: Postoperative Pain after Endodontic Treatment of Asymptomatic Teeth Using Rotary Instruments: A Randomized Clinical Trial. Iran Endod J. 2016; 11(1): 38–43. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPai S, Vivekananda Pai AR, Thomas MS, et al.: Effect of calcium hydroxide and triple antibiotic paste as intracanal medicaments on the incidence of inter-appointment flare-up in diabetic patients: An in vivo study. J Conserv Dent. 2014; 17(3): 208–11. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWalton RE, Holton IF Jr, Michelich R: Calcium hydroxide as an intracanal medication: effect on posttreatment pain. J Endod. 2003; 29(10): 627–9. PubMed Abstract \| Publisher Full Text\n\nTarale K: Post-operative pain analysis between single visit and two visit root canal treatments using visual analogue Scale: an in vivo study. J Dent Allied Sci. 2013; 2(1): 8–15. Publisher Full Text\n\nGhoddusi J, Javidi M, Zarrabi MH, et al.: Flare-ups incidence and severity after using calcium hydroxide as an intra canal dressing. Iran Endod J. 2010; 1(1): 7–13.\n\nQuadir F, Amin F, Shahbaz U: Comparison of intracanal medications for the assessment of pain after root canal treatment. Pakistan Oral Dent J. 2015; 35(2): 286–9. Reference Source\n\nSiqueira JF Jr, de Uzeda M: Intracanal medicaments: evaluation of the antibacterial effects of chlorhexidine, metronidazole, and calcium hydroxide associated with three vehicles. J Endod. 1997; 23(3): 167–9. PubMed Abstract \| Publisher Full Text\n\nSamir Abouelenien S, Mohamed Ibrahim S, Gameel Shaker O, et al.: Dataset 1 in: Evaluation of postoperative pain in infected root canals after using double antibiotic paste versus calcium hydroxide as intra-canal medication: A randomized controlled trial. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16820.d224359" }	[ { "id": "40440", "date": "12 Nov 2018", "name": "Reham Hassan", "expertise": [ "Reviewer Expertise Endodontics", "intracanal medication", "obturation", "rotary endodontic files" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article addresses the links between intra-canal medication used and the post operative pain after endodontic treatment, as mentioned in the article, post operative pain is multifactorial, however the design of the experiment is adequate including experimental and control groups, the primary outcome was measured preoperatively and after 6, 12, 24 and 48 hours only, however in the conclusion its mentioned that the 7 days was sufficient for reducing the post-operative pain. (The use of intra-canal medication for 7 days in necrotic teeth with apical periodontitis was efficient in reducing postoperative pain regardless of the type of intra-canal medication used).\n\nApproved: No changes are required. The experimental design, including controls and methods, is adequate; results are presented accurately and the conclusions are justified and supported by the data. The research is clear and cites the current literature in the topic of post-operative pain after intra-canal medication. The data published and the supplemental data provided is sufficient and clear to allow replication The conclusion is sufficiently supported by the results but could omit the (7 days) to prevent confusion.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "41046", "date": "24 Jan 2019", "name": "Brenda P F A Gomes", "expertise": [ "Reviewer Expertise endodontics", "endodontic microbiology", "endotoxins" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe objective of this study was to assess the ability of double antibiotic paste versus calcium hydroxide used as intra-canal medication in reducing postoperative pain. Post-operative pain is currently a relevant subject in the dental practice. Overall, the study was well written and designed. Additionally, it fits in a Randomized Control Trial, which increases the significance of the research. The results are well described, and no differences were observed between the groups; however, clinically, the antibiotic pastes present some disadvantages compared to calcium hydroxide. The sample size is adequate for this kind of research. I recommend the manuscript for indexing.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1768
https://f1000research.com/articles/7-1617/v1	08 Oct 18	{ "type": "Method Article", "title": "Using zebrafish larval models to study brain injury, locomotor and neuroinflammatory outcomes following intracerebral haemorrhage", "authors": [ "Siobhan Crilly", "Alexandra Njegic", "Sarah E. Laurie", "Elisavet Fotiou", "Georgina Hudson", "Jack Barrington", "Kirsty Webb", "Helen L. Young", "Andrew P. Badrock", "Adam Hurlstone", "Jack Rivers-Auty", "Adrian R. Parry-Jones", "Stuart M. Allan", "Paul R. Kasher", "Siobhan Crilly", "Alexandra Njegic", "Sarah E. Laurie", "Elisavet Fotiou", "Georgina Hudson", "Jack Barrington", "Kirsty Webb", "Helen L. Young", "Andrew P. Badrock", "Adam Hurlstone", "Jack Rivers-Auty", "Adrian R. Parry-Jones", "Stuart M. Allan" ], "abstract": "Intracerebral haemorrhage (ICH) is a devastating condition with limited treatment options, and current understanding of pathophysiology is incomplete. Spontaneous cerebral bleeding is a characteristic of the human condition that has proven difficult to recapitulate in existing pre-clinical rodent models. Zebrafish larvae are frequently used as vertebrate disease models and are associated with several advantages, including high fecundity, optical translucency and non-protected status prior to 5 days post-fertilisation. Furthermore, other groups have shown that zebrafish larvae can exhibit spontaneous ICH. The aim of this study was to investigate whether such models can be utilised to study the pathological consequences of bleeding in the brain, in the context of pre-clinical ICH research. Here, we compared existing genetic (bubblehead) and chemically inducible (atorvastatin) zebrafish larval models of spontaneous ICH and studied the subsequent disease processes. Through live, non-invasive imaging of transgenic fluorescent reporter lines and behavioural assessment we quantified brain injury, locomotor function and neuroinflammation following ICH. We show that ICH in both zebrafish larval models is comparable in timing, frequency and location. ICH results in increased brain cell death and a persistent locomotor deficit. Additionally, in haemorrhaged larvae we observed a significant increase in macrophage recruitment to the site of injury. Live in vivo imaging allowed us to track active macrophage-based phagocytosis of dying brain cells 24 hours after haemorrhage. Morphological analyses and quantification indicated that an increase in overall macrophage activation occurs in the haemorrhaged brain. Our study shows that in zebrafish larvae, bleeding in the brain induces quantifiable phenotypic outcomes that mimic key features of human ICH. We hope that this methodology will enable the pre-clinical ICH community to adopt the zebrafish larval model as an alternative to rodents, supporting future high throughput drug screening and as a complementary approach to elucidating crucial mechanisms associated with ICH pathophysiology.", "keywords": [ "Intracerebral haemorrhage", "zebrafish", "neuroinflammation", "animal models", "pre-clinical" ], "content": "\n\nNo surgery is required so the models more accurately recapitulate spontaneous ICH\n\nEase of genetic manipulation associated with zebrafish allows for the use of transgenic and mutant lines\n\nTransparency of zebrafish larvae allows for non-invasive monitoring of cellular processes over time (live imaging)\n\nUnprotected zebrafish larvae (before 5 dpf) can be used as an alternative to rodent models in pre-clinical ICH research\n\nHigh fecundity of zebrafish pairings means that a large sample size of genetically comparable siblings can be produced easily\n\nZebrafish are a lower cost organism to host than rodents\n\nExperimental timeline is less time consuming than typical rodent studies which look at long term outcomes of stroke commonly up to 3 months\n\nElucidating immediate pathological outcomes of spontaneous ICH in the brain of intact animals\n\nHigh throughput drug screens for ICH therapies\n\nReplacing the current rodent models for analysis of immediate ICH pathology\n\n\nIntroduction\n\nIntracerebral haemorrhage (ICH) accounts for 10–15% of strokes and has the worst stroke outcomes, with a 1-month case fatality of 40% and disability in most survivors (An et al., 2017). The effects of ICH in the brain are biphasic. Primary injury following an ICH event arises due to an influx of blood into the brain and haematoma expansion which increases intracranial pressure on cerebral structures causing neuronal death and cell necrosis (Mracsko & Veltkamp, 2014; Xi et al., 2006). A secondary wave of injury is induced by the breakdown of blood compounds which activates the immune system, further exacerbating cellular damage and death in the brain parenchyma and induces a breakdown in blood-brain barrier integrity (Lok et al., 2011). This inflammatory component is considered a viable therapeutic target in ICH and other forms of stroke (Veltkamp & Gill, 2016) and targeting the toxic insult of blood components after bleed onset is being investigated clinically (Yeatts et al., 2013). However at present, apart from acute and chronic blood pressure lowering, we have no specific treatments to prevent ICH or improve patient outcomes once bleeding has occurred.\n\nDespite representing a significant public health burden (WHO, 2017), an understanding of the fundamental pathogenesis of ICH is still lacking. Pre-clinical studies to-date have depended heavily on rodent models of ICH, which have improved our knowledge of the basic mechanisms associated with the disease (Casals et al., 2011). However, current rodent models of ICH involve severe surgical intervention, poorly recapitulating the spontaneous and immediate nature of the human disease and presenting welfare implications associated with severe experimental procedures in mammals (ASPA, 1986 amendments 2012). Autologous blood injection and collagenase injection models (Andaluz et al., 2002) are used worldwide, and typical experimental groups include 6-8 rats, sacrificed at various time points for ex vivo histological analysis, which can result in ~150 animals used per publication (Wang et al., 2018) highlighting scope for a change of strategy from a 3Rs perspective. Unfortunately, this research has not yet resulted in the translation of any specific drugs to the clinic (Kellner & Connolly, 2010; Kirkman et al., 2011). Potential reasons for this include difficulties in observing cellular responses in ‘real-time’ within whole brains of intact live animals, and the invasive and artificial procedures required to induce cerebral haematomas (MacLellan et al., 2010; Selim et al., 2018). Mammalian models of spontaneous ICH models do exist, such as cerebral amyloid angiopathy co-morbidity studies and use of hypertensive mice, but their usefulness is limited due to variability in haematoma size, timing and location (Alharbi et al., 2016). Alternative and complementary approaches are therefore needed to bridge the ‘translational gap’ for novel drug target discovery in ICH.\n\nZebrafish (Danio rerio) are becoming an increasingly popular tool for studying cerebrovascular disease (Walcott & Peterson, 2014). Due to the production of hundreds of offspring from a single adult pairing, zebrafish larvae can be utilised for high-throughput drug screening, thus offering an attractive model for pre-clinical research. Larval transparency and the availability of numerous transgenic reporter lines amount to an extremely powerful system for studying and visualising cellular responses and disease processes in-vivo in real time. Prior to 5 days post fertilisation (dpf), larval zebrafish are not a protected species (in the UK) and could therefore replace a significant number of protected mammals used for pre-clinical study. Furthermore, spontaneous brain-specific bleeding can be observed in zebrafish larvae using non-invasive techniques (Eisa-Beygi et al., 2013; Liu et al., 2007), thereby eliminating specific constraints associated with mammalian models. As such, the use of larval zebrafish models of ICH could offer critical insight into the immediate cellular responses after a bleed to support the rodent community and provide a potential platform for future drug discovery addressing pre-clinical ICH priorities (Selim et al., 2018).\n\nAs previously described, zebrafish larvae exposed to atorvastatin (ATV) at 24 hours post-fertilisation (hpf) exhibit spontaneous cerebral-specific blood vessel rupture at the onset of circulation (~33 hpf) (Eisa-Beygi et al., 2013; Huang et al., 2018; Li et al., 2017; Shen et al., 2013). Comparably, the ‘bubblehead’ (bbh) mutant line, which expresses a hypomorphic mutation in the arhgef7 gene, encoding the Rac GEF βpix, also exhibit spontaneous ICH and hydrocephalus within a similar time frame to the ATV model (Liu et al., 2007; ten Klooster et al., 2006). ICH is induced through comparable mechanistic defects in both ATV and bbh models (Eisa-Beygi & Rezaei, 2016). Although several groups have utilised these models to study the development and integrity of the cerebrovasculature (Buchner et al., 2007; Huang et al., 2018; Li et al., 2017; Liu et al., 2007; Yang et al., 2017), they have not yet been used to study the pathological and neuroinflammatory consequences of bleeding in the zebrafish larval brain in the context of human ICH. Furthermore, drug intervention studies have focussed on preventing cerebrovascular rupture in zebrafish rather than targeting the disease outcomes, which represent a more realistic therapeutic avenue. In this study, we show that spontaneous ICH in non-protected zebrafish larvae induces quantifiable pathological and inflammatory phenotypes that relate to the human condition. As such, these data support the use of this model species as a valuable complementary system for pre-clinical ICH research.\n\n\nMethods\n\nA detailed protocol of the experimental procedure is available in Supplementary File 1\n\nZebrafish were raised and maintained at The University of Manchester Biological Services Unit under standard conditions as previously described (Westerfield, 2000). Adult zebrafish husbandry was approved by the University of Manchester Animal Welfare and Ethical Review Board. All experiments were performed in accordance with U.K. Home Office regulations (PPL:70/9091) and reported according to ARRIVE guidelines. Transgenic lines used in this study include macrophage-specific lineage mpeg1:mCherry (constructed in-house as previously described (Ellett et al., 2011)), neutrophil-specific mpo:GFP (Renshaw et al., 2006), erythroid-specific gata1:dsRed (Traver et al., 2003) and ubiq:secAnnexinV-mVenus, a reporter for apoptosis (re-derived in house (Morsch et al., 2015)) on wild-type, nacre (mitfaw2/w2) and mutant (bbhm292) backgrounds. Fertilized embryos were collected from natural spawning and incubated at 28°C in standard E3 embryo medium and staged according to standard guidelines (Kimmel et al., 1995). At experiment end, zebrafish larvae were terminated prior to protected status using a lethal overdose of MS222 anaesthesia and freezing at -20°C.\n\nA completed ARRIVE checklist is available in Supplementary File 2\n\nICH was modelled using genetic (bbh) and chemical (ATV) approaches. For the bbh line (Liu et al., 2007), embryos were obtained from adult in-crosses from heterozygous bbhm292 mutant animals (maintained on wild-type and transgenic reporter backgrounds). For the ATV model, nacre or transgenic embryos were dechorionated at 24 hpf and transferred to clean petri dishes in E3 embryo medium. ATV (Sigma-Aldrich, PZ0001) was solubilised in distilled water to a stock concentration of 0.5 mM. Embryos (n=100) were treated with a final concentration of 1 μM ATV through water bath incubation at 28°C for 24 hours and equivalent numbers were left as untreated controls. A proportion of ATV-treated embryos did not develop ICH and therefore these animals were used as controls for the treatment (ICH-) alongside untreated (UNT) siblings. For both bbh and ATV models, embryos with evident haemorrhages (ICH+) were separated from non-haemorrhaged (ICH-) siblings at ~52 hpf for downstream analyses.\n\nLocomotion was measured at 120 hpf to determine if ICH resulted in a physical phenotype. To remove locomotor function bias, larvae were briefly anesthetised at 72 hpf using 0.02% MS222 in embryo water and selected at random for plating. Following recovery, n=24 larvae were individually transferred to each well of a 24-well plate in 1 ml of fresh methylene-blue-free E3 medium. Cumulative time spent mobile was measured using the DanioVision camera chamber and Ethovision XT software (Noldus, version 11) at room temperature. Analyses were performed on larvae at 72, 96 and 120 hpf. Swimming movement of each individual larva was tracked in the x and y plane for 10 minutes using a white light stimulus to initiate a startle response every 60 seconds. Cumulative time spent swimming was measured from three independent replicates.\n\nTransgenic ICH+ and ICH- larvae were imaged using light sheet microscopy to analyse apoptotic cell death (ubiq:secAnnexinV-mVenus), neutrophils (mpo:GFP) and macrophages (mpeg1:mCherry). At ~72 hpf, randomly selected larvae were anaesthetised using 0.02% MS222 and mounted in 1.5% low-melt agarose (Promega), maintained at room temperature. Images were acquired using a W Plan-Apochromat 20X magnification/1.0 UV-VIS objective for light-sheet microscope (Carl Zeiss Lightsheet Z.1) and processed with ZEN imaging software (version 2.3). Maximum intensity projection (MIP) composites were made from z-stack images and brain regions (excluding the eyes) were analysed for average intensity fluorescence of cells with image background removal using an ImageJ (version 1.52a) macro (Supplementary File 1). Numbers of fluorescent cells in the brain were also verified by blind manual counts from MIPs. Data were collected from n=6-12 randomly selected larvae per group from 3 independent replicates for ATV studies and verified in two replicates for bbh. For time-lapse recording, MIP composites were stitched from a series of successive z-stack images over a period of 18 hours.\n\nExperimental sample sizes were determined by using power calculations from preliminary data using α=0.05 and β=0.80. All statistical analysis was first performed using GraphPad Prism 7.0 and then verified using R (R Core Team, 2018) for non-parametric data and subsequent significance values plotted. Linear mixed modelling (LMM) was used to evaluate the effects of independent factors on the continuous dependent variables (Bates et al., 2015). All factors and interactions were modelled as fixed effects. As there is a lack of independence in fish from the same clutch, “Clutch” was treated as a random effect, modelled with random intercepts for all models. The significance of inclusion of an independent variable or interaction terms were evaluated using log-likelihood ratio. Holm-Sidak post-hocs were then performed for pair-wise comparisons using the least square means (Lenth, 2016). Homoscedasticity and normality of the Pearson residuals were evaluated graphically using predicted vs residual and Q-Q plots, respectively, and transformations were applied when necessary.\n\nFor discrete data and data with non-normal distributions, generalized linear mixed modelling was used (GLMM) (Bates et al., 2015; Fournier et al., 2012; Skaug et al., 2013). Again “Clutch” was treated as a random effect modelled with random intercepts for all models. Appropriate families were selected based on the data distribution. Numerous families and link functions were evaluated where necessary and the optimal parameters were selected based on the Akaike information criterion (AIC). For mobile or non-mobile (yes/no) data, a logistic regression with cloglog link was selected; for count data (number of dead cells) a negative binomial family was selected. The significance of inclusion of an independent variable or interaction terms were evaluated using log-likelihood ratio. Holm-Sidak post-hocs were then performed for pair-wise comparisons using the least square means (Lenth, 2016). Pearson residuals were evaluated graphically using predicted vs level plots. All analyses were performed using R (R Core Team, 2018) (Supplementary File 3).\n\n\nResults\n\nIt has been shown that both the ATV and bbh models share similar underlying mechanisms that are responsible for neuroendothelial weakness in the developing larvae and spontaneous cranial vessel rupture (Eisa-Beygi et al., 2013; Liu et al., 2012). In this study we compared these models in the context of ICH and characterised the pathological outcomes of haemorrhage to develop a new platform for pre-clinical interrogation of post-bleed consequences. Using the translucent nature of the zebrafish embryo we observed brain-specific bleeding non-invasively, using light and fluorescent microscopy (Figure 1A). Bleeds were observed in fore, mid and hindbrain regions at comparable frequencies, as described by others (Eisa-Beygi et al., 2013; Liu et al., 2007). ATV absorption induced haemorrhages in a dose-dependent manner when embryos were treated at 24 hpf (Figure 1B). Homozygous mutant bbh embryos and ATV-treated embryos both exhibited ICH between 38 and 48 hpf (Figure 1C, D). In homozygous mutant bbh embryos, ICH was frequently accompanied with severe cranial oedema (Liu et al., 2007) (Supplementary Figure 1). Wild-type and heterozygous bbh siblings had no haemorrhages and were utilised as ICH- controls.\n\n(A) Brain-specific bleeds were observed in both ATV and bbh models maintained on the transgenic gata1:DsRed reporter background using both brightfield (top panels) and fluorescence (bottom panels) microscopy. Bleeds formed in both forebrain and mid-hindbrain regions, as described by others (Eisa-Beygi et al., 2013) (arrows denotes haemorrhages). Bleeds in bbh mutants are frequently associated with severe cranial oedema making blood pooling more disperse. Original magnification, x20. (B) ATV treatment causes ICH to occur in a dose-dependent manner. (C) Timeline of ICH development in ATV-treated and untreated embryos and (D) bbh homozygotes.\n\nTo determine the pathological consequences of ICH in zebrafish larvae, we next assessed brain injury using a transgenic ubiq:secAnnexinV-mVenus cell death reporter line (Morsch et al., 2015). AnnexinV binding was assessed between 48 and 120 hpf, to characterise the timeline of cell death following an ICH event (Supplementary Figure 2). We observed peri-haematomal brain-damaged lesions as ‘clusters’ of dying cells formed by 72 hpf, which had receded before 96 hpf (Supplementary Figure 2). We quantified brain lesions from images taken at 72 hpf (Figure 2A). In both ATV and bbh models, bleeding was associated with a significant, two-fold increase in intensity of fluorescent signal in the brain compared to ICH- controls (Figure 2B, C). These data were verified using blinded, manual counts of annexinV-positive cells from MIP images (Supplementary Figure 3). Taken together, these data provide convincing evidence that ICH in zebrafish larvae induces a reproducible cell death phenotype that can act as a quantifiable readout of brain injury.\n\n(A) Representative images of the brain injury phenotype in ICH+ larvae (right panels), in comparison to ICH- siblings (left panels), at 72 hpf. Bright-field images (bottom panels) demonstrate the presence of brain bleeds (arrows) in ICH+ larvae. Fluorescent microscopy was performed to visualise cell death in the ubiq:secAnnexinV-mVenus reporter line (top panels). Clusters of dying cells were observed in peri-haematomal regions. Images were cropped to brain only regions and analysed for total green fluorescence intensity in round particles bigger than 30 pixels in diameter (white line). (B) Quantification of fluorescent signal in the brains of untreated, ICH- and ICH+ larvae obtained through the ATV model (n=12 per group; 3 independent replicates) at 72 hpf. Significant differences were observed when comparing ICH+ with untreated (*p=0.004) and with ICH- (p=0.03) siblings. (C) Quantification of fluorescent signal as a read out for annexinV binding in the brains of ICH- and ICH+ larvae obtained through the bubblehead (bbh) model (n=12 per group; 2 independent replicates) at 72 hpf. A significant difference in mVenus fluorescence was observed between ICH+ and ICH- age-matched siblings (p=0.002). Original magnification, x20.\n\nTo investigate whether ICH-induced brain injury was associated with a locomotor deficit, as frequently seen in stroke patients (Kloter et al., 2011; Saulle & Schambra, 2016), we tracked swimming behaviour between 72 and 120 hpf. Larvae were analysed across 3 days to account for the improvement in baseline swimming performance associated with increasing age. Representative tracks are shown in Figure 3A. ICH+ larvae spent significantly less time swimming in the cumulative time spent mobile during the 10 minute recording period at both 72 and 96 hpf compared to ICH- siblings, implying a persistent physical deficit (Figure 3B). Although a reduction was observed in ICH+ larvae at 120 hpf, this was not significantly different to ICH- siblings (p=0.08). However ATV-treated larvae assayed at 120 hpf did show a significant reduction in swimming in ICH+ larvae and not in controls. Comparable motility phenotypes in both models imply that the swimming deficit is due to cerebral bleed and not statin treatment. Reproducing these results in both models suggests that the impairment in locomotion is not caused by any one mechanistic factor but due to ICH itself.\n\n(A) Representative examples of the swimming tracks in ICH- and ICH+ larvae at 72, 96 and 120 hpf. (B) ICH+ larvae exhibited a significant decrease in the cumulative time spent mobile during the 10 minute recording period at both 72 and 96 hpf. Significance was lost at the 120 hpf time point potentially alluding to recovery from brain injury (n=24 larvae per group; 3 independent replicates; p=0.00006; p=0.003 ns p=0.08) (C) Quantification of cumulative time spent moving in untreated and ATV-treated ICH- and ICH+ larvae at 120 hpf. ICH+ larvae exhibited a significant decrease in the cumulative time spent mobile during the 10 minute recording period. Three technical replicates (n=24 larvae per group) were used to calculate s.d from the mean (*p=0.00004, p=0.0003).\n\nIn order to determine whether ICH in zebrafish larvae initiated inflammation at the cellular level, neutrophils and macrophages in the brain were quantified using the mpo:GFP and mpeg1:mCherry reporter lines. At 72 hpf, the number of macrophages increased significantly, doubling in the brains of ICH+ larvae compared to ICH- siblings (Figure 4). At the same time point, neutrophil numbers were greater in ICH+ larvae than ICH- larvae; however, the difference did not reach significance.\n\nNumbers of leukocytes quantified within the brains of mpo:GFP;mpeg1:dsRed double transgenic larvae (n=8 per group; 2 independent replicates) at 72 hpf reveals a significant increase in macrophages (p=0.01), but not neutrophils (p=0.5), in response to ICH.\n\nWe investigated the phagocytic response of activated macrophages to the brain lesion sites in ubiq:secAnnexinV-mVenus;mpeg1:mCherry ICH+ larvae using real time microscopy. The formation of a new brain lesion site was recorded between 55 and 65 hpf (Supplementary Video 1) and macrophages were seen migrating to sites of injury and actively phagocytosing annexinV positive dying cells (Figure 5A). We next analysed total macrophage activity within the brain, using morphology as an indication of phagocytic activation. Amoeboid, rounded cells were considered phagocytic and ramified cells considered inactive, as previously defined (Morsch et al., 2015). We found a substantial increase in the proportion of activated, amoeboid macrophages in the ICH+ brain compared to ICH- siblings (Figure 5B, C). These results indicate that the innate immune response to ICH can be examined at the cellular level in zebrafish ICH models using real-time microscopy.\n\n(A) Representative time-lapse stills (from Supplementary Video 1) showing a ramified patrolling macrophage migrating towards an annexinV positive cell (i - vi). The macrophage acquired an amoeboid morphology (v) before phagocytosing the annexinV-positive cell (vi, vii). After phagocytosis the macrophage resumes a ramified morphology and migrates away and the annexinV-positive cell can no longer be seen (viii). Ramified macrophage (#), annexinV positive cell (arrow), amoeboid macrophage (). (B) Representative images of mpeg1-positive cells in the intracerebral haemorrhage (ICH)- and ICH+ larval brain exhibiting amoeboid and ramified morphologies. (C) An increased proportion of amoeboid (phagocytic) and decreased proportion of ramified (inactive) macrophages was observed in ICH+ brains in comparison to ICH- siblings.\n\n\nDiscussion\n\nHere, we reveal that ICH in zebrafish larvae induces quantifiable pathological and inflammatory consequences that mimic aspects of human pathophysiology. Several groups have previously described the mechanisms underlying neurovascular instability in zebrafish larval models of ICH (Buchner et al., 2007; Eisa-Beygi et al., 2013; Liu et al., 2007), and drug intervention studies have attempted to identify compounds that can inhibit cerebral bleeding (Li et al., 2017; Yang et al., 2017). However, the relevance and translational impact of intervention before onset of haemorrhage is unclear. As such, we focussed our attention on characterising the pathological and immunological consequences of blood in the brain in zebrafish models, which we consider to be a more realistic therapeutic target.\n\nIn general, ICH can be predominantly considered as a disorder associated with older age. Consequently, it is remarkable that the phenotypes observed in developing zebrafish recapitulate those associated with the aged human brain, indicating the innate injury response to blood in the brain is both evolutionarily conserved between species and analogous during development and adulthood. Zebrafish models have been employed in other neurological and neuropsychiatric disease investigation, including epilepsy, schizophrenia, Alzheimer’s and Parkinson’s disease, because of these conserved fundamental mechanisms and behaviours (Fontana et al., 2018; Vaz et al., 2018). Although these models present their own limitations, the use of non-invasive in vivo imaging, ease of genetic manipulation and availability of large sample sizes offers new insights and overcomes some of the common restrictions associated with rodent models. Given that ICH occurs spontaneously using non-invasive techniques in zebrafish, it can be argued this system more accurately models some aspects of the human condition than the most commonly used rodent models (Andaluz et al., 2002; Rosenberg et al., 1990). The use of rodents has not been successful in terms of identifying translatable therapeutic targets for ICH. As such, we propose that the post-ICH pathologies presented in this study (Figure 6) represent an alternative, complementary platform for pre-clinical ICH research and future drug discovery.\n\nICH, intracerebral haemorrhage; bbh, bubblehead.\n\nThe transferability of employing these zebrafish larval models would address the 3Rs welfare issue of prolonged severe surgical procedures in mammals by replacing some of these animals with zebrafish larvae of unprotected status. We have used equipment and procedures that are commonplace and would be relatively simple to adopt in other labs. We would like to see these models used to determine the translatability of drugs that prevent cerebral bleeding in zebrafish (Huang et al., 2018; Yang et al., 2017), to investigate the clinical relevance of post-ICH treatment. We also propose these endpoint assays can be used to develop medium/high-throughput drug screening to identify new compounds for pre-clinical investigation.\n\nIn humans, an influx of blood into the brain causes primary brain injury through neuronal death and cell necrosis, inducing a secondary phase of injury triggered by the production of inflammatory mediators and innate immune cell migration towards the site of injury (Mracsko & Veltkamp, 2014). We show that cerebral bleeding in zebrafish larvae causes an increase in cell death in the brain not shown before in zebrafish haemorrhage models. This lesion was associated with a physical impairment, as observed by a reduction in swimming ability, which we propose is either due to a defect in stimulus perception or through a motor deficiency. This impairment is seen to begin to recover at 3 days post injury, suggesting the zebrafish larval model system as a useful tool for further investigation to reveal recovery processes responding to brain cell death. Importantly, these observations mimic outcomes that are exhibited in rodent models (MacLellan et al., 2008) and ICH patients (Saulle & Schambra, 2016). In clinical presentation of ICH the initial haematoma mass effect and increase in intracranial pressure exacerbates brain damage, compressing surrounding structures (Xi et al., 2006) and increases risk of death following ICH (Yang et al., 2015). In zebrafish larvae, cerebral oedema is regularly associated with ICH (Kasher et al., 2015; Liu et al., 2007). However without a fully developed cranium, oedema is unlikely to result in the same injury severity seen in humans, which may be one limitation of this particular model system.\n\nAn increase in recruitment and activation of macrophages in the brain was also observed in zebrafish following haemorrhage corresponding to time of cell death. A trend towards increased recruitment of neutrophils was also observed, but this result did not reach statistical significance. It is possible that a rapid temporal neutrophil response was not fully captured during our analytical time frame. However, it has been shown that neutrophils are not involved in the clearance of cellular debris in a zebrafish larval model of brain injury (van Ham et al., 2014), suggesting neutrophils are less vital to early injury responses in the brain than macrophages. It remains unclear whether activation of macrophages post-ICH is overall beneficial or deleterious in the short term after injury. Pro-inflammatory cells contribute to the breakdown of the blood-brain barrier (Abbott, 2000); however, phagocytic cells promote clearance of red blood cells and tissue debris, which occurs from 7 days post haemorrhage in rodent models of ICH (Mracsko & Veltkamp, 2014; Yang et al., 2016b). Early responses to laser-induced cerebral bleeding in zebrafish have shown that macrophages are essential for vessel repair (Liu et al., 2016) and in zebrafish stab wound brain injury models, inflammation is necessary for regeneration and recovery (Kyritsis et al., 2012). Studies show that polarisation of macrophages to M1-like and M2-like states change over time in rodent models of ICH in response to brain damage molecules (Wan et al., 2016; Yang et al., 2016a) and some drugs in clinical trials target microglial polarisation in an attempt to prevent the pro-inflammatory phenotype (Lan et al., 2017). In vivo imaging of ICH-induced inflammatory processes has barely been explored (Mracsko & Veltkamp, 2014), and so the translucent nature of zebrafish larvae and availability of transgenic lines offer an accessible model for further interrogation. Observations of cellular interactions within whole, intact rodent brains at ongoing time points are currently not possible, therefore utilising the zebrafish system will allow us to learn more about leukocyte behaviour after spontaneous ICH. Furthermore, immune responses observed in real time in this system will have been elicited by spontaneous vessel rupture and not as an artefact of ICH surgery or ex vivo analysis (Kirkman et al., 2011; Xue & Del Bigio, 2003). For better translation of therapies from pre-clinical to patients, understanding of early innate immune responses to spontaneous bleeding needs to be improved, thus zebrafish offer a powerful resource to facilitate this.\n\nIn conclusion, given the advantages associated with zebrafish larvae and the potential for a 3Rs approach to pre-clinical stroke research, we propose that this model organism can provide critical insight into ICH pathophysiology during the early phases of injury and offers a future platform for drug discovery.\n\n\nData availability\n\nDataset 1. All raw data from the present study. Data include all raw microscopy images of cell death in Annexin and bubblehead groups; fluorescence intensities, cell counts and motility times for ATV and bubblehead groups; and leukocyte cell counts. DOI: https://doi.org/10.5256/f1000research.16473.d220415 (Crilly et al., 2018).", "appendix": "Grant information\n\nThis study was supported by the NC3Rs (NC/N002598/1), Stroke Association (TSA LECT 2017/02), ERA-NET NEURON (MR/M501803/1) and The British Heart Foundation (FS/14/64/31291, FS/15/67/32038). We are also particularly thankful to The Natalie Kate Moss Trust for their continued financial support.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe would like to thank Prof Yanick Crow and Dr Gillian Rice for use of equipment and reagents, Dr David Spiller and the University of Manchester Systems Microscopy Core Facility for use of the light sheet microscope, and Prof Richard Baines for the use of DanioVision. The bbh line was kindly shared by Nicole Munsie from Dr Sarah Child’s lab at the University of Calgary. We also thank Prof. Stephen Renshaw for sharing the mpo:GFP line.\n\n\nSupplementary material\n\nSupplementary File 1. Full detail protocol for experimental methodology.\n\nClick here to access the data\n\nSupplementary File 2. Completed ARRIVE guidelines checklist.\n\nClick here to access the data\n\nSupplementary File 3. Supplementary statistical analysis.\n\nClick here to access the data\n\nSupplementary Figure 1. Severe cranial oedema (affected regions denoted by hashtags) associated with cranial bleeds in ‘bubblehead’ homozygous mutants (right panel) and subsequently exhibit more dispersed bleeds (arrows) compared to heterozygous and wild-type intracerebral haemorrhage (ICH) (left panel).\n\nOriginal magnification, x20.\n\nClick here to access the data\n\nSupplementary Figure 2. Representative images of the brain injury phenotype in intracerebral haemorrhage (ICH)+ larvae (bottom panels), in comparison to ICH- age-matched siblings (top panels), at 48-120 hpf.\n\nFluorescent microscopy was performed to visualise cell death in the ubiq:secAnnexinV-mVenus reporter line to determine the timeline of cell injury following an ICH event. Original magnification, x20.\n\nClick here to access the data\n\nSupplementary Figure 3. Manual counts of fluorescent cells from blinded analysis of the ubiq:secAnnexinV-mVenus reporter line in atorvastatin (ATV) and bubblehead (bbh) models show a significant increase in intracerebral haemorrhage (ICH)+ (**p=0.0097, ###p=0.0005).\n\nClick here to access the data\n\nSupplementary Video 1. A time lapse recording of phagocytosis occurring in an intracerebral haemorrhage+ ubiq:secAnnexinV-mVenus;mpeg1:mCherry larvae between 55 and 72 hpf.\n\nCluster of dying cells (green annexinV positive) in forebrain visible at 55 hpf and active macrophages (red) can be observed phagocytosing cells. A second cluster in the mid brain develops over recording period and macrophages can be seen migrating to the new lesion site. Original magnification, x20.\n\nClick here to access the data\n\n\nReferences\n\nAbbott NJ: Inflammatory mediators and modulation of blood-brain barrier permeability. Cell Mol Neurobiol. 2000; 20(2): 131–147. PubMed Abstract\n\nAlharbi BM, Tso MK, Macdonald RL: Animal models of spontaneous intracerebral hemorrhage. Neurol Res. 2016; 38(5): 448–455. PubMed Abstract \| Publisher Full Text\n\nAn SJ, Kim TJ, Yoon BW: Epidemiology, Risk Factors, and Clinical Features of Intracerebral Hemorrhage: An Update. J Stroke. 2017; 19(1): 3–10. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAndaluz N, Zuccarello M, Wagner KR: Experimental animal models of intracerebral hemorrhage. Neurosurg Clin N Am. 2002; 13(3): 385–393. PubMed Abstract \| Publisher Full Text\n\nASPA 1986 amendments 2012: Animals (Scientific Procedures) Act 1986. Reference Source\n\nBates D, Mächler M, Bolker B, et al.: Fitting Linear Mixed-Effects Models using lme4. J Stat Softw. Computation. 2015; 67(1): 1–48. Publisher Full Text\n\nBuchner DA, Su F, Yamaoka JS, et al.: pak2a mutations cause cerebral hemorrhage in redhead zebrafish. Proc Natl Acad Sci U S A. 2007; 104(35): 13996–14001. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCasals JB, Pieri NC, Feitosa ML, et al.: The use of animal models for stroke research: a review. Comp Med. 2011; 61(4): 305–313. PubMed Abstract \| Free Full Text\n\nCrilly S, Njegic A, Laurie SE, et al.: Dataset 1 in: Using zebrafish larval models to study brain injury, locomotor and neuroinflammatory outcomes following intracerebral haemorrhage. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16473.d220415\n\nEisa-Beygi S, Hatch G, Noble S, et al.: The 3-hydroxy-3-methylglutaryl-CoA reductase (HMGCR) pathway regulates developmental cerebral-vascular stability via prenylation-dependent signalling pathway. Dev Biol. 2013; 373(2): 258–266. PubMed Abstract \| Publisher Full Text\n\nEisa-Beygi S, Rezaei M: Etiology of intracerebral hemorrhage (ICH): novel insights from Zebrafish embryos. Int J Dev Biol. 2016; 60(4–6): 119–126. PubMed Abstract \| Publisher Full Text\n\nEllett F, Pase L, Hayman JW, et al.: mpeg1 promoter transgenes direct macrophage-lineage expression in zebrafish. Blood. 2011; 117(4): e49–e56. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFontana BD, Mezzomo NJ, Kalueff AV, et al.: The developing utility of zebrafish models of neurological and neuropsychiatric disorders: A critical review. Exp Neurol. 2018; 299(Pt A): 157–171. PubMed Abstract \| Publisher Full Text\n\nFournier DA, Skaug HJ, Ancheta J, et al.: AD Model Builder: using automatic differentiation for statistical inference of highly parameterized complex nonlinear models. Optim Methods Softw. 2012; 27(2): 233–249. Publisher Full Text\n\nHuang B, Zhou ZY, Li S, et al.: Tanshinone I prevents atorvastatin-induced cerebral hemorrhage in zebrafish and stabilizes endothelial cell-cell adhesion by inhibiting VE-cadherin internalization and actin-myosin contractility. Pharmacol Res. 2018; 128: 389–398. PubMed Abstract \| Publisher Full Text\n\nKasher PR, Jenkinson EM, Briolat V, et al.: Characterization of samhd1 morphant zebrafish recapitulates features of the human type I interferonopathy Aicardi-Goutières syndrome. J Immunol. 2015; 194(6): 2819–2825. PubMed Abstract \| Publisher Full Text\n\nKellner CP, Connolly ES Jr: Neuroprotective strategies for intracerebral hemorrhage: trials and translation. Stroke. 2010; 41(10 Suppl): S99–S102. PubMed Abstract \| Publisher Full Text\n\nKimmel CB, Ballard WW, Kimmel SR, et al.: Stages of embryonic development of the zebrafish. Dev Dyn. 1995; 203(3): 253–310. PubMed Abstract \| Publisher Full Text\n\nKirkman MA, Allan SM, Parry-Jones AR: Experimental intracerebral hemorrhage: avoiding pitfalls in translational research. J Cereb Blood Flow Metab. 2011; 31(11): 2135–2151. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKloter E, Wirz M, Dietz V: Locomotion in stroke subjects: interactions between unaffected and affected sides. Brain. 2011; 134(Pt 3): 721–731. PubMed Abstract \| Publisher Full Text\n\nKyritsis N, Kizil C, Zocher S, et al.: Acute inflammation initiates the regenerative response in the adult zebrafish brain. Science. 2012; 338(6112): 1353–1356. PubMed Abstract \| Publisher Full Text\n\nLan X, Han X, Li Q, et al.: Modulators of microglial activation and polarization after intracerebral haemorrhage. Nat Rev Neurol. 2017; 13(7): 420–433. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLenth RV: Least-squares means: the R package lsmeans. J Stat Softw. 2016; 69(1): 1–33. Publisher Full Text\n\nLi S, Ai N, Shen M, et al.: Discovery of a ROCK inhibitor, FPND, which prevents cerebral hemorrhage through maintaining vascular integrity by interference with VE-cadherin. Cell Death Discov. 2017; 3: 17051. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiu J, Fraser SD, Faloon PW, et al.: A betaPix Pak2a signaling pathway regulates cerebral vascular stability in zebrafish. Proc Natl Acad Sci U S A. 2007; 104(35): 13990–13995. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiu C, Wu C, Yang Q, et al.: Macrophages Mediate the Repair of Brain Vascular Rupture through Direct Physical Adhesion and Mechanical Traction. Immunity. 2016; 44(5): 1162–1176. PubMed Abstract \| Publisher Full Text\n\nLiu J, Zeng L, Kennedy RM, et al.: βPix plays a dual role in cerebral vascular stability and angiogenesis, and interacts with integrin αvβ8. Dev Biol. 2012; 363(1): 95–105. PubMed Abstract \| Publisher Full Text\n\nLok J, Leung W, Murphy S, et al.: Intracranial hemorrhage: mechanisms of secondary brain injury. Acta Neurochir Suppl. Intracerebral Hemorrhage Research. Springer. 2011; 111: 63–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMacLellan CL, Silasi G, Auriat AM, et al.: Rodent models of intracerebral hemorrhage. Stroke. 2010; 41(10 Suppl): S95–S98. PubMed Abstract \| Publisher Full Text\n\nMacLellan CL, Silasi G, Poon CC, et al.: Intracerebral hemorrhage models in rat: comparing collagenase to blood infusion. J Cereb Blood Flow Metab. 2008; 28(3): 516–525. PubMed Abstract \| Publisher Full Text\n\nMorsch M, Radford R, Lee A, et al.: In vivo characterization of microglial engulfment of dying neurons in the zebrafish spinal cord. Front Cell Neurosci. 2015; 9: 321. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMracsko E, Veltkamp R: Neuroinflammation after intracerebral hemorrhage. Front Cell Neurosci. 2014; 8: 388. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRenshaw SA, Loynes CA, Trushell DM, et al.: A transgenic zebrafish model of neutrophilic inflammation. Blood. 2006; 108(13): 3976–3978. PubMed Abstract \| Publisher Full Text\n\nRosenberg GA, Mun-Bryce S, Wesley M, et al.: Collagenase-induced intracerebral hemorrhage in rats. Stroke. 1990; 21(5): 801–807. PubMed Abstract \| Publisher Full Text\n\nSaulle MF, Schambra HM: Recovery and Rehabilitation after Intracerebral Hemorrhage. Semin Neurol. NIH Public Access, 2016; 36(3): 306–12. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSelim M, Hanley D, Broderick J, et al.: Basic and Translational Research in Intracerebral Hemorrhage: Limitations, Priorities, and Recommendations. Stroke. 2018; 49(5): 1308–1314. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShen M, Yu H, Li Y, et al.: Discovery of Rho-kinase inhibitors by docking-based virtual screening. Mol Biosyst. 2013; 9(6): 1511–1521. PubMed Abstract \| Publisher Full Text\n\nSkaug H, Fournier D, Nielsen A, et al.: Generalized linear mixed models using AD model builder. R package version 0.7. 2013; 7.\n\nTeam RC: R: A Language and Environment for Statistical Computing. 2018. Reference Source\n\nten Klooster JP, Jaffer ZM, Chernoff J, et al.: Targeting and activation of Rac1 are mediated by the exchange factor beta-Pix. J Cell Biol. 2006; 172(5): 759–769. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTraver D, Paw BH, Poss KD, et al.: Transplantation and in vivo imaging of multilineage engraftment in zebrafish bloodless mutants. Nat Immunol. 2003; 4(12): 1238–46. PubMed Abstract \| Publisher Full Text\n\nvan Ham TJ, Brady CA, Kalicharan RD, et al.: Intravital correlated microscopy reveals differential macrophage and microglial dynamics during resolution of neuroinflammation. Dis Model Mech. 2014; 7(7): 857–869. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVaz RL, Outeiro TF, Ferreira JJ: Zebrafish as an Animal Model for Drug Discovery in Parkinson's Disease and Other Movement Disorders: A Systematic Review. Front Neurol. 2018; 9: 347. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVeltkamp R, Gill D: Clinical Trials of Immunomodulation in Ischemic Stroke. Neurotherapeutics. 2016; 13(4): 791–800. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWalcott BP, Peterson RT: Zebrafish models of cerebrovascular disease. J Cereb Blood Flow Metab. 2014; 34(4): 571–577. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWan S, Cheng Y, Jin H, et al.: Microglia Activation and Polarization After Intracerebral Hemorrhage in Mice: the Role of Protease-Activated Receptor-1. Transl Stroke Res. 2016; 7(6): 478–487. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWang J, Zhai W, Yu Z, et al.: Neuroprotection Exerted by Netrin-1 and Kinesin Motor KIF1A in Secondary Brain Injury following Experimental Intracerebral Hemorrhage in Rats. Front Cell Neurosci. 2018; 11: 432. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWesterfield M: The zebrafish book: a guide for the laboratory use of zebrafish. 2000. Reference Source\n\nWHO: The top 10 causes of death [Online]. 2017; [Accessed 28/02/17]. Reference Source\n\nXi G, Keep RF, Hoff JT: Mechanisms of brain injury after intracerebral haemorrhage. Lancet Neurol. 2006; 5(1): 53–63. PubMed Abstract \| Publisher Full Text\n\nXue M, Del Bigio MR: Comparison of brain cell death and inflammatory reaction in three models of intracerebral hemorrhage in adult rats. J Stroke Cerebrovasc Dis. 2003; 12(3): 152–159. PubMed Abstract \| Publisher Full Text\n\nYang J, Arima H, Wu G, et al.: Prognostic significance of perihematomal edema in acute intracerebral hemorrhage: pooled analysis from the intensive blood pressure reduction in acute cerebral hemorrhage trial studies. Stroke. 2015; 46(4): 1009–13. PubMed Abstract \| Publisher Full Text\n\nYang J, Ding S, Huang W, et al.: Interleukin-4 Ameliorates the Functional Recovery of Intracerebral Hemorrhage Through the Alternative Activation of Microglia/Macrophage. Front Neurosci. 2016a; 10: 61. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYang SS, Lin L, Liu Y, et al.: High Morphologic Plasticity of Microglia/Macrophages Following Experimental Intracerebral Hemorrhage in Rats. Int J Mol Sci. 2016b; 17(7): pii: E1181. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYang R, Zhang Y, Huang D, et al.: Miconazole protects blood vessels from MMP9-dependent rupture and hemorrhage. Dis Model Mech. 2017; 10(3): 337–348. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYeatts SD, Palesch YY, Moy CS, et al.: High dose deferoxamine in intracerebral hemorrhage (HI-DEF) trial: rationale, design, and methods. Neurocrit Care. 2013; 19(2): 257–266. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "39725", "date": "24 Oct 2018", "name": "Claire L. Gibson", "expertise": [ "Reviewer Expertise Referee suggested by the NC3Rs for their scientific expertise and experience in assessing 3Rs impact." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article describes the development of a zebrafish model of intracerebral haemorrhage (ICH). ICH is a significant contributor to human mortality and morbidity and effect treatment options are limited. Research in this area has been hindered by the lack of appropriate models – although rodent models have contributed to our understanding of the pathology underlying ICH such models are based on a surgical intervention which does not mirror the spontaneous event, in humans, of ICH. Here the authors compare an existing genetic (bubblehead) and chemically inducible (atorvastatin) zebrafish models of spontaneous ICSH and importantly assess brain injury, locomotor function and neuroinflammation.\n\nThe article is well described with sufficient experimental detail and the authors clearly articulate the 3Rs benefits of such work. One of the particularly novel aspects of this work is the ability to image in vivo the ICH-induced inflammatory processes and cellular interactions – this is not currently feasible in whole animal models, such as rodents. This may allow improved understanding of the early innate responses to spontaneous bleeding as occurs in ICH patients but which we cannot model in rodents.\n\nAlthough the authors state that this paper used equipment and procedures that are commonplace and relatively simple to adopt in other labs it is worth mentioning what barriers need to be overcome in order to facilitate ‘traditional’ rodent researchers adopting such a model and species.\n\nSpecific comments: Authors state that a proportion of ATV-treated embryos did not develop ICH and therefore were used a controls – what proportion was unsuccessful? Can the authors suggest why not all were successful? Did this result in a group of animals forming a control group which was treated separately to another control group (i.e. the untreated siblings)?\n\nNot clear to me (although I am a rodent researcher!) how a brief exposure to anaesthesia at 72hpf would remove locomotor function bias or was necessary? Couldn’t a baseline measurement be taken and data compared to this? The authors states that the exposure to this brief anaesthesia was at 72hpf but also state functional analyses were done at 72, 96 and 120 hpf – did the first time point occur coincidentally with the brief anaesthesia?\n\nAuthors have made available all underlying source data.\nAre a suitable application and appropriate end-users identified? Yes\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\nAre the 3Rs implications of the work described accurately? Yes\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4102", "date": "08 Nov 2018", "name": "siobhan crilly", "role": "Author Response", "response": "Dear Prof Gibson,Many thanks indeed for reviewing our manuscript and your helpful comments. Please see below for responses to your specific questions.1. Authors state that a proportion of ATV-treated embryos did not develop ICH and therefore were used as controls – what proportion was unsuccessful? Can the authors suggest why not all were successful? Did this result in a group of animals forming a control group which was treated separately to another control group (i.e. the untreated siblings)?The ATV treated non-haemorrhaged group formed the ICH- control group for the ATV model. As you suggest, this group was separate from the untreated control group, and our subsequent analyses compared all 3 groups (i.e. untreated, ICH- and ICH+ groups), as presented in Figure 2B and 3C. The proportion of ICH- controls varied with ATV dose, as presented in Figure 1B. For our cell death and motility assays, we used an ATV dosage of 1.0µg/ml, which resulted in approximately 20% of non-haemorrhaged animals, which we used as our ICH- group. We currently cannot explain why a proportion of ATV-treated larvae do not develop ICH. However as zebrafish are a relatively outbred model species, we hypothesise that a genetic component may explain increased neuroendothelial stability in some animals.2. Not clear to me (although I am a rodent researcher!) how a brief exposure to anaesthesia at 72hpf would remove locomotor function bias or was necessary? Couldn’t a baseline measurement be taken and data compared to this? The authors states that the exposure to this brief anaesthesia was at 72hpf but also state functional analyses were done at 72, 96 and 120 hpf – did the first time point occur coincidentally with the brief anaesthesia?Following fertilisation, we routinely plate n=50-100 embryos per petri dish. At 72hpf, natural variation in swimming behaviour/ability exists between individual larvae. For allocation to the locomotion assay, larvae are transferred from the petri dish into a 24 well plate using a pipette. To avoid always ‘catching’ the weakest/slowest swimmers, we briefly anaesthetise the animals so that larvae are randomly allocated to the assay. These fish were segregated at 72 hpf for repeat assays at 96 and 120 hpf and so there was no need to re-anaesthetise the animals for these later timepoints.Kind regardsSiobhan Crilly & Paul Kasher" } ] }, { "id": "39224", "date": "30 Oct 2018", "name": "Marietta Zille", "expertise": [ "Reviewer Expertise Intracerebral haemorrhage", "stroke researcher", "cell death mechanisms" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this manuscript, the authors propose two models of intracerebral hemorrhage (ICH) in zebrafish larvae to study post-hemorrhage brain injury. The authors compare a genetic (bubblehead) and chemically inducible (atorvastatin) model of spontaneous ICH and elegantly show assessment of cell death and inflammation in these models using live non-invasive imaging. In addition, they present data on locomotor activity as a functional outcome measurement.\nFirst of all, I want to congratulate the authors on this high quality study that is timely and relevant to the ICH community as spontaneous models that allow higher throughput screening of potentially protective or regenerative compounds are clearly needed in the field (Hemorrhagic Stroke Academia Industry (HEADS) Roundtable Participants, 2018). Although new to the ICH community, zebrafish models have multiple advantages that may be of great use in this research area (as well as in others).\nI have a couple of suggestions that I hope will be beneficial to further improve the quality of this excellent publication:\nWe and others have recently established that neurons die by multiple different regulated cell death mechanisms after ICH, including ferroptosis and necroptosis (Zille et al, 2017, Li et al, 2018). Therefore, the authors should discuss the limitation of the cell death marker, Annexin V, used in their study. Annexin V detects exposure of phosphatidiyl serine on the cell surface. On the one hand, this is not specific to apoptosis since when the cell membrane is disrupted phosphatidylserine can be detected intracellularly (and hence should not be claimed to analyze apoptotic cell death, e.g. in Methods section) (Munoz et al, 2013, Sawai et al, 2011, Zille et al, 2012). On the other hand, this only reflects a specific subset of cell death. This may also explain why cell death was detected only in a subset of larvae and at one specific time point which is also a relatively narrow time window. Is there a possibility to determine the involvement of other cell death mechanisms in zebrafish (e.g., GPX4 for ferroptosis or pRIP1/3 for necroptosis)? Undoubtedly, the exact cell death mechanisms warrant further exploration in future studies including pharmacological and molecular interventions. The authors nicely show side-by-side comparison of both bubblehead and atorvastatin model. What is the main difference between the two? Therefore, the authors should make fluorescence pictures from both available in Fig. 2, the same time course in Fig. 3 as well as subgroup analysis in Fig. 4 and 5. Was only one of the two investigated regarding macrophages or are they pooled in this analysis? One of the differences seems to be the occurrence of edema in the bubblehead model. This is very interesting as it is also a key feature in the ICH pathophysiology. The authors should provide pictures on both models over time. Is it possible to quantify the edema? Locomotor activity: How does swimming time represent functional outcome of the larvae? Would distance travelled be an alternative readout? What is the size of the hematoma and does it correlate with cell death, locomotor activity or macrophage activation? For the atorvastatin model, was the 1µM dose used in Figure 2-5? If not, maybe the clusters of positive and negative for Annexin V correspond to the dose of atorvastatin? One of the major disadvantages of the larvae model is that it uses young animals. Is it possible to use atorvastatin to induce ICH in old zebrafish? Regarding modeling human disease, another difference is that there is no mortality in this model at least not to the time points investigated. Do they die at earlier time points than their wildtype/untreated littermates? Do higher dosages of atorvastatin induce mortality in the larvae due to ICH?\nMinor:\n\nFrom the data where replicates were acquired, please specify whether the mean or median was chosen. What is the % MS222 used as lethal overdose?\n\nAre a suitable application and appropriate end-users identified? Yes\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\nAre the 3Rs implications of the work described accurately? Yes\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4103", "date": "08 Nov 2018", "name": "siobhan crilly", "role": "Author Response", "response": "Dear Dr Zille, Many thanks indeed for reviewing our manuscript and your helpful comments. Please see below for responses to your specific questions. Where necessary, we have made amendments in a second version of our manuscript, which we also refer to below. We and others have recently established that neurons die by multiple different regulated cell death mechanisms after ICH, including ferroptosis and necroptosis (Zille et al, 2017, Li et al, 2018). Therefore, the authors should discuss the limitation of the cell death marker, Annexin V, used in their study. Annexin V detects exposure of phosphatidiyl serine on the cell surface. On the one hand, this is not specific to apoptosis since when the cell membrane is disrupted phosphatidylserine can be detected intracellularly (and hence should not be claimed to analyze apoptotic cell death, e.g. in Methods section) (Munoz et al, 2013, Sawai et al, 2011, Zille et al, 2012). On the other hand, this only reflects a specific subset of cell death. This may also explain why cell death was detected only in a subset of larvae and at one specific time point which is also a relatively narrow time window. Is there a possibility to determine the involvement of other cell death mechanisms in zebrafish (e.g., GPX4 for ferroptosis or pRIP1/3 for necroptosis)? Undoubtedly, the exact cell death mechanisms warrant further exploration in future studies including pharmacological and molecular interventions. We thank the reviewer for raising this point, and we appreciate that only using Annexin V in this study limits the information that we can obtain about cell death mechanisms. As suggested, we have updated our discussion in version 2 to address this, and removed the term ‘apoptosis’ throughout. Of interest, in addition to the Annexin V reporter line, we have also verified cell death using an acridine orange assay in live larvae at the same time points and found an identical pattern of brain lesions in haemorrhaged animals only at 72 hpf (data not shown). The primary purpose of this paper was to demonstrate that a range of ‘translatable’ disease outcomes exist following ICH in zebrafish larvae. However, our ongoing studies will aim to build a more accurate understanding of the precise mechanisms underlying brain injury following ICH in zebrafish larvae, which will include an investigation into other forms of cell death, such as necroptosis and ferroptosis. The authors nicely show side-by-side comparison of both bubblehead and atorvastatin model. What is the main difference between the two? Exposure to the HMG-CoA reductase inhibitor, atorvastatin (ATV) at 24 hours post-fertilisation (hpf) leads to spontaneous cerebral-specific blood vessel rupture at ~33hpf in zebrafish larvae at the onset of circulation (Eisa-Beygi et al., 2013, Shen et al., 2013, Huang et al., 2017, Li et al., 2017). By inhibiting HMG-CoA in neovascular neuroendothelial cells, ATV disrupts isoprenylation and hence activation of Rac1 (Xiao et al., 2013) and subsequent actin remodelling to stabilise VE cadherin mediated tight junctions (Eisa-Beygi and Rezaei, 2016). The result of this impairment is a ‘leaky’ cerebrovasculature and ICH between 33-48hpf. Comparably, the ‘bubblehead’ (bbh) mutant line, which expresses a hypomorphic mutation in the arhgef7 gene, encoding the Rac GEF βpix, also exhibit spontaneous ICH, as well as hydrocephalus, within a similar time frame to the ATV model (Liu et al., 2007, ten Klooster et al., 2006). Therefore it appears ICH is induced through comparable mechanistic defects in both ATV and bbh models. We have included a summary of this difference in new supplementary figure 1 in version 2. Therefore, the authors should make fluorescence pictures from both available in Fig. 2, the same time course in Fig. 3 as well as subgroup analysis in Fig. 4 and 5. Was only one of the two investigated regarding macrophages or are they pooled in this analysis? In our updated version, we have now included images of the ATV model with Annexin V cell death lesions at 3 dpf in supplementary figure 2. We have monitored locomotor function in the ATV model larvae over the same time course as figure 3, however the baseline measurements of control larvae at 3 and 4 dpf was too low for a measurable outcome, and so we continued with the bbh mutant. We hypothesise this could be due to individual strain effect, as we have observed variable locomotor function in normal larvae between different strains. The ATV model was not used for the studies in figure 4 and 5 due to the anti-inflammatory effects of statins (Bu et al., 2011). One of the differences seems to be the occurrence of edema in the bubblehead model. This is very interesting as it is also a key feature in the ICH pathophysiology. The authors should provide pictures on both models over time. Is it possible to quantify the edema? For the ATV model of ICH, larvae do not exhibit any oedema at any age. Also, although oedema occurs exclusively in ICH+ bbh homozygotes it occurs in a stochastic fashion. Liu et al (2007) have shown that oedema and haemorrhage occur as separate entities in bbh mutants and develop independently of each other. For these reasons and the limitations raised in our discussion, we avoided using oedema as a pathological outcome for haemorrhagic stroke. Locomotor activity: How does swimming time represent functional outcome of the larvae? Would distance travelled be an alternative readout? We have recorded distance travelled for both models in conjunction to time spent mobile and the trend is identical to that shown in Figure 3. We decided to use cumulative time spent mobile as this would be inclusive of seizure-like behaviour as well as normal thigmotactic behaviour (Turrini et al., 2017) What is the size of the hematoma and does it correlate with cell death, locomotor activity or macrophage activation? This is an extremely interesting question, and will form the basis of a future grant application. We have not accurately measured haematoma size, however we will use 3D rendering and light sheet microscopy to do so in the future. We are also interested in investigating haematoma location within the brain, to determine if this effects behavioural outcomes/specific neuronal populations. For the atorvastatin model, was the 1µM dose used in Figure 2-5? If not, maybe the clusters of positive and negative for Annexin V correspond to the dose of atorvastatin? Yes, we used 1uM ATV, however we only performed ATV experiments in Figure 1-3, as discussed in methods and in response to your earlier question. Based on observation, we do not think ATV dosage affects severity of haemorrhage or haemorrhage-induced injury; rather it just affects the frequency of haemorrhage occurrence. We propose that once ICH occurs, the pathogenic response to blood in the brain is comparable between larvae, irrespective of ATV dosage. One of the major disadvantages of the larvae model is that it uses young animals. Is it possible to use atorvastatin to induce ICH in old zebrafish? We have used ATV to treat larvae at later time points to verify the findings in Eisa-Beygi et al (2013) and saw that haemorrhages were only induced when embryos are exposed to ATV prior to the critical neuroendothelial developmental time point at ~36-48hpf. Animals treated after this time do not haemorrhage. We believe that the use of young animals is an advantage, especially given the potential for impacting the 3Rs. Furthermore as the animals age, pigmentation develops and the larvae are no longer transparent – therefore this significant experimental advantage will be lost using older animals. Admittedly, ICH is predominantly a disease associated with aging; however this condition does occur in children and young adults, making our model even more pertinent for understanding those types of presentation. Ultimately, however, our article does strongly indicate that ICH induces brain injury phenotypes in young zebrafish that are comparable to the aged human brain. Furthermore, through our observations, we know that zebrafish larvae recover from ICH at an astonishing rate and therefore offer a potentially powerful tool for studying resolution and recovery processes in the future. Regarding modelling human disease, another difference is that there is no mortality in this model at least not to the time points investigated. Do they die at earlier time points than their wildtype/untreated littermates? The bbh mutants do not die due to ICH. Indeed, these animals develop into healthy, fertile adults, which we propose is due to either developmental processes or the regenerate properties of zebrafish. Either way, understanding this recovery at the molecular/cellular level could provide essential insight into how we may one day recover the injured human brain. Do higher dosages of atorvastatin induce mortality in the larvae due to ICH? In the ATV model, high doses of ATV will kill the larvae; however we propose this is due to toxicity rather than ICH-related. From the data where replicates were acquired, please specify whether the mean or median was chosen. Mean was presented throughout. This has now been clarified in the latest version of the manuscript. What is the % MS222 used as lethal overdose? 4% MS222 followed by maceration or freezing, which we have made clear in version 2. Kind regards Siobhan Crilly & Paul Kasher" } ] } ]	1	https://f1000research.com/articles/7-1617
https://f1000research.com/articles/7-1766/v1	08 Nov 18	{ "type": "Research Article", "title": "Effect of continuous-infusion antibiotic therapy on pulmonary function of patients with cystic fibrosis: A cross-sectional study", "authors": [ "Somaya Albhaisi", "Fei-Pi Lin", "Nauman Chaudary", "Fei-Pi Lin", "Nauman Chaudary" ], "abstract": "Background: Cystic fibrosis (CF) is associated with frequent pulmonary exacerbations which increase the mortality risk. Therefore, most CF patients are chronically colonized with respiratory pathogens, the most common being Pseudomonas aeruginosa. Multidrug-resistant organisms are a major problem in CF patients. It’s been hypothesized that continuous-infusion antipseudomonal beta-lactam therapy in CF maintains serum concentrations above the minimum inhibitory concentration of susceptible strains and is more likely than intermittent infusion to achieve optimal pharmacodynamic targets for some intermediate and resistant strains of P. aeruginosa. The most extensively studied antibiotic for continuous-infusion protocol in CF is ceftazidime, which has been shown to improve lung function (forced expiratory volume in 1 second and forced vital capacity) and to increase pulmonary exacerbation free time. There have been no studies to evaluate the cost effectiveness or impact on quality of life of continuous infusion versus intermittent infusion antibiotic therapy in patients with CF. Our study aims to investigate the effect of continuous-infusion antibiotic therapy on pulmonary function. Methods: Cross sectional study of CF patients who were admitted to our hospital with acute pulmonary exacerbations between 1/1/2010 and 12/31/2016 and received parenteral antibiotics. We investigated the effect of use of continuous versus intermittent infusion of intravenous antibiotics on the pulmonary function (FEV1% predicted). Results: Intermittent infusion protocol was found to have a very small advantage over continuous infusion protocol on pulmonary function; however this difference is not statistically significant (p=0.0049). The longer the duration of antibiotics, the slightly better the pulmonary function at the end of the treatment, but the difference was not significant (p=0.2543). Conclusions: Even though we could not draw meaningful conclusions from our data, we would like bring attention to this subject because it carries an important therapeutic value for CF patients.", "keywords": [ "Cystic Fibrosis", "Pulmonary Function Test", "Continuous Infusion", "Intermittent Infusion", "Antibiotics", "Forced expiratory volume" ], "content": "Introduction\n\nCystic fibrosis (CF) patients commonly suffer infections from multidrug resistant organisms, which increases their risk of treatment failure as a result of inability to meet pharmacodynamic targets (time above the minimum inhibitory concentration (T > MIC))1. Continuous-infusion antibiotic therapy is believed to be more effective in achieving those targets for resistant organisms in comparison to intermittent infusion.\n\nStudies have shown that ceftazidime, which is the most studied antibiotic for continuous infusion in CF patients, could improve forced expiratory volume in 1 second (FEV1) and forced vital capacity (FVC), and reduce the frequency of pulmonary exacerbations2. Continuous infusion has the potential to optimize the efficacy and safety of antimicrobial treatment during CF pulmonary exacerbations while potentially decreasing the costs of therapy3.\n\nDespite the promising results of studies comparing the two infusion protocols, there is insufficient evidence recommend the routine use of continuous infusion for patients with pulmonary exacerbations, which supports the position of the Cystic Fibrosis Foundation on this matter4. There is little information regarding the impact of continuous infusion on quality of life in patients with CF. Our study aims to investigate the effect of continuous-infusion antibiotic therapy on pulmonary function.\n\n\nMethods\n\nThe study was reviewed and approved by the Institutional Review Board (IRB) of Virginia Commonwealth University (approval number HM20011248). The data was analyzed in its entirety by the investigators who are fully responsible for the data and conclusion.\n\nThis was a cross-sectional study. The source of data was the CF Foundation Patient Registry. We obtained data for patients registered in our institute. We randomly selected 42 CF patients who were hospitalized for acute pulmonary exacerbations between 1st January 2010 and 31st December 2016 at Virginia Commonwealth University were retrospectively reviewed. We attempted to reduce potential bias by choosing a random sample of CF patients through the simple random sampling technique. We selected a small sample size by statistical formula that assumed p value is < 0.05 and 95% confidence Interval. In addition, we compared our study size with the previously published studies.\n\nInclusion criteria: (1) presence of CF pulmonary exacerbation, defined as: the need for antibiotic treatment as indicated by a recent change in at least two of the following: change in sputum volume or color; increased cough; increased malaise, fatigue or lethargy; anorexia or weight loss; decrease in pulmonary function by ≥10%; radiographic changes; increased dyspnea or at the physician’s discretion. (2) Age ≥ 18 years; (3) Patient has stable vital signs; (4) Patient is not a ward of the state; (5) Suspected or proven infection; (6) The presence of a 10% decline in lung function as assessed by spirometry.\n\nExclusion criteria: (1) Age < 18 years; (2) Critically ill patients or patients with unstable vital signs; (3) Inmates; (4) Chronic respiratory failure (PaCO2 > 60 mm Hg as outpatient).\n\nFor each patient, we collected the following variables: age, gender, ethnicity, race, body mass index, date of admission, date of discharge, antibiotic name, antibiotic infusion protocol, antibiotic start and end dates, culture data, pulmonary function test upon initiation and completion of antibiotics.\n\nThe effect of use of continuous infusion of intravenous antibiotics on pulmonary function (FEV1% predicted) was investigated.\n\nThe continuous infusion protocol has been adopted in our institute since 2013. For each patient, we collected data related to the use of intermittent infusion protocol which was the adopted protocol in our institute until 2013. From 2013 onwards, we collected data related to the use of continuous infusion protocol in order to investigate the differences in pulmonary function between the two protocols.\n\nFrom each and every hospital stay (inpatient visits) for each patient, values of FEV1% predicted at the start (baseline) and end dates of the antibiotics are collected. The name of antibiotic therapies that were used, as well as the type of infusion protocol with start and end dates of each of them are also recorded.\n\nIn this study, we attempted to compare the difference in FEV1% predicted between the two infusion protocols and for each individual visit. The major confounding factor is that many patients receive the intermittent-infusion at the same time as the continuous-infusion protocol.\n\nThe primary outcome of the study was a change in FEV1% predicted upon completion of antibiotic therapy. Categorical variables are reported as numbers and percent. Comparisons between groups for continuous variables were made using student t-test, while categorical variables were compared using κ2 test. A nominal p-value of <0.05 was considered statistically significant. Data analysis was performed using SPSS 24.0 (IBM, Chicago, IL).\n\n\nResults\n\nA total of 42 patients met entry criteria and were enrolled into the study. The study cohort consisted of 21 female and 21 male patients with a mean age of 33.9 years (age range 22–61). Mean FEV 1 upon initiation of antibiotics was 0.82 (FEV1/FVC ratio 43%), mean FEV1 upon completion of antibiotics was 1.9 (FEV1/FVC ratio 53%) (Table 1).\n\nIntermittent infusion protocol was found to have a very small advantage over continuous infusion protocol on pulmonary function (based on FEV1 outputs); however, this difference is not statistically significant (p=0.0049).\n\nThe longer the duration of antibiotics, the slightly better the pulmonary function at the end of the treatment, but the difference was not significant (p=0.2543). Patients with a history of multi-drug resistant organisms had lower pulmonary function test values and difficulty regaining their pulmonary function after treatment. As expected, patients who had higher pulmonary function test values before starting the antibiotics, had higher test values upon completion of the antibiotics, regardless of the infusion protocol (p<0.0001). When matched for other variables, older patients (age ≥ 65 years) had worse pulmonary function despite antibiotic therapy (p=0.0248). Those who had longer duration of hospital stay were noted to have a worse pulmonary function despite antibiotics (p=0.0140). African Americans had worse pulmonary function tests compared to Caucasians, but the difference was not significant (p=0.8998). Similarly, females had worse pulmonary function tests compared to males, but the difference was not significant (p=0.0930).\n\n\nDiscussion\n\nThe duration that β-lactams concentrations exceed the MIC for the bacteria determines its activity. Therefore, the current practice of intermittent infusion of β-lactam antibiotics may not be the optimal administration technique in CF patients. Moreover, many antibiotics have relatively short half-lives in this patient population.\n\nThere were several limitations to our study, which included the small sample size, the difficulty identifying which antibiotic protocol was adopted for each and every patient, interrupted or incomplete continuous infusion protocol and the significant difference in underlying patient demographics. We never collected or measured the beta lactam levels which is a huge limiting factor as well, as beta lactams need to be up titrated based on serum levels which we never had in this study. Without checking beta lactam levels, continuous infusion protocols seemed suboptimal to utilize so our concept of giving the same dose slowly in CF may not work.\n\nIn our opinion, although the data showed a small advantage for administration of antibiotics via intermittent infusion protocol, this finding was not statistically significant. In reality, most patients who received the continuous infusion protocol had interruptions to the protocol due to many factors, such as taking breaks for their pulmonary rehabilitation or other personal reasons. Therefore, this protocol was almost never completed as intended.\n\nHowever, our conclusions are mainly based on crude associations more than statistical ones. We could not make meaningful conclusions due to the above mentioned limitations. Therefore, the study lacks generalizability.\n\nOur study is directing attention towards a very important subject which could revolutionize the management of CF in the future. More research is really needed.\n\n\nEthical approval\n\nThe study was reviewed and approved by the Institutional Review Board (IRB) of Virginia Commonwealth University, approval number HM20011248. Consent from participants was waived by the IRB because the study is a retrospective review.\n\n\nData availability\n\nF1000Research: Dataset 1. Raw data for ‘Effect of continuous-infusion antibiotic therapy on pulmonary function of patients with cystic fibrosis: A cross-sectional study’, 10.5256/f1000research.15598.d2222935", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nMany thanks to Le Kang, PhD from the Department of Biostatistics for his assistance with the data analysis.\n\n\nReferences\n\nPrescott WA Jr, Gentile AE, Nagel JL, et al.: Continuous-infusion antipseudomonal Beta-lactam therapy in patients with cystic fibrosis. P T. 2011; 36(11): 723–63. PubMed Abstract \| Free Full Text\n\nGrupper M, Kuti JL, Nicolau DP: Continuous and Prolonged Intravenous β-Lactam Dosing: Implications for the Clinical Laboratory. Clin Microbiol Rev. 2016; 29(4): 759–72. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nParkins MD, Elborn JS: Newer antibacterial agents and their potential role in cystic fibrosis pulmonary exacerbation management. J Antimicrob Chemother. 2010; 65(9): 1853–61. PubMed Abstract \| Publisher Full Text\n\nFlume PA, Mogayzel PJ Jr, Robinson KA, et al.: Cystic fibrosis pulmonary guidelines: treatment of pulmonary exacerbations. Am J Respir Crit Care Med. 2009; 180(9): 802–8. PubMed Abstract \| Publisher Full Text\n\nAlbhaisi S, Lin F, Chaudary N: Dataset 1 in: Effect of continuous-infusion antibiotic therapy on pulmonary function of patients with cystic fibrosis: A cross-sectional study. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15598.d222293" }	[ { "id": "42994", "date": "21 Jan 2019", "name": "Martin J Walshaw", "expertise": [ "Reviewer Expertise cystic fibrosis" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis study compared continuous with intermittent beta lactam therapy for the treatment of CF exacerbations. There are theoretical reasons why continuous therapy is preferred, as the authors state. It seems that some of their data for intermittent therapy are not contemporaneously matched with the continuous therapy, which could confound the results. Furthermore, the authors state that many of the continuous infusion admissions were interrupted - i.e. - they ended up becoming intermittent. In any event, the study showed no difference between the two methods of administration. It is incorrect to state that one method had \"a very small advantage\" or \"slightly better pulmonary function\" than the other, when in fact there was no statistical difference - by definition the outcome was the same whichever method of administration was used. The data in the results alluding to race, age, length of stay etc. are irrelevant, since it is not expressed in relation to which type of therapy was used. The main thrust of this paper should be that continuous and intermittent therapy are both equally efficiacious in the treatment of an exacerbation.", "responses": [] }, { "id": "43983", "date": "15 Feb 2019", "name": "Zhe Hui Hoo", "expertise": [ "Reviewer Expertise Cystic fibrosis", "medication adherence", "registry analysis (and other observational data)" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nSpecific comments for each question on the peer review form:\n1. In terms of study design, what is the justification for \"choosing a random sample\" (page 3) instead of analysing all available data? Also, the authors mentioned that a power calculation was done, but the results of the power calculation was not presented. The authors implied their sample size is adequate but then mentioned that \"small sample size\" as one of the limitations (page 4).\n2. In terms of details of methods and analysis, there are several issues that should be clarified. First, what are the justification for the inclusion and exclusion criteria, in particular 10% FEV1 decline as one of the inclusion criterion and chronic respiratory failure as an one of the exclusion criterion.\nSecond, the authors claimed that \"the continuous infusion protocol has been adopted in our institution since 2013\", but \"the major confounding factor is that many patients receive intermittent-infusion at the same time as the continuous-infusion protocol\". Presumably this refers to beta-lactam antibiotics only. Also, presumably continuous infusion protocol was introduced in 2013 but beta-lactam may still be given via intermittent-infusion (i.e. continuous infusion was not blanketly applied). Is it not possible to restrict the analysis to people receiving only 1 type of beta-lactam as a sensitivity analysis to try understand the potential bias from the subgroup of participants that receive both intermittent and continuous beta-lactams?\nThird, if the 42 participants are only a subset of those eligible, it is worth including a \"CONSORT diagram\" (though this is not a RCT) to help readers understand the flow from all eligible participants to participants that were finally analysed.\nFourth, what is the equation used to calculate % predicted FEV1?\nFifth, the authors claimed that \"the longer the duration of antibiotics, the slightly better the pulmonary function at the end of the treatment\" but what was the analysis method used to determine the correlation between treatment duration and extent of FEV1 improvement?\n3. In terms of statistical analysis, the authors \"attempted to compare the difference in FEV1 % predicted between the two infusion protocols and for each individual visit\". Given that a participant can have multiple courses of IV antibiotics, what is the method used to account for repeated measures within an individual? In the 'data analysis' subsection, only student t-test and chi-square test were mentioned and both these tests assume independence.\n\nAlso, for the results, it is important to display the effect size and confidence intervals, not just the p-values. The authors claimed that FEV1 difference is not statistically significant with p-value 0.0049 (in the abstract and the text in page 3), yet p-value <0.05 was considered statistically significant. Is there a typo?\n4. With regards to the source data, it would be helpful if the authors could include the key for the labels e.g. what are antibiotics Type 1 and Type 2? Also, labels such as Chest_PT and VEST were neither explained in the text nor elaborated in the data source file.\n5. With regards to the conclusion drawn, it is crucial to report the difference (and confidence intervals) between group, otherwise readers are unable to judge whether there is any advantage at all favouring either the intermittent or continuous infusion group. It is also crucial to use the correct analysis method in order to justify the conclusion drawn.\nOther general comments: 1. The author stated that \"our study aims to investigate the effect of continuous-infusion antibiotic therapy on pulmonary function\". It seems more accurate to describe the study aim as to investigate the effect on FEV1 recovery post IV treatment.\n2. Some of the sentences in the introductions and discussion may benefit from being re-phrased. For example, multidrug resistant organisms \"may increase\" the risk of treatment failure (introduction) and I think it is hyperbolic to say that continuous beta-lactam infusion \"could revolutionize the management of CF\".", "responses": [] }, { "id": "43451", "date": "26 Feb 2019", "name": "Ghassan Kamel", "expertise": [ "Reviewer Expertise COPD exacerbations", "COPD", "CF exacerbations", "quality improvement" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a very interesting study and raises very important point: optimal antibiotic therapy for CF exacerbations. As this is a major source of morbidity in CF patients and the introduction of antibiotics clearly improved outcomes in CF patients.\nI have several comments/points to make regarding this paper:\nWhat is the advantage of choosing a random sample as opposed to analysing all CF exacerbations? This is a retrospective study and larger sample would've increased the power. The authors enrolled CF patients from 2010-2016 and they mentioned that continuous infusion was introduced in 2013. They did not mention what percentage of patients received continuous infusion versus intermittent. It is unclear how the authors came to conclude that \"the longer the duration of antibiotics, but slightly better the pulmonary function at the end of the treatment\". p value for FEV1 change was 0.0049, however, the authors reported that this was not statistically significant.The raw data do not report that actual change in FEV1.\n\nOverall, it is understandable that the authors were unable to draw strong conclusions likely due to small sample size, however, further evaluation of the analysis maybe needed.", "responses": [] } ]	1	https://f1000research.com/articles/7-1766
https://f1000research.com/articles/7-1760/v1	07 Nov 18	{ "type": "Case Report", "title": "Case Report: Dentigerous cyst marsupialization for a child with Hunter’s syndrome", "authors": [ "Shaimaa Sabry", "Dalia Moheb", "Osama El Shahawy", "Dalia Moheb", "Osama El Shahawy" ], "abstract": "Hunter’s syndrome or mucopolysaccharidosis (MPS) type II is an inherited disorder caused by enzyme iduronate-2-sulfatase deficiency. It is characterized by involvement of the nervous, cardiovascular, respiratory, and musculoskeletal systems, along with numerous oral manifestations. This is a case report of an eight year-old girl diagnosed with Hunter’s syndrome, who was referred to the Pediatric Dentistry Department, Faculty of Dentistry, Cairo University with a chief complaint of hard swelling related to the lower left posterior area. Radiographic examination revealed well defined corticated radiolucency surrounding an unerupted lower left first molar. Aspiration was done and cytopathologic examination revealed cystic fluid mixed with blood. The case was diagnosed as a dentigerous cyst. Cyst marsupialization was done under general anaesthesia. From this case report we concluded that in Hunter’s syndrome patients more conservative approaches are more valuable. Regular dental follow up is advised to maintain good oral hygiene, and to detect any complications as early as possible.", "keywords": [ "Dentigerous cyst", "Marsupialization", "Mucopolysaccharidosis", "Hunter’s syndrome", "Case report." ], "content": "Introduction\n\nHunter syndrome, or mucopolysaccharidosis type II (MPS II) is a rare metabolic disorder. It was first described in 1917, named after physician Charles A Hunter1. It is inherited as an autosomal recessive trait, and results in lysosomal enzyme iduronate-2-sulfatase deficiency. Prevalence of MPS II-H has been reported to be 1 in 170,000 cases, and no predilection for sex and ethnicity has been found2. Recently enzyme replacement therapy has emerged as a new treatment. Supportive management and physical therapy are also very important in the management of MPS type II3.\n\nSystemic manifestations of Hunter’s syndrome are macrocephaly, moderate to severe developmental delay, dysmorphic facies, skeletal abnormalities, joint contractures, hepatosplenomegaly, cardiac valvular disease, as well as corneal clouding, and puffy eyelids. Intraoral manifestations include an enlarged tongue, hyperplastic gingivae, broad arches with interdental spacing, anterior open bite, hypoplasticity, peg-shaped teeth with delayed development and eruption4,5.\n\n\nCase report\n\nAn eight year-old girl diagnosed with Hunter’s syndrome. She was born to healthy, consanguineous parents as their second child. She was referred to the Pediatric Dentistry Department, Cairo University in June, 2017 with a chief complaint of hard swelling related to the lower left posterior area that can be easily felt on palpation. The patient has no previous dental history, it was her first dental visit.\n\nClinical examination showed a hard bony swelling obliterating buccal vestibule related to the unerupted lower left fist permanent molar. Adjacent primary teeth; lower left first and second primary molars were sound with normal mobility and no pain on percussion.\n\nRadiographic examination revealed well defined corticated radiolucency surrounding the unerupted lower left first molar. There was significant root resorption related to the roots of the first and second primary molars. Delayed eruption of the permanent first molars was also present, with the lower right molar displaced towards the inferior border of the mandible. Enlarged dental follicles of the lower second permanent molars, as well as shortened lower permanent incisors. (Figure 1).\n\nAspiration was performed using a sterile plastic syringe; cystic fluid mixed with blood was found (Figure 2). Examination of aspirated fluid was done performed using a light microscope, by a pathologist in the Department of Oral and Maxillofacial Pathology. It revealed red blood cells with cholesterol crystals. Final diagnosis was reached by exclusion. List of differential diagnosis was as follow: dentigerous cyst, unicyctic ameloblastoma, odntogenic keratocyst. Odontogenic keratocyst; was excluded because it gives white cheesy material on aspiration. Dentigerous cyst and unicyctic ameloblastoma, have similar clinical and radiographic presentation, but it was diagnosed as a dentigerous cyst as they are common in patients with Hunter’s syndrome.\n\nCyst marsupialization was the choice of treatment, and the patient was referred to her physician to be prepared for the surgical intervention under general anaesthesia. Two weeks later, surgery was performed by the dental team; oral surgeon, pedodontist and dental assistants. Marsupialization was performed and a drain was placed. The patient was scheduled for follow up visits every two weeks to check the drain for two months post-surgery.\n\nAfter a one year follow up period it was found that clinically on palpation, there was no hard bony swelling related to the lower left fist permanent molar, but the tooth was yet to erupt. Radiographically there was an increase in bone density around the lower left first permanent molar indicating normal bone healing and shrinkage of cyst size. The lower left first permanent molar had moved towards occlusal plane (Figure 3).\n\n\nDiscussion\n\nHunter’s syndrome patients suffer from permanent, progressive cellular damage which affects all organs and systems functioning6. The patient was very apprehensive, exhibited aggressive behavior, had a large tongue, and limited mouth opening. Consultation with the physician and general anesthesia team was done to ensure the surgical procedure would have least possible risk.\n\nDelayed development and eruption of teeth, dentigerous cysts, and enlarged dental follicles are common features in Hunter’s syndrome patients7. In the presented case enlarged dental follicles are evident in the lower second permanent molars. Enlarged dental follicles are due to pools of chondroitin sulfate B. These lesions contain dense, fibrous connective tissues and large amounts of acid mucopolysaccharides. These areas of destruction tend to worsen with age8.\n\nDowns et al.7 stated that cystically involved teeth in Hunter’s syndrome patients should be removed, however, in our case cyst marsupialization was considered for a number of reasons. First was to be more conservative, to avoid total loss of the first permanent molar. Second was to avoid risk of pathologic fracture of mandible, which may have occurred if cyst inoculation was performed. The enlarged tongue and limited mouth opening also made invasive surgery more difficult.\n\nMarsupialization is not an aggressive technique but it requires meticulous postoperative care. Wound infection and delayed healing were expected. A drain was placed at the site of surgery and both the patient and her parents were educated and motivated to apply good oral hygiene measures and to maintain the drain in place. Surprisingly, normal healing with no infection occurred. This agrees with Savitha et al.4 who stated that aggressive surgery is usually not recommended for Hunter’s syndrome patients to avoid any complications.\n\nAlthough, the presented treatment modality doesn’t eliminate the pathologic condition at the time of surgery, and requires multiple postoperative follow up visits, it provides a conservative, non-aggressive line of treatment with good healing for a systemically compromised, syndromic case.\n\n\nConclusion\n\nHunter’s syndrome is a complex medical condition that necessitates regular dental follow up to maintain good oral hygiene, and to detect any complications as early as possible. With regards to dental lesions, less aggressive with more conservative approaches are recommended.\n\n\nPatient perspective\n\nAlthough the treatment required long and exhausting follow up visits, the patient and her parents were pleased with the more conservative treatment performed. The hard swelling gradually disappeared with time, and the patient kept her tooth.\n\n\nConsent\n\nWritten informed consent for publication of the clinical details and images was obtained from the patient’s father.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgement\n\nWe want to thank all the team worked in this operation; general anaesthesia team, oral surgeon and dental assistants for their great effort. Our gratitude also extends to the patient and her parents, who were very cooperative.\n\n\nReferences\n\nHunter C: A Rare Disease in Two Brothers. Proc R Soc Med. 1917; 10(Sect Study Dis Child): 104–116. PubMed Abstract \| Free Full Text\n\nTuschl K, Gal A, Paschke E, et al.: Mucopolysaccharidosis type II in females: case report and review of literature. Pediatr Neurol. 2005; 32(4): 270–272. PubMed Abstract \| Publisher Full Text\n\nShah GS, Mahal T, Sharma S: Atypical clinical presentation of mucopolysaccharidosis type II (Hunter syndrome): a case report. J Med Case Rep. 2010; 4: 154. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSavitha NS, Saurabh G, Krishnamoorthy SH, et al.: Hunter's syndrome: A case report. J Indian Soc Pedod Prev Dent. 2015; 33(1): 66–8. PubMed Abstract \| Publisher Full Text\n\nAnekar J, C DN, A C R, et al.: A rare case of mucopolysaccharidosis: hunter syndrome. J Clin Diagn Res. 2015; 9(4): ZD23–ZD26. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChinawa J, Adimora G, Obu H, et al.: Clinical Presentation of Mucopolysaccharidosis Type II (Hunter's Syndrome). Ann Med Health Sci Res. 2012; 2(1): 87–90. PubMed Abstract \| Free Full Text\n\nDowns AT, Crisp T, Ferretti G: Hunter's syndrome and oral manifestations: a review. Pediatr Dent. 1995; 17(2): 98–100. PubMed Abstract\n\nLiu KL: The oral signs of Hurler-Hunter syndrome: report of four cases. ASDC J Dent Child. 1980; 47(2): 122–7. PubMed Abstract" }	[ { "id": "41616", "date": "17 Dec 2018", "name": "Nedal Abu-Mostafa", "expertise": [ "Reviewer Expertise Oral surgery" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis report presented a case of an eight year-old girl diagnosed with Hunter’s syndrome. She had hard bony swelling on the buccal vestibule related to the unerupted lower left first permanent molar. Radiographically: there was a well-defined radiolucency surrounding the unerupted lower left first molar with radio-opaque margin. Marsupialization was performed and a drain was placed. Follow up was done.\nThe case report is interesting however, I have some comments:\nSigns and symptoms of Hunter syndrome on the patient face weren’t mentioned. As well, there are no intra-oral and extra-oral clinical pictures. The part of cystic lining that has been removed during marsupialization should be taken as incision biopsy. So the final diagnosis should depend on the histo-pathologic examination The panoramic radiograph shows only very minimal movement of lower left 1st molar. The eruption of the tooth and other permanent 1st molars could be achieved if orthodontic treatment has been carried out after 3 months of marsupialization. Discussion does not mention if there are similar previous case reports of hunter syndrome with dentigerous cyst that was treated by marsupialization. The conclusion should reveal the effectiveness of marsupialization as a conservative method for treatment of dentigerous cyst with Hunter syndrome with evidenced bone deposition. However the written conclusion put first the necessity of follow up and oral hygiene which was not evaluated the case report!\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Partly\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Partly\n\nIs the case presented with sufficient detail to be useful for other practitioners? Partly", "responses": [] }, { "id": "41617", "date": "19 Dec 2018", "name": "Onur Şahin", "expertise": [ "Reviewer Expertise implantology", "oral and maxillofacial surgery", "TMJ disorders", "medication related osteonecrosis of the jaw", "orthognathic surgery" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI appreciate your submission. This study is of interest to dentistry and medicine, to those who diagnose and treat paediatric patients with Hunter's syndrome. But, the manuscript should be revised and the author should provide the more detail data, to give useful clinical information to readers. So, I suggest the following revisions and advice before accepting it for indexing:\nHow was Hunter's syndrome diagnosed in this patients? You should give more detailed data about patient's medical history. If patient's physical appearance and intra-oral views could be presented, your paper would be perfect. More information is needed on how to apply marsupialization. Bone removed? Or buccal bone resorbed? If the view of placed drain could be presented, your paper would be perfect. In these patients, orthodontic treatment may be needed to erupt the tooth after marsupialization. This should be added to the discussion section.\nHunter's syndrome is a very serious disease affecting children. Aggressive surgery is usually not advocated for Hunter's syndrome. This study was to present that marsupialization could be the first choice treatment for dentigerous cyst in preadolescents. It should be useful article to oral & maxillofacial surgeons and general dental practitioners, if the authors perform more research and revision.\n\nThank you again for your submission.\n\nIs the background of the case’s history and progression described in sufficient detail? Partly\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Partly\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes", "responses": [] }, { "id": "40386", "date": "02 Jan 2019", "name": "Iyad Hussein", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article presents a case of a dentigerous cyst in a rare case of Hunter's syndrome. Overall it is good. Some corrections are suggested here.\nParagraph 1 Line 4 Add \"Hunter’s syndrome clinically resembles other mucopolysaccharide disorders such as Hurler’s, San Filippo’s, Morquio’s, and Schere’s syndromes,which are autosomal recessive. Hunter’s syndrome is the exception being transmitted as an X-linked recessive disorder, thus mainly seen in males. Thus this does have a preference for males.\"\n\nThis is mentioned in the reference the authors quoted (number 7)1.\nParagraph 1 Line 6: sex - as above comment.\nParagraph 3: English needs rewording.\nI suggest the following: \"An eight year old girl, born as a second child to healthy consanguineous parents, was diagnosed with Hunter's syndrome. She was referred to the Paediatric Dentistry Department, Cairo University in June 2017 with a chief complaint of a hard swelling related to the lower left posterior area of her mandible, that was easily felt on palpation.\"\nParagraph 4 Line 2 should read: \"first\" not \"fist\"\nParagraph 4 Line 3 should read: \"The adjacent...\" not \"Adjacent..\"\nParagraph 5 Line 1 should read: \"revealed a well defined...\"\nParagraph 5 Line 8 should read: \"..lower permanent incisors were also noted (Figure 1).\"\nParagraph 6 Line 2 should read: \"Examination of the aspirated..\"\nParagraph 6 Line 6 should read: \"A list..\"\nParagraph 6 Line 8 should read: \"keratinizing odontogentic tumour\" not \"Odontogentic keratocyst\" according to Welbury et al2.\n\nParagraph 6 Line 11 should read: \"presentations\"\n\nIs the background of the case’s history and progression described in sufficient detail? Yes\n\nAre enough details provided of any physical examination and diagnostic tests, treatment given and outcomes? Yes\n\nIs sufficient discussion included of the importance of the findings and their relevance to future understanding of disease processes, diagnosis or treatment? Yes\n\nIs the case presented with sufficient detail to be useful for other practitioners? Partly", "responses": [] } ]	1	https://f1000research.com/articles/7-1760
https://f1000research.com/articles/7-318/v1	14 Mar 18	{ "type": "Research Article", "title": "Loss of Zbtb32 in NOD mice does not significantly alter T cell responses.", "authors": [ "William D. Coley", "Yongge Zhao", "Charles J. Benck", "Yi Liu", "Chie Hotta-Iwamura", "M. Jubayer Rahman", "Kristin V Tarbell", "William D. Coley", "Yongge Zhao", "Charles J. Benck", "Yi Liu", "Chie Hotta-Iwamura", "M. Jubayer Rahman" ], "abstract": "Background: We previously identified the transcriptional regulator Zbtb32 as a factor that can promote T cell tolerance in the Non-Obese Diabetic (NOD) mouse, a model of Type 1 diabetes. Antigen targeted to DCIR2+ dendritic cells (DCs) in vivo inhibited both diabetes and effector T cell expansion in NOD mice. Furthermore, Zbtb32 was preferentially induced in autoreactive CD4 T cells stimulated by these tolerogenic DCIR2+ DCs, and overexpression of Zbtb32 in islet-specific T cells inhibited the diabetes development by limiting T cell proliferation and cytokine production. Methods: To further understand the role of Zbtb32 in T cell tolerance induction, we have now used CRISPR to target the Zbtb32 gene for deletion directly in NOD mice and characterized the mutant mice. We hypothesized that the systemic loss of Zbtb32 in NOD mice would lead to increased T cell activation and increased diabetes pathogenesis. Results: Although NOD.Zbtb32-/- male NOD mice showed a trend towards increased diabetes incidence compared to littermate controls, the difference was not significant. Furthermore, no significant alteration in lymphocyte number or function was observed. Importantly, in vitro stimulation of lymphocytes from NOD.Zbtb32-/- mice did not produce the expected hypersensitive phenotype observed in other genetic strains, potentially due to compensation by homologous genes. Conclusions: The loss of Zbtb32 in the NOD background does not result in the expected T cell activation phenotype.", "keywords": [ "Zbtb32", "ROG", "NOD", "diabetes", "CRISPR", "CRISPR/Cas9" ], "content": "Introduction\n\nAlthough some autoimmune diseases can be managed by the use of immunosuppressive drugs, current treatment options for individuals with Type 1 diabetes (T1D) are largely limited to controlling the disease symptoms instead of addressing the underlying autoimmune assault1. Reestablishing immune tolerance to the targeted self-antigens could provide more specific and durable treatment options2. The Non-Obese Diabetic (NOD) mouse strain spontaneously develops autoimmune diabetes, and can be used to identify how tolerance is defective in chronic autoimmune environments or to find potential therapies that can overcome these defects3,4.\n\nPreviously, we targeted self antigens to specific dendritic cell (DC) subsets by using chimeric antibodies that recognize antigen uptake receptors (either DEC205 for cDC1s5 or DCIR2 for cDC2s) with the relevant self peptide covalently attached6. The entire antibody is internalized and the antigenic self peptide is then processed and presented on MHC class II. This allowed identification of DC subsets able to induce tolerance against β-islet antigens within the context of chronic autoimmunity found in NOD mice. Using this approach, we demonstrated that self-antigen presentation by DCIR2+ DCs but not DEC205+ DCs could promote self-tolerance via increased apoptosis and decreased effector functions in islet-specific CD4 T cells6. A gene expression analysis of the affected autoreactive T cells showed increased early expression of Zbtb32 in T cells stimulated by DCIR2+ DCs. In addition, transient overexpression of Zbtb32 in islet-specific CD4 T cells delayed diabetes development, and decreased both proliferation and IFNγ production in islet-specific CD4+ T cells6. Together these studies suggest T cell expression of Zbtb32 plays a role in tolerance induction in NOD mice, and could be a target for treatment.\n\nThe Zbtb32 gene (also known as ROG or PLZP) was first identified as a homologue of the leukemia oncogene PLZF (Zbtb16)7. Zbtb32 was then shown to act as a transcriptional repressor of GATA-38,9. Since GATA-3 regulates thymocyte development and TH2 differentiation, Zbtb32 knockout mice were produced to test if Zbtb32 played a role in modulating lymphocyte development and response to stimuli. In three genetic backgrounds (129/SV, C57Bl6, and BALB/c), the homozygous loss of Zbtb32 significantly altered T cell responses resulting in increased proliferation, increased cytokine production, and hypersensitive responses to inflammatory stimuli10–13. NK cells and B cells from C57Bl6 mice lacking Zbtb32 similarly displayed hyperproliferation and increased effector functions14,15. Therefore, data from these knockout strains show that Zbtb32 functions as a brake for lymphocyte activation and differentiation, and is consistent with our data from transient overexpression of Zbtb32 in T cells.\n\nTo better understand the role of Zbtb32 for T cell tolerance in autoimmune NOD mice, we introduced a targeted deletion using CRISPR/CAS9 techniques. The creation of mutant NOD strains has been severely limited by the lack of a robust ES line, necessitating backcrossing knockouts made in other genetic backgrounds onto the NOD background16,17. Because the region surrounding the knockout locus will retain alleles from the original strain, this can confound interpretation of such mice. NOD mice carry dozens of genetic susceptibility loci that contribute to the onset of T1D and some of those linked alleles could alter diabetes susceptibility or immune phenotype of any new mutant strain18. These issues can be avoided by using CRISPR/CAS9 technology to directly edit DNA in NOD embryos, an approach utilized only very recently in the NOD strain19. For the present study, we deleted a portion of exon 2 to cause a frameshift mutation in the Zbtb32 gene, and validated the loss of the Zbtb32 protein in our colony of NOD.Zbtb32-/- mice. These mice were used to test our hypothesis that the transcription repressor Zbtb32 plays a critical role in the inhibition of T-cell mediated autoimmune T1D. Surprisingly, T cells from NOD.Zbtb32-/- mice were not hyperreactive as would have been expected based on both our prior overexpression data and knockout mice in other strains. Although male NOD.Zbtb32-/- mice showed a trend of increased diabetes incidence compared to littermate controls, the difference was not significant, and no significant difference in either time of onset or overall incidence was observed in female NOD.Zbtb32-/- mice. Overall, most of our experiments suggested mild phenotypic changes perhaps as a result of either compensation by other family members or by effects on multiple cell types that is different on the NOD genetic background.\n\n\nMethods\n\nNon-Obese Diabetic mice (NOD, JAX #001976) were initially obtained from The Jackson Laboratory, Bar Harbor ME, and bred in the facility at NIH. Additional NOD females were regularly delivered every two weeks to compensate for the loss of breeders due to diabetes in the wildtype NOD colony. Approximately 630 NOD.Zbtb32-/- mice (about 90 litters with an average liter size of 7) were bred during the characterization the mutant strain. All mice were housed in a specific pathogen free vivarium with a 12 h on/12 h off light cycle (6 am on and 6 pm off). The care, use, and disposition of all mice used in this study were reviewed and approved (protocol K024-DEOB-16) by the Institutional Animal Care and Use Committee of the NIDDK, NIH. Diabetes was monitored and mice were considered diabetic on the first of two blood glucose level measurements above 250 mg/dl.\n\nThe NOD.Zbtb32-/- genetic alteration was carried out as follows: NOD female mice (6–8 weeks old) were super ovulated and mated overnight with NOD male mice (>8 weeks old). Zygotes were harvested from the ampullae of super ovulated females and were placed in KSOM medium (Millipore, Billerica MA) before microinjection. Microinjection was performed in M2 medium (Sigma, St Louis, MO) using a micromanipulator (Narishige) and microscope (Nikon). The second exon of the Zbtb32 gene was targeted using the two following RNA guide sequences; 5'-UACAG UUAGC GGCUA GACUC-3' and 5'-CAAUC AUGGA UCCCC CAUUG-3'. Binding sites for single guide RNA (sg RNA) are indicated in Figure 1A. The modified Cas9n enzyme with nickase activity (System Biosciences) and sgRNA were co-injected into the pronucleus and cytoplasm of 238 zygotes (1st round 37, 2nd round 47, and 3rd round 154) and resulted in a total of 35 pups. The final injection concentration in the mixture was 10 ng/μl Cas9n and 5 ng/μl of each sgRNA. After injection and incubation in 5.5% CO2 at 37°C overnight for 24 h, surviving 2-cell stage embryos were transferred to female NOD pseudopregnant recipients via oviduct transfer.\n\nThe CRISPR/Cas9 technique was used to target exon 2 of the Zbtb32 gene directly in NOD mice. A portion of the Zbtb32 gene was deleted and a frameshift introduced in NOD mice using the CRISPR/Cas9 technique, as illustrated here (A). The scissors represent the location of cutting based on sgRNA binding and black arrows indicate binding sites for the standard primers, which were used for both sequencing and genotyping. The location of the deletion maps onto the beginning of exon2 of the wildtype mRNA (B). Non-coding portions are shown as thinner than the coding portions. The CRISPR/Cas9-treated embryos develop into mice with genetic mosaicism of the Zbtb32 gene in the F0 generation. Four genetically heterogeneous females (indicated with black arrows) were examined further before one mutant allele was chosen to establish the NOD.Zbtb32-/- colony (C). The targeted deletion of Zbtb32 removed 146 bp from exon 2 as shown by genetic screening of knockout, heterozygous, and wildtype animals in the F2 generation (D). The loss of the Zbtb32 protein after CRISPR-mediated deletion was confirmed by Western blotting extracts from stimulated CD4+ splenocytes (E) (from three biologically independent replicates).\n\nGenomic DNA was extracted from tail snips of the 35 21 day-old F0 mice using the DNeasy blood and tissue kit (Qiagen). The target region was PCR amplified by PuReTaq Ready-To-Go PCR beads (GE Healthcare) followed by IndelCheck CRISPR/TALEN insertion or deletion detection system (GeneCopoeia). Successful genetic editing of the Zbtb32 gene was observed in 15 of the 35 pups. Genomic DNA from pups with evidence of deletion was further PCR amplified and subsequently cloned by TOPO T/A subcloning (Invitrogen, Carlsbad, CA) followed by DNA sequencing. Compound heterozygous null founders were bred with wildtype NOD animals to provide four potential alleles (named A, B, C, and D) for the creation of a NOD.Zbtb32-/- colony. All four mutant alleles were sequenced to verify a successful deletion (Supplementary Table 1). As expected, the mutant alleles displayed a variety of deletion sizes due to differential DNA repair. Sibling crosses with matching mutant alleles were set up but, only mice carrying the mutant B allele were reliable breeders. Therefore, all NOD.Zbtb32-/- mice described in this paper are homozygous for the B allele.\n\nFor routine genotyping and sequencing, we used the following PCR primers: forward 5’-AGCTG GCCTT TGGCT TAGTT-3’ and reverse 5’-CAAAG GTGGA AGGGC TTATG-3’. Binding sites for these primers in relation to the CRISPR/Cas9 deletion are indicated in Figure 1A. PCR conditions followed a typical 3-temp program; 95°C for 5 min, [95°C 30s, 55°C 30s, 72°C 45s] repeat 34 times, and finally 72°C for 10 min. The PCR results were run out on a 2% agarose gel. Single pass DNA sequencing services (ACGT, Inc., Germantown, MD) were carried out on genomic DNA using the reverse primer.\n\nDuring dissections, mice were euthanized by CO2 asphyxiation and then the spleen was removed and placed in ice-cold PBS. Splenocytes were negatively enriched for CD4+ T cells using magnetic beads via the mouse CD4+ T Cell Isolation Kit (Miltenyi Biotech, Bergisch Gladbach, Germany) and then were incubated with Armenian Hamster anti-mouse CD3 (clone 145-2C11, Biolegend cat#100301) and Syrian Hamster anti-mouse CD28 (clone 37.51, Biolegend cat#102111) for either 14 h or 24 h at 37°C in T-cell Media (RPMI-1640+10% FBS+L-glutamine +Penn/Strep). The cells were then pelleted and lysed used RIPA buffer supplemented with Protease Inhibitor Cocktail (Roche, Indianapolis, IN). A total of 35 µg of protein was loaded into a NuPAGE 4–12% Bis-Tris gel, transferred onto nitrocellulose, blocked in 5% milk for 1 hour at 25°C, and blotted for Zbtb32 using rabbit anti-mouse sc-25358 (Santa Cruz Biotech, Santa Cruz CA) at a 1:500 dilution overnight at 4°C. The membrane was then washed prior to the addition of goat-anti-rabbit HRP (Jackson Immunoresearch cat#111-035-003) at a dilution of 1:5,000 dilution at 1 hour at 25°C. The membrane was then stripped, reblocked and reblotted for Lamin B using goat anti-mouse sc-6217 at a 1:5,000 dilution for 1 hour at 25°C. The membrane was then washed prior to the addition of donkey anti-goat HRP (Jackson Immunoresearch cat#705-035-003) at a dilution of 1:5,000 dilution for 45 minutes at 25°C. All protein bands were imaged on a BioRad ChemiDoc CCD system and processed with the BioRad Image Lab 6.0 software.\n\nSplenocytes and lymphocytes were stimulated with anti-mouse CD3 (145-2C11; BioLegend, San Diego, CA). A total of 5×105 cells were incubated for 18 hours at 37°C in T-cell media with either 0, 10, 100, or 1000 ng of anti-CD3 in a 300 µl volume. For cytokine staining, cells were stimulated with PMA and Ionomycin in T-cell Media for 5 hours at 37°C. Brefeldin A was added to all cells for the final 4 hours of stimulation. Cells were then fixed and stained for flow cytometry. FMO staining controls were performed using a mixture of both unstimulated and stimulated cells.\n\nAll cells were washed in Wash Buffer [PBS+2% FBS] and then blocked with LEAF purified anti-CD16/32 (clone 93) on ice for 30 minutes prior to staining for flow cytometry. The Foxp3 (FJK-16s) antibodies were purchased from eBioscience (San Diego, CA). Foxp3 antibodies were diluted at 1:50 in Wash Buffer for cell staining, while all other antibodies were diluted at 1:100 in Wash Buffer. Staining was carried out for a minimum of 1 hour at 4°C. Antibodies purchased from Biolegend (San Diego, CA) are listed alongside their clone and Antibody Registry numbers. We used antibodies raised against mouse B220 (RA3-6B2, RRID:AB_11203907), CD1d (CD1.1, RRID:AB_2715919), CD4 (GK1.5, RRID:AB_312690), CD5 (53-7.3, RRID:AB_312736), CD11b (M1/70, RRID:AB_312798), CD8 (53-6.7, RRID:AB_2561352), CD19 (6D5, RRID:AB_130884), CD23 (B3B4, RRID:AB_10060129), CD25 (PC61, RRID:AB_2562611), CD43 (S11, RRID:AB_2563698), CD44 (IM7, RRID:AB_2621762), CD62L (MEL-14, RRID:AB_10555750), CD69 (H1.2F3, RRID:AB_10683447), CD335 (29A1.4, RRID:AB_10827686), IFNγ (XMG1.2, RRID:AB_469504). Cell viability was assessed by Fixable Live/Dead Aqua staining (Life Technologies, Grand Island, NY) according to the manufacturer's protocol. Cell proliferation was measured using the Cell Trace Violet Kit (Life Technologies, Grand Island, NY) according to the manufacturer's protocol. For cell fixation, permeabilization, and intracellular staining, we utilized the Intracellular Fixation & Permeabilization Buffer Set from eBiosciences. All samples were run on a BD Biosciences LSR-II cytometer and later analyzed using FlowJo v9.4.1. For analysis, CD44 expression was measured by mean fluorescence intensity (MFI) rather than by the percentage of positive expression.\n\nStatistical analysis for all experiments was carried out using GraphPad Prism v5. Where appropriate, statistical significance was calculated using either a Kaplan-Meier survival method (for disease incidence), or a two-way ANOVA tests for independent samples with post hoc Bonferroni comparisons for each possible pair of groups (used for all other data). Any significant differences (P < 0.05) between the two control strains are denoted with an asterisk. All dot plots show exact data points with a horizontal line denoting the average. All histograms show the average and standard deviation for the data.\n\n\nResults\n\nThe NOD mouse is an important model for examining spontaneous and chronic autoimmune disease, but it has been difficult to obtain genetically modified NOD mice due to the lack of a robust NOD ES cell line19. Directly targeting mutations into the NOD genetic background via CRISPR gene editing in NOD embryos is now possible20. Matching sgRNAs in exon 2 of Zbtb32 (Figure 1A and B) and CAS9n were injected into 2 cell-stage embryos. Using the CAS9n with nickase activity ensures that an insertion or deletion can only occur if CAS9n binds to both sgRNAs, vastly reducing off-target events21. Four female mice carrying a suitable deletion were selected to be potential F0 founders (indicated in Figure 1C) and were crossed with wildtype NOD mice to produce heterozygous mice in the F1 generation. Two of these potential F0 founders were found to be germline compound heterozygous for the loss of Zbtb32, providing up to four potential alleles (listed in Supplementary Table 1 as alleles A, B, C, and D) for the creation of a NOD.Zbtb32-/- colony. Sibling crosses with matching mutant alleles were set up using the F1 generation. Because mice carrying the mutant B allele were the most reliable breeders, all NOD.Zbtb32-/- mice described in this paper are homozygous for the B allele containing a 146 bp frameshift deletion.\n\nWe designed PCR primers to serve for both routine genotyping purposes (demonstrated in Figure 1D) and as sequencing primers. Genomic sequencing on select NOD.Zbtb32-/- mice from multiple generations verified that the deletion within the Zbtb32 gene was stable (Supplementary Figure 1). In wildtype NOD mice, Zbtb32 was not detectable at baseline in resting CD4+ lymphocytes, but could be detected by Western blot as early as 14 hours post anti-CD3/CD28 stimulation (Figure 1E). No Zbtb32 protein was observed in NOD.Zbtb32-/- mice at either baseline or post-stimulation (Figure 1E).\n\nTo test the hypothesis that absence of Zbtb32 would increase diabetes pathogenesis in NOD mice, we monitored diabetes via blood glucose levels in NOD.Zbtb32-/- mice and their control littermates for up to 35 weeks for both female (Figure 2A) and male mice (Figure 2B). NOD mice normally spontaneously develop hyperglycemia, between 12 and 30 weeks of age due to autoimmune destruction of the insulin secreting β-islet cells within the pancreas. The diabetes incidence rates showed no statistical differences between NOD.Zbtb32-/- mice and their wildtype littermate controls when using a standard Kaplan-Meir survival test, but the male NOD.Zbtb32-/- trended toward a higher than expected total diabetes incidence (Figure 2C). Therefore, the trend in higher male disease incidence may indicate a mild phenotype in NOD.Zbtb32-/- mice may impact disease onset over long timescales.\n\nThe onset of Type 1 diabetes was monitored in our NOD.Zbtb32-/- colony by weekly blood glucose monitoring for both female (A) and male (B) mice. Disease incidence rates were followed for littermate wildtype, and homozygous knockout mice. Disease incidence rates for males and females were plotted in (C).\n\nPotential differences between wildtype and Zbtb32-/- lymphocyte phenotype of both male and female mice at 8 weeks of age were tested using flow cytometry. Activation markers on Foxp3+ regulatory T cells (Tregs), conventional CD4+ T cells (Tcons), and CD8+ T cells were measured without exogenous stimulation. As expected, higher activation was measured in the pancreatic-draining lymph nodes (pLNs) (Figure 3) compared to the spleen (Supplementary Figure 2), consistent with an autoimmune reaction to antigens acquired from pancreatic tissue. However, no statistically significant differences were observed between the NOD.Zbtb32-/- mice and their control littermates for percent CD25+ (Figure 3A), CD69+ (Figure 3B), or CD44 MFI (Figure 3C) (and Supplementary Figure 2A – C). T cells from the NOD.Zbtb32-/- pLN, but not the spleen were more variable than the wildtype for these phenotypes, and in some comparisons trend higher (e.g. elevated CD25 expression in female CTLs seen in Figure 3A), suggesting a mild activation phenotype.\n\nLymphocytes from the pancreatic draining lymph node of male and female NOD and NOD.Zbtb32-/- mice were stained for activation markers. Flow cytometry was used to determine cell surface expression of CD25 (A) CD69 (B) and CD44 (C) on CD4+ Foxp3+ (Tregs), CD4+ Foxp3- (Tcon) and CD8+ T cells. The data were combined from two independent repeats of this experiment.\n\nIn previous reports describing Zbtb32 gene deletion in other genetic backgrounds, reported phenotypes included T-cell hypersensitivity, increased proliferation, and increased cytokine production10–12,14. We therefore carried out parallel experiments using splenocytes from the NOD.Zbtb32-/- mice. Splenocytes from both NOD.Zbtb32-/- mice and NOD wildtype mice were stimulated with anti-CD3, and cell activation was assessed after 18 h by staining for surface CD69 (Figure 4A–C), CD44 (Figure 4D–F) and CD25 (Supplementary Figure 3A–C). T cell activation markers increased with anti-CD3 stimulation, but no significant differences between wildtype and Zbtb32-/- lymphocytes were observed. Levels of cell proliferation after 48 hours of anti-CD3 stimulation, as measured by dilution of Cell Trace Violet dye, were also the same between NOD wildtype and Zbtb32-/-(Supplementary Figure 3D and E). We next measured cytokine production by stimulating splenocytes with PMA and Ionomycin, and detecting IFNγ production in CD4+, CD8+, and Nkp46+ cells via intracellular cytokine staining. In all cases, the stimulation with PMA and Ionomycin produced the expected production of IFNγ, with higher expression in CD8+ and Nkp46+ cells. However, we did not find statistically significant differences between NOD.Zbtb32-/- splenocytes and their matching controls (Figure 4G–I). Therefore, unlike other Zbtb32-/- strains, ex vivo T cell responses from NOD.Zbtb32-/- mice were not significantly altered.\n\nFreshly isolated splenocytes from female NOD.Zbtb32-/- mice and control littermates were stimulated with the indicated dosages of anti-CD3 for 18 h. Surface expression of CD69 (A–C) and CD44 (D–F) was measured on CD4+, CD8+, and NKp46+ splenocytes. For IFNγ staining, splenocytes from female NOD.Zbtb32-/- mice and their control littermates were stimulated with PMA and Ionomycin for 4 hours (G–I). The percentage of cells positive for IFNγ in CD4+ (G), CD8+ (H), and NKp46+ cells (I) is shown. The data were combined from two independent repeats of this experiment.\n\nBecause the T cell and diabetes phenotypes did not match the expected result based on overexpression of Zbtb32 in T cells6, potential effects of Zbtb32 deletion on other cell types was considered. B1 cells, which primarily reside in the peritoneal cavity, display a constitutively high expression of Zbtb3222,23. A related transcription factor, Zbtb20, was recently implicated as a regulator of terminal differentiation in germinal center B cells24. The authors proposed that either Zbtb20 or Zbtb32 may play a critical role in either B1 cell proliferation or differentiation. B1 cells can be further divided into B1a and B1b cells based on their surface expression of CD525. In NOD mice, B1 cells can be identified as CD19+B220+CD11b+CD43+, with CD5+ and CD5- subpopulations being labeled as B1a and B1b, respectively (Figure 5). Since prior reports had indicated that Zbtb32 can regulate early B cell proliferation26, B1 populations in the peritoneal cavity of NOD.Zbtb32-/- mice and their wildtype controls were examined at 4 weeks of age (Figure 5). NOD.Zbtb32-/- mice trended toward fewer total B cells in their peritoneal cavity. But no statistically significant differences were observed in the proportion of B1 cells among all CD19+ cells between wildtype NOD and NOD.Zbtb32-/- mice (analysis not shown). Therefore, the mild diabetes phenotype and the absence of the expected T cell activation phenotype are unlikely to be due to effects on B1 cells.\n\nThe number of B cells in the peritoneal cavity of 4 week-old mice were measured, including all CD19+ cells, all B1 cells, and the subpopulations of B1a and B1b cells. The data were combined from two independent experiments.\n\n\nDiscussion\n\nWe previously focused on Zbtb32 as a potential regulator of T cell tolerance in the NOD mouse model, and found that this transcription regulator was preferentially induced in autoreactive CD4+ T cells after interacting with tolerogenic DCIR2+ DCs, and that overexpression in T cells inhibited both diabetes and effector T cell expansion6. The available literature corroborated these findings but did not provide insight into how Zbtb32 might control T cell development and tolerance in a spontaneous autoimmune disease such as T1D9–11,27. We initially hypothesized that the Zbtb32-/- mutation in NOD mice would result in hyperactive T cells and increased diabetes. Instead, no significant changes in disease incidence in the NOD.Zbtb32-/- mice were observed, and the increased T cell activation and differentiation observed in mutants from other genetic backgrounds were absent. No significant shifts in the number or percentage of the major lymphocyte populations were observed. Furthermore, we observed that the loss of Zbtb32 had no impact on intraperitoneal B1 cells, despite the fact that it is constitutively expressed in those cells in wildtype animals. In most of these readouts, a trend towards the expected phenotype was observed, suggesting a very mild phenotype, with only the long-term male disease incidence rates provided any evidence that the loss of Zbtb32 affected T1D disease pathogenesis. Although NOD females normally have higher incidence than males, this phenotype can vary greatly, and 50% incidence is within range of what we have observed for our colony (20–50% in males and 40–90% in females).\n\nPrior investigations of Zbtb32 knockout mice in other genetic backgrounds consistently showed an activated T cell phenotype that included hypersensitivity, increased cytokine production, and increased proliferation10–14. Furthermore, our results showed that overexpression of Zbtb32 delayed diabetes onset, limited T-cell proliferation, and decreased IFNγ production from autoreactive T cells in an adoptive transfer model of T1D6. There are several possible reasons that may account for the observed lack of expected phenotype. One potential explanation for our unexpected results is that the CRISPR/Cas9 mediated deletion produced either an off-target effect or otherwise caused genetic instability at the deletion site. The use of CRISPR techniques in the NOD genetic background is a new approach that has, to our knowledge, only been described in two previous publications to date19,28. The CRIPSR/Cas9 mediated deletion was significantly smaller than other previously engineered Zbtb32 deletions (Compare Figure 1B to Supplementary Figure 4), but still disrupted the reading frame of the Zbtb32 gene. Our western blots (Figure 1E) demonstrated that the protein had been eliminated, and the use of CAS9n minimized other potential mutations. Nevertheless, we re-examined our colony for any evidence of genetic instability or off-target effects by re-sequencing archived genomic DNA from founder mice and five subsequent generations of sibling crosses and found that the mutations remained stable (Supplementary Figure 1) in every sequenced member of our NOD.Zbtb32-/- colony. To identify potential off-target effects, we compared sequence homologies among all the transcription factors in the Zbtb family as well as sequence homologies for just the guide RNA strands used by the Cas9n enzyme. Both homology analyses indicated that only Zbtb16 was a potential target for unintended genetic editing. However, sequencing the Zbtb16 gene in multiple generations of the NOD.Zbtb32-/- mouse colony revealed no off-target mutations of Zbtb16 (data not shown). These results therefore argue against the idea that our overall lack of phenotype can be explained by erroneous CRISPR/Cas9 activity.\n\nGenetic backgrounds can have a dramatic effect as genetic alterations combine to either mask or exacerbate disease phenotypes29. Interestingly, the role of Zbtb32 in immune tolerance was examined in the BALB/c strain by using the Experimental Autoimmune Encephalitis (EAE) model. The loss of Zbtb32 in BALB/c mice had no significant impact on the EAE clinical score or timing of onset of paralysis suggest that the loss of Zbtb32, despite the observed activated T cell phenotype, was not sufficient to precipitate severe autoimmunity11. A rapid and induced autoimmune model like EAE differs significantly from a spontaneous, chronic disease model like NOD, but the parallel results between our findings and the findings in Kang et al.11 suggest that, in some contexts, other genes may be able to compensate for the loss of Zbtb32. Further support for this explanation comes from an examination of regulatory transcription factors in TH17-mediated autoimmune pathology30. The authors found that although Zbtb32 gene expression was highly correlated with a protective phenotype, the loss of the Zbtb32 gene in the 129/Sv background did not produce the expected hypersensitive TH17 response and instead observed a diminished TH17 response after repeated stimulations. Based on their gene expression data, Gaublomme et al. hypothesized that while Zbtb32 normally suppresses pro-inflammatory responses, other unidentified transcription factors continued to upregulate pro-regulatory genes and therefore mask the expected phenotype in TH17 responses30. The NOD.Zbtb32-/- mice differed from past knockout strains in terms of the T cell response. Examinations of lymphocytes from the pLNs (Figure 3) and the spleen (Supplementary Figure 2) revealed only mild differences between wildtype and knockout littermates in NOD mice. Considering that the target pancreatic antigens are trafficked into the pLNs and by 8 weeks of age NOD mice display chronic autoimmunity and lymphocytic infiltration in the pancreatic islets31,32, we had expected to find increased levels of lymphocyte activation in the pLNs. The major phenotype observed in other Zbtb32-/- strains was a hyperreactivity to in vitro T cell stimulation, manifest in higher proliferation and cytokine production10,11,14,26. However, T cells from NOD.Zbtb32-/- mice showed no increase in proliferation or IFNγ production in response to in vitro stimulation.\n\nIt is possible that the lack of phenotype in the NOD genetic background, but not other backgrounds, was due to compensation from homologous proteins. The loss of Zbtb32 from birth could activate compensatory mechanisms in NOD mice that obscure any overt abnormalities in adult mice. It is even possible that a compensatory mechanism was selected for in the course of breeding NOD.Zbtb32-/- mice. Our incidence data includes one mutant mouse from our first generation of the colony that developed hyperglycemia at an abnormally early age (8 weeks), and three out of the four mutated alleles proved to be poor breeders. Zbtb32 is constitutively expressed in the testis33, but its role in fertility is not clear34. These unexpected results could potentially be explained by compensatory mechanisms that were either already present within the NOD genetic background, or were selected for during the generation of this NOD.Zbtb32-/- mouse colony. Therefore, an alternative approach targeting expression in adults could yield a different result.\n\nTaken together, our results show that the systemic loss of Zbtb32 in NOD mice does not lead to a hypersensitive T cell phenotype and increased diabetes pathogenesis. Although the NOD.Zbtb32-/- male mice displayed some increase in diabetes incidence compared to littermate controls, these trends did not reach statistical significance. Therefore, we conclude that NOD mice with the loss of Zbtb32 have a mild phenotype that may be a result of compensation by a currently unknown mechanism. In a broader context, these data do not support a major role for Zbtb32 as a target for treatment of autoimmune diabetes.\n\n\nData availability\n\nDataset 1: Raw images for Figure 1 The raw images for the genotyping gels and western blots seen in Figure 1. 10.5256/f1000research.13864.d19744135\n\nDataset 2: Raw data for blood glucose from Figure 2 The raw data values and the GraphPad Prism fole for blood glucose measurements. 10.5256/f1000research.13864.d19744236\n\nDataset 3: Raw flow cytometry data from 8 week-old mice: Raw flow cytometry sample data and the FlowJo analysis file for resting splenocytes and lymphocytes from 8 week old mice. 10.5256/f1000research.13864.d19744337\n\nDataset 4: Raw flow cytometry data from ex vivo stimulated splenocytes: Raw flow cytometry sample data and the FlowJo analysis file for ex vivo stimulated splenocytes, including cell proliferation and cytokine production 10.5256/f1000research.13864.d19744438\n\nDataset 5: Diabetes incidence in NOD.Zbtb32-/- mice 10.5256/f1000research.13864.d19744539\n\nDataset 6: Alignment 10.5256/f1000research.13864.d19744640", "appendix": "Competing interests\n\n\n\nKristin Tarbell is currently an employee of Amgen; all work described here was performed at NIDDK and has received no financial support from Amgen.\n\n\nGrant information\n\nThis work was supported by the Intramural Research Program of the National Institute of Diabetes and Digestive and Kidney Diseases, Project # 1ZIADK075065-06.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe would like to thank Mrs. Alice Franks (DEOB, NIDDK) for mouse colony management, the NHLBI/NIDDK flow core for assistance with flow cytometry, System Biosciences for help designing the sgRNA sequences, and Huiyan Lu for assistance with carrying out genetic manipulation and transference of embryos.\n\n\nSupplementary material\n\nSupplementary Table 1. Alignment of NOD.Zbtb32 mutant alleles. The four mutant alleles of Zbtb32 that were produced by CRISPR/Cas9 were then sequenced and aligned. Allele B was chosen to serve as the only mutation in the colony.\n\nClick here to access the data.\n\nSupplemental Figure 1. The CRISPR/Cas9 deletion remained stable in the NOD.Zbtb32 colony. We archived breeder genealogies and genomic DNA from all mice in the NOD.Zbtb32 mouse colony. A genealogy of breeding cages carrying the B allele of the mutated Zbtb32 gene is depicted in (A). A total of 16 archived DNA samples were sequenced and then aligned in order to verify that the mutation remained stable in the colony (B). These archived samples included 1 founder of the B line, 13 homozygous knockout descendants, and 2 homozygous wildtype offspring.\n\nClick here to access the data.\n\nSupplemental Figure 2. Activation markers from resting splenocytes in NOD.Zbtb32-/- mice. We examined populations of Treg, Tcon, and CD8+ T cells in freshly dissected spleens from both male and female mice and their littermate controls. Flow cytometry was used to determine the ratio of CD25+ (A) CD69+ (B) and cell surface expression of CD44 (C) on the indicated splenocytes.\n\nClick here to access the data.\n\nSupplemental Figure 3. Activation markers on ex vivo-stimulated lymphocytes from pancreatic draining lymph nodes. After isolating lymphocytes from the pancreatic draining lymph nodes of NOD.Zbtb32-/- mice and stimulating them with anti-CD3 as indicated, we stained for activation markers in both male and female mice as well as their matching controls. Flow cytometry was used to determine the ratio expressing CD25 (A–C) for CD4+, CD8+, and Nkp46+ splenocytes. Additionally, we measured cell proliferation in splenocytes after 48 h in response to 1000 ng of anti-CD3 (D&E).\n\nClick here to access the data.\n\nSupplemental Figure 4. Other zbtb32 KO maps. For comparison purposes, the mutant mRNA sequences that arise from each of the previous 3 mutant strains are depicted here. The mature mRNAs contain no introns, and non-coding portions are shown as thinner than the coding portions. Each of the previous three mutants was produced by deleting large portions of the Zbtb32 gene and replacing the sequence with selection and/or reporter genes.\n\nClick here to access the data.\n\n\nReferences\n\nMiller KM, Foster NC, Beck RW, et al.: Current state of type 1 diabetes treatment in the U.S.: updated data from the T1D Exchange clinic registry. Diabetes Care. 2015; 38(6): 971–978. PubMed Abstract \| Publisher Full Text\n\nAudiger C, Rahman MJ, Yun TJ, et al.: The importance of dendritic cells in maintaining immune tolerance. J Immunol. 2017; 198(6): 2223–2231. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHotta-Iwamura C, Tarbell KV: Type 1 diabetes genetic susceptibility and dendritic cell function: potential targets for treatment. J Leukoc Biol. 2016; 100(1): 65–80. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYamazaki S, Inaba K, Tarbell KV, et al.: Dendritic cells expand antigen-specific Foxp3+ CD25+ CD4+ regulatory T cells including suppressors of alloreactivity. Immunol Rev. 2006; 212: 314–329. PubMed Abstract \| Publisher Full Text\n\nPrice JD, Beauchamp NM, Rahir G, et al.: CD8+ dendritic cell-mediated tolerance of autoreactive CD4+ T cells is deficient in NOD mice and can be corrected by blocking CD40L. J Leukoc Biol. 2014; 95(2): 325–336. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPrice JD, Hotta-Iwamura C, Zhao Y, et al.: DCIR2+ cDC2 DCs and Zbtb32 Restore CD4+ T-Cell Tolerance and Inhibit Diabetes. Diabetes. 2015; 64(10): 3521–3531. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHoatlin ME, Zhi Y, Ball H, et al.: A novel BTB/POZ transcriptional repressor protein interacts with the Fanconi anemia group C protein and PLZF. Blood. 1999; 94(11): 3737–3747. PubMed Abstract\n\nTang CJ, Chuang CK, Hu HM, et al.: The zinc finger domain of Tzfp binds to the tbs motif located at the upstream flanking region of the Aie1 (aurora-C) kinase gene. J Biol Chem. 2001; 276(22): 19631–19639. PubMed Abstract \| Publisher Full Text\n\nMiaw SC, Choi A, Yu E, et al.: ROG, repressor of GATA, regulates the expression of cytokine genes. Immunity. 2000; 12(3): 323–333. PubMed Abstract \| Publisher Full Text\n\nPiazza F, Costoya JA, Merghoub T, et al.: Disruption of PLZP in mice leads to increased T-lymphocyte proliferation, cytokine production, and altered hematopoietic stem cell homeostasis. Mol Cell Biol. 2004; 24(23): 10456–10469. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKang BY, Miaw SC, Ho IC: ROG negatively regulates T-cell activation but is dispensable for Th-cell differentiation. Mol Cell Biol. 2005; 25(2): 554–562. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHirahara K, Yamashita M, Iwamura C, et al.: Repressor of GATA regulates TH2-driven allergic airway inflammation and airway hyperresponsiveness. J Allergy Clin Immunol. 2008; 122(3): 512–20.e11. PubMed Abstract \| Publisher Full Text\n\nHirasaki Y, Iwamura C, Yamashita M, et al.: Repressor of GATA negatively regulates murine contact hypersensitivity through the inhibition of type-2 allergic responses. Clin Immunol. 2011; 139(3): 267–276. PubMed Abstract \| Publisher Full Text\n\nBeaulieu AM, Zawislak CL, Nakayama T, et al.: The transcription factor Zbtb32 controls the proliferative burst of virus-specific natural killer cells responding to infection. Nat Immunol. 2014; 15(6): 546–553. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJash A, Wang Y, Weisel FJ, et al.: ZBTB32 Restricts the Duration of Memory B Cell Recall Responses. J Immunol. 2016; 197(4): 1159–1168. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMarkel P, Shu P, Ebeling C, et al.: Theoretical and empirical issues for marker-assisted breeding of congenic mouse strains. Nat Genet. 1997; 17(3): 280–284. PubMed Abstract \| Publisher Full Text\n\nOhta H, Ohinata Y, Ikawa M, et al.: Male germline and embryonic stem cell lines from NOD mice: efficient derivation of GS cells from a nonpermissive strain for ES cell derivation. Biol Reprod. 2009; 81(6): 1147–1153. PubMed Abstract \| Publisher Full Text\n\nRidgway WM: A new tool for dissecting genetic control of type 1 diabetes. Diabetes. 2014; 63(1): 56–58. PubMed Abstract \| Publisher Full Text\n\nLin X, Pelletier S, Gingras S, et al.: CRISPR-Cas9-Mediated Modification of the NOD Mouse Genome With Ptpn22R619W Mutation Increases Autoimmune Diabetes. Diabetes. 2016; 65(8): 2134–2138. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFellmann C, Gowen BG, Lin PC, et al.: Cornerstones of CRISPR-Cas in drug discovery and therapy. Nat Rev Drug Discov. 2017; 16(2): 89–100. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMali P, Aach J, Stranges PB, et al.: CAS9 transcriptional activators for target specificity screening and paired nickases for cooperative genome engineering. Nat Biotechnol. 2013; 31(9): 833–838. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHeng TS, Painter MW, Immunological Genome Project Consortium: The Immunological Genome Project: networks of gene expression in immune cells. Nat Immunol. 2008; 9(10): 1091–1094. PubMed Abstract \| Publisher Full Text\n\nMabbott NA, Gray D: Identification of co-expressed gene signatures in mouse B1, marginal zone and B2 B-cell populations. Immunology. 2014; 141(1): 79–95. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChevrier S, Emslie D, Shi W, et al.: The BTB-ZF transcription factor Zbtb20 is driven by Irf4 to promote plasma cell differentiation and longevity. J Exp Med. 2014; 211(5): 827–840. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHardy RR: B-1 B cell development. J Immunol. 2006; 177(5): 2749–2754. PubMed Abstract \| Publisher Full Text\n\nYoon HS, Scharer CD, Majumder P, et al.: ZBTB32 is an early repressor of the CIITA and MHC class II gene expression during B cell differentiation to plasma cells. J Immunol. 2012; 189(5): 2393–2403. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMiaw SC, Kang BY, White IA, et al.: A repressor of GATA-mediated negative feedback mechanism of T cell activation. J Immunol. 2004; 172(1): 170–177. PubMed Abstract \| Publisher Full Text\n\nRatiu JJ, Racine JJ, Hasham MG, et al.: Genetic and Small Molecule Disruption of the AID/RAD51 Axis Similarly Protects Nonobese Diabetic Mice from Type 1 Diabetes through Expansion of Regulatory B Lymphocytes. J Immunol. 2017; 198(11): 4255–4267. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nColey WD, Bogdanik L, Vila MC, et al.: Effect of genetic background on the dystrophic phenotype in mdx mice. Hum Mol Genet. 2016; 25(1): 130–145. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGaublomme JT, Yosef N, Lee Y, et al.: Single-Cell Genomics Unveils Critical Regulators of Th17 Cell Pathogenicity. Cell. 2015; 163(6): 1400–1412. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHöglund P, Mintern J, Waltzinger C, et al.: Initiation of autoimmune diabetes by developmentally regulated presentation of islet cell antigens in the pancreatic lymph nodes. J Exp Med. 1999; 189(2): 331–339. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMiyazaki A, Hanafusa T, Yamada K, et al.: Predominance of T lymphocytes in pancreatic islets and spleen of pre-diabetic non-obese diabetic (NOD) mice: a longitudinal study. Clin Exp Immunol. 1985; 60(3): 622–630. PubMed Abstract \| Free Full Text\n\nLin W, Lai CH, Tang CJ, et al.: Identification and gene structure of a novel human PLZF-related transcription factor gene, TZFP. Biochem Biophys Res Commun. 1999; 264(3): 789–795. PubMed Abstract \| Publisher Full Text\n\nJiang X, Zhang H, Yin S, et al.: Specific deficiency of Plzf paralog, Zbtb20, in Sertoli cells does not affect spermatogenesis and fertility in mice. Sci Rep. 2014; 4(1): 7062. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTarbell K, Coley WD, Zhao Y, et al.: Dataset 1 in: Loss of Zbtb32 in NOD mice does not significantly alter T cell responses. F1000Research. 2018. Data Source\n\nTarbell K, Coley WD, Zhao Y, et al.: Dataset 2 in: Loss of Zbtb32 in NOD mice does not significantly alter T cell responses. F1000Research. 2018. Data Source\n\nTarbell K, Coley WD, Zhao Y, et al.: Dataset 3 in: Loss of Zbtb32 in NOD mice does not significantly alter T cell responses. F1000Research. 2018. Data Source\n\nTarbell K, Coley WD, Zhao Y, et al.: Dataset 4 in: Loss of Zbtb32 in NOD mice does not significantly alter T cell responses. F1000Research. 2018. Data Source\n\nTarbell K, Coley WD, Zhao Y, et al.: Dataset 5 in: Loss of Zbtb32 in NOD mice does not significantly alter T cell responses. F1000Research. 2018. Data Source\n\nTarbell K, Coley WD, Zhao Y, et al.: Dataset 6 in: Loss of Zbtb32 in NOD mice does not significantly alter T cell responses. F1000Research. 2018. Data Source" }	[ { "id": "31946", "date": "09 Apr 2018", "name": "Hubert M. Tse", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nDefining how tolerance is lost in Type 1 diabetes (T1D) will significantly contribute to our understanding of autoimmune dysregulation and the generation of beta-cell-specific effector T cell responses. Previous studies by the Tarbell lab demonstrated that DCIR2+ tolerogenic dendritic cells could promote self-tolerance to autoreactive CD4 T cells partly due to the upregulation of the transcription factor Zbtb32 in CD4 T cells. Overexpression of Zbtb32 in CD4 T cells was effective in delaying T1D by decreasing T cell proliferation and IFN-gamma synthesis. To further confirm the role of Zbtb32 in autoimmune diabetes, the current manuscript by Coley, et al. attempts to test the hypothesis that loss of Zbtb32 will exacerbate diabetogenic T cell responses and accelerate T1D in newly derived NOD.Zbtb32-/- mice. Contrary to their hypothesis, the data presented displays a moderate acceleration in spontaneous diabetes in male NOD.Zbtb32-/- mice, but no difference in female mice. In addition, there were no significant differences observed in activation markers, IFN-gamma synthesis, or proliferation between CD4, CD8, NK, and B cells in NOD and NOD.Zbtb32-/- mice. The manuscript is well-written, results are interesting, and the conclusions are supported by the provided data. While the majority of the data is “negative”, the results are insightful and informative to the T1D research community. However, there are minor concerns that need to be addressed with the manuscript in its current form:\n\nZbtb32 is a negative regulator of GATA3, did the authors ever examine GATA3 expression levels and GATA3-dependent signaling pathways to determine if Th2 cytokine responses were elevated in NOD.Zbtb32-/- mice? The spontaneous incidence of T1D in female mice was stopped at 35 weeks of age, but the male incidence stopped at 45 weeks of age, if the female incidence was extended to the same time point, was there a difference? I agree with the authors’ discussion concerning strain differences and potential compensatory mutations that may account for the absence of accelerated T1D in NOD.Zbtb32-/- mice in contrast to other inbred mouse models. Out of curiosity, was EAE ever induced in NOD.Zbtb32-/- mice to determine if this mouse may exhibit an increased susceptibility to ascending paralysis? The figure legend for 4 needs to be consistent since the symbols are different for NOD and NOD.Zbtb32-/- mice.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "3593", "date": "16 Apr 2018", "name": "Kristin Tarbell", "role": "Author Response", "response": "Thank you for taking the time to review our article and for your thoughtful responses. We truly appreciate your review and will endeavor to provide satisfactory responses to each the concerns that you raised.Query 1) Zbtb32 is a negative regulator of GATA3, did the authors ever examine GATA3 expression levels and GATA3-dependent signaling pathways to determine if Th2 cytokine responses were elevated in NOD.Zbtb32-/- mice? Answer: It is our understanding that Zbtb32 normally inhibits the function of GATA3 by physically occupying the DNA binding site for GATA3, and not by regulating its gene expression. Consequently, we did not examine GATA3 gene expression. We did examine cytokine production in NOD.Zbtb32-/- lymphocytes (see Fig. 4), but did not observe any differences compared to control mice.Query 2) The spontaneous incidence of T1D in female mice was stopped at 35 weeks of age, but the male incidence stopped at 45 weeks of age, if the female incidence was extended to the same time point, was there a difference? Answer: We monitored most of the female mice (50 out of 55) past 35 weeks but did not encounter any cases of disease onset in females beyond 31 weeks. The figure for disease incidence in females was truncated at 35 weeks for the sake of clarity.Query 3) I agree with the authors’ discussion concerning strain differences and potential compensatory mutations that may account for the absence of accelerated T1D in NOD.Zbtb32-/- mice in contrast to other inbred mouse models. Out of curiosity, was EAE ever induced in NOD.Zbtb32-/- mice to determine if this mouse may exhibit an increased susceptibility to ascending paralysis? Answer: No, we never attempted to induce EAE in the NOD.Zbtb32-/- mice.Query 4) The figure legend for 4 needs to be consistent since the symbols are different for NOD and NOD.Zbtb32-/- mice. Answer: Thank you for the feedback. In Figure 4, data is shown for 3 cell types from both wild type and knockout animals. Each cell type was assigned its own symbol, but the coloration is consistent with the genotype. All wildtype data uses black, filled symbols; all knockout data uses open symbols. We will submit a revision to make this more clear to future readers.Thank you again for your feedback and constructive criticism." }, { "c_id": "3597", "date": "17 Apr 2018", "name": "Hubert Tse", "role": "Reviewer Response", "response": "Thank you for your responses! I have no other concerns." } ] }, { "id": "32633", "date": "19 Apr 2018", "name": "Ivan C. Gerling", "expertise": [ "Reviewer Expertise Autoimmunity of Type 1 Diabetes" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an interesting \"negative results\" report. The authors had previously reported data in support of the hypothesis that systemic loss of the expression of the gene Zbtb32 would increase T-cell activation and increase diabetic pathogenesis. The expected hypersensitivity of leukocytes was not observed. A minor but non-significant increase in diabetes incidence was observed in males, but only at later than 30 weeks of age. The lack of hyper-responsiveness of immune cells in these NOD mice (investigating an impressive panel of different leukocyte types and assays) is surprising since hyper-responsiveness had been found in 3 other strains with a knock out of Zbtb32. This raise an important question of whether immune responses and their regulation in NOD mice are different compared to other commonly used mice strains with respect to regulation by Zbtb32. However, as noted and discussed by the authors, the mutation on the NOD background was created using CRISPR/CAS9 whereas the 3 strains where the KO was successful in enhancing immune reactivity were more traditional KO mice. Furthermore the immune reactivity assays in this report are only comparing activation of cells from NOD mice without direct comparison to cells from the other strain backgrounds. As such this question may warrant a more detailed investigation before it can be answered conclusively. The manuscript is clearly written and experimental designs are technically sound.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-318
https://f1000research.com/articles/7-1741/v1	02 Nov 18	{ "type": "Research Article", "title": "Molecular and in-silico analysis of single nucleotide polymorphism targeting human TP53 gene exon 5-8 in Sudanese esophageal cancer patients", "authors": [ "Rihab M. Elfaki", "Mohammed S. Abdelaziz", "Hisham N. Altayb", "Munsoor M. Munsoor", "Ahmed A. Gameel", "Mohammed S. Abdelaziz", "Hisham N. Altayb", "Munsoor M. Munsoor", "Ahmed A. Gameel" ], "abstract": "Background: The protein product of the normal TP53 gene performs an essential function in cell cycle control and tumor suppression, and the mutation of a TP53 gene is an essential step in the development of many cancers. Despite the reported association of TP53 gene mutations with many human cancers, the comprehensive computational analysis of single nucleotide polymorphisms (SNPs), and their functional impacts, still remains rare. Methods: In this study DNA were extracted from formalin fixed paraffin embedded samples followed by the conventional polymerase chain reaction and DNA sequencing. Computational analysis was performed using different algorithms to screen for deleterious SNPs. Results: The results demonstrate that there are synonymous SNPs (sSNPs) and non-synonymous SNPs (nsSNPs) in the TP53 gene that may be deleterious to p53 structure and function. Additionally, TP53 gene mutations were found in 40% of samples. Six out of ten of TP53 gene mutations occurred in exon 5, two mutation in exon 6 and other two were present in exon 8. Only one SNP in position E298Q was predicted to have a neutral effect and other SNPs were predicted to be disease related according to Mutation Taster software. A total of 37.2% of squamous cell carcinoma (SCC) samples were found to be mutated, 87.5% of them exist in exon 5, 12.5% in exon 6 and 6.3% in exon 8, whereas adenocarcinoma (AC) achieved a higher rate of mutation (57.1%) with 100% exon 5 involvement. Conclusions: Mutation of TP53 exon 5 in esophageal cancer patients were the most frequent. Genomic results have identified a higher TP53 mutation rate in esophageal AC in contrast to SCC.", "keywords": [ "Protein 53", "SNPs", "in silico Analysis", "Esophageal cancer", "Sudan." ], "content": "Introduction\n\nEsophageal cancer is considered one of the eight most common cancers throughout the world, and is also one of the most fatal cancers, taking into account its aggressiveness and reduced survival rate. Because of its poor prognosis with 5-year survival rates ranging between 10–13%, it ranks sixth among all cancers in mortality rate1–4.\n\nKnockout of TP53 in mice leads to the development of different tumors, including lymphomas, sarcomas adenocarcinoma and benign tumors such as hemangioma, before they reach 6 month of age5.\n\nTP53 gene encodes a tumor suppressor protein which plays an important role inside the cell especially in DNA transcription and repair, senescence, apoptosis, tumor suppression, treatment response and also the response to changes in metabolism6,7. Protein domains represent independently folding units of protein with a size between 40 to 200 amino acids. Human p53 protein contains three domains; transcriptional activation, DNA binding, and oligomerization domains. These domains are edged by a connecting region. A proline-rich region links the transcriptional activation and DNA binding domains, a second proline-rich region links the DNA binding and oligomerization domains and a basic region form the C-terminus of the protein8. The evolutionarily highly conserved core domain (amino acids ~100 to ~300) is involved in sequence-specific binding to promoters of p53-regulated genes9.\n\nSingle nucleotide polymorphisms (SNP) are a significant type of genetic variation commonly detected in the human genome. SNPs occur in non-coding regions as well as in coding regions of the genome10,11. A total of 336,845,724 SNPs have been identified in humans so far, and have been deposited in NCBI dbSNP. The human TP53 gene has 3115 identified SNPs. SNPs arise in coding regions may cause an amino acid change in the corresponding protein and in such case it is called as non-synonymous SNP (nsSNP) or may not change the amino acid and here it is called a synonymous SNP (sSNP); these nsSNPs change the protein structure and hence its function, causing a specific disease12,13.\n\nRecently a number of articles have demonstrated the association of SNPs in the TP53 gene with different cancer types, but in silico analysis has not yet been discussed on the functional, interactional and structural aspects of different types of SNPs in this gene. In the current study, we used different bioinformatics prediction tools and databases for analysis of these SNPs in TP53 gene. As a significant number of mutations have an impact on protein stability and interactions with the corresponding proteins, we also offered a structural model of the mutant protein. Here in this study the main objective is to detect mutations of TP53 gene focusing on exons 5 to 8 among esophageal cancer patients as these has been reported as the most mutated exons in this gene14.\n\n\nMethods\n\nSections of 30–40 µm thickness from 50 formalin fixed paraffin embedded (FFPE) tissue samples were obtained from esophageal cancer patients representing different hospitals and clinics in Khartoum State, Sudan, from July 2013 to June 2017. All patients have been previously diagnosed with squamous cell carcinoma (SCC) and adenocarcinoma (AC).\n\nGenomic DNA for PCR analysis was extracted from FFPE tissue blocks. Using commercial DNA extraction kits for fast isolation of genomic DNA from FFPE samples as per manufacturer’s instructions. Extraction procedure is based on combination of an efficient lysis step with a subsequent binding of genomic DNA on a Spin Filter surface followed by washing of the bound DNA and finally eluting of the DNA (845-BP-0020250, black PREP FFPE DNA Kit, Analytik Jena Company).\n\nFor amplification of exon 5, 6,7 and 8 of TP53 gene, four pairs of primers (catalogue numbers: 171002-009_D5, 171002-009_D6, 171002-009_D7, 171002-009_D8, 171002-009_D9, 171002-009_D10, 171002-009_D11, 171002-009_D12; Macrogen, Korea) were used15,16 (Table 1). A total of 2–5 µl of genomic DNA, 0.5 µl forward and 0.5 µl reverse primer and 25 μl double distilled water (DDW) was combined to make up the final reaction volume. The mixture was amplified using Heal force thermal cycler (Model No: K960) with the following amplification conditions; 95°C for 5 min, followed by 37 cycles at 95°C for 45 sec, primer-specific annealing temperature for 45 sec, 72°C for 45 sec and a final extension at 72°C for 5 min. 5 μl of the PCR products were applied on 2 % agarose gel and remaining PCR products were sequenced by BGI company (China).\n\nThe SNP information SNP ID, mRNA accession number NM_000546, and Protein accession number NP_000537 of the human TP53 gene used in our computational analysis were retrieved from the National Center for Biotechnology Information (NCBI) database and catalogue of somatic mutation in cancer (COSMIC) database (TP53_ENST00000269305). The nucleotide and amino acid sequence of the p53 protein were obtained and investigated using nucleotide (NG_017013), Gene (Gene ID: 7157) database NCBI and UniProt database (P04637).\n\nCodon code aligner. Sequences were assembled into contigs end clipped and edited using Codon Code Aligner software version 8.0.1 (Dedham, MA, USA). Sequence data are available at GenBank under accession numbers MH366303 to MH36648317.\n\nSIFT Program. SIFT (Sorting Intolerant from Tolerant) tool uses sequence homology to calculate the probability of affecting protein function in case of amino acid change. It uses the concept of evolutionarily conserved regions which is less tolerant to mutations, and therefore amino acid change or frame shift mutations in these regions are expected to affect protein function the most. SIFT tool works by introducing a query protein into SIFT program to be searched against protein database aligned with homologous protein sequences. Then the program calculates SIFT score based on amino acid changes in that position. A SIFT score ranges from 0 to 1. Score less than 0.05 is predicted to affect protein function and considered functionally deleterious, whereas any score more than or equal to 0.05 represents a neutral substitution18–20.\n\nPolyPhen -2. PolyPhen-2 (Polymorphism Phenotyping version 2) is a structural and functional predicting tool that predicts the effect of an amino acid change on protein characteristics based on SNPs functional annotations, protein structural properties with sequence annotation, and finally predict if the coding non-synonymous SNPs are considered damaging or not21,22.\n\nPolyPhen-2 workflow requires protein sequence, mutational position, and substitution. The PolyPhen output is represented with a score that ranges from 0 to 1, with zero score indicating a neutral effect of amino acid substitution on protein function and a higher score representing a mutation that is more likely to be damaging23.\n\nI-Mutant 3.0. I-Mutant 3.0 is a support vector machine (SVM) based tool, which was used to calculate the stability changes of specific SNP upon protein sequence. Information of wild and mutated residue, protein sequence, temperature, and pH was used as input parameters to this server, and finally, the outputs reports if a point mutation is stable or not. The program categorizes the prediction into: neutral mutation (DDG = 0.5 kcal/mol), large decrease of stability (0.5 kcal/mol). The output is a free Gibbs energy change value (ΔΔG) of protein before and after mutation24–27.\n\nPhD-SNP. PhD-SNP (Predictor of Human Deleterious Single Nucleotide Polymorphisms) software is a prediction tool that predicts disease association of nsSNP by dividing those SNPs into disease-related or neutral polymorphism based on a score ranged from (0-1); SNPs with a score above 0.5 are considered disease associated according to the program algorithm. PhD-SNP outputs depend on a number of sequences aligned, conservation index of SNP position, frequencies of wild and mutant residues19,20,28.\n\nProject HOPE. Structural and biochemical analysis for mutations was accomplished using Project HOPE is a web-server used to give a comprehensive report on the effect of the specific mutation on the 3D structure of the native protein and the variant model using different software and sources. The user can submit a protein sequence or an accession number of specific protein after specifying the wild-type residue and the new mutant form to create the report29,30.\n\nMutation Taster. Mutation Taster calculates the pathogenic consequences of variations in DNA sequence. It predicts the functional impact of amino acid alterations, intronic and synonymous substitutions, in addition to INDEL mutations and variants covering intron-exon connection region. Mutation Taster prediction system divides alterations as; Disease-causing: which is probably deleterious, Disease-causing automatic: the alteration here is known to be deleterious, Polymorphism: probably harmless alteration and polymorphism automatic: known to be harmless31–33.\n\nFATHMM. FATHMM (Functional Analysis Through Hidden Markov Models) is a web-server predicts the functional significances of both coding and non-coding variants. We selected the cancer option to display predictions that can distinguish between cancer-promoting/driver mutations and other germline polymorphisms. It uses a default prediction threshold of -0.75. Predictions with scores less than this indicate that the mutation is potentially cancer associated34.\n\n\nResults\n\nEsophageal squamous cell carcinoma cases represent 43 (86%) of all cases, whereas adenocarcinoma made up 7 cases (14%). Mutation analysis results demonstrate a higher TP53 mutation rate in esophageal adenocarcinoma compared to squamous cell carcinoma. This were illustrated in (Table 2) in the results section which describe Histopathological diagnosis, mutational status and exons affected in esophageal cancer patients.\n\nDistribution of TP53 coding synonymous SNPs (sSNPs), coding non-synonymous SNPs (nsSNPs) and INDEL through exon 5-8 in P53 gene were represented by (Figure 1).\n\nTP53 gene SNPs were found in 20 out of 50 sample of esophageal carcinomas (40%). Six out of ten SNPs (60%) were missense SNPs leading to amino acid substitution, four SNPs (40%) were silent mutations without any amino acid change. Six out of ten (60%) TP53 SNPs occurred in exon 5 at the following codons 160, 161 (twice), 163, 164, 175. Three of them were missense and three were silent. Two SNPs (20%) were located in exon 6 at codon 215 (missense) and 222 (silent). Two SNPs (20%) were present in exon 8 at codon 298 (missense) and 305 (silent).\n\n15 out of 43 (34.9%) SSC samples were found to be mutated, 13/15 (86.7%) of them existed in exon 5, 2/15 (13.3 %) in exon 6 and 1/15 (6.7%) in exon 8. Whereas adenocarcinoma had a higher rate of mutations 4/7 (57.1%) with 100% of SNPs occurring in exon 5.\n\nThe percentage of deleterious nsSNPs predicted by SIFT and PolyPhen was 66.7% (Figure 2), those SNPs were A161D, K164E, R175P and S215N according to SIFT and PolyPhen (Table 3–Table 4). I-Mutant suite also give the same percentage for deleterious nsSNPs which is 66.7% (Table 5) but in case of using I-Mutant suite the predicted SNPs to be deleterious were, A161D, R175P, S215N and M160V (Table 6). The PhD-SNP report defines 5/6 (83.3%) of nsSNPs as disease related polymorphism (Table 7).\n\nThe magenta cylindrical bar indicates the percentage of nsSNPs that were found to be deleterious by SIFT, damaging (Possibly/Probably) by PolyPhen, and largely unstable by I-Mutant Suite. The pink cylinder indicates the percentage of nsSNPs that were found to be tolerated by SIFT, benign by PolyPhen, and largely stable/neutral by I-Mutant Suite.\n\nSIFT: Sorting Intolerant From Tolerant. SIFT score ≤ 0.05 considered damaging. SIFT Tolerance Index: Ranges from 0 to 1. AA: Amino acid\n\nnsSNPs: Non-synonymous single nucleotide polymorphisms, PolyPhen: Polymorphism Phenotyping v2.\n\nSIFT: Sorting Intolerant From Tolerant. The amino acid substitution is predicted deleterious if the score is <= 0.05, and tolerated if the score is > 0.05.\n\nI-mutant RI (Reliability Index): 0–10, where 0 is the lowest reliability and 10 is the highest reliability.\n\nThe results reveal that SNPs in positions E298Q were predicted to be a neutral polymorphism which represent 10% of all mutation detected. All other SNPs M160V, A161D, A161A, Y163Y, K164E, R175P, S215N, P222P, and K305K representing 90% of all SNPs were predicted to be disease related according to MutationTaster software (Table 8).\n\n*according to MutationTaster.\n\nFrequency of SNPs among different samples was shown in details in (Table 9). FATHMM server; cancer association predictions result of the non-synonymous changes found in TP53 gene exon 5-8 were demonstrated in (Table 10).\n\nDifferent mutations with their position, wild type and mutant form in addition to alignment and chromatogram were illustrated in (Figure 3–Figure 8).\n\n\nDiscussion\n\nMutations of TP53 gene can lead to loss of functional characteristics in tumor cells, as mutant TP53 may not play the assigned role in repairing cellular machinery leading to a loss of normal function and, subsequently, cells with mutant gene may express uncontrolled replication which leads to accumulation of protein 5313.\n\nThere were 8 (40%) males with TP53 gene mutations and 12 (60%) females. TP53 mutations were found in 22.9% of esophageal SCCs and 14.3% of esophageal ACs in a study done by Zheng et al.35 here the ratio was (34.9% vs. 57.1%) respectively. Sample size may be the main factor affecting the result.\n\nShi et al.36 in Henan province, China showed that the p53 mutations were detected in 30 out of 43 (70%) SCC cases but in this study it represents 16 out of 43 (37.2%); this difference can be attributed to different factors e.g. geographical zone differences or type of exons studied as China has a high-incidence of esophageal carcinoma. Another study in Henan province, China conducted by Li-Ya15 and his colleagues stated that p53 mutations were detected in 40.9% of their esophageal cancer specimens and this is compatible with our finding on mutations being detected in 20 out of 50 specimens of esophageal carcinomas which represent 40%.\n\nIn a study conducted by Zheng et al.35 Qiqihar City, China, a higher mutation rate was found in SCC samples compared to AC samples (31.4% vs. 21.4%, respectively). And this partially differs from our result which found 4 out of 7 (57.1%) of adenocarcinoma patients had mutations which revealed higher mutation rate among AC samples, and this result may differ if had a larger sample size with more adenocarcinoma cases.\n\nExon 5 of the TP53 gene was observed to be the most mutated exon among the other exons investigated here in this study, with 17/20 (85%) of all mutations detected found in exon 5, 2/20 (10%) in exon 6 and 1/20 (5%) in exon 8, while exon 7 showed no mutations. Uchino et al.37 results for mutation distribution across exons was 39.3% in exon 5, 32.1% in exon 6, 17.9% in exon 7 and (10.7%) forexon 8 mutations. Which has partly agreed with our result in exon 5 having the higher mutation rate in addition to reasonable differences in other exons rate of mutation, and this can be attributed to the small sample size used in this study.\n\nA total of 10% of all detected mutations were classified as neutral polymorphism, while 90% were considered disease-causing according to Mutation Taster, considered high rate of disease-causing mutations. This rate is lesser using other software due to different algorithms used by any one of them, their linked databases, and characteristics of the different software.\n\n\nConclusions\n\nMutation of exon 5 in p53 gene were the most frequent in esophageal cancer. Genomic results have identified a high TP53 mutation rate in esophageal Adenocarcinoma compared to squamous cell carcinoma.\n\n\nEthical considerations\n\nThis study was approved by the Institutional Ethics Committee, Sudan University of Science and Technology (reference number for the ethical committee is DSR-IEC-13-05). Patients consent cannot be obtained because most of the patients were dead and the rest cannot be traced due to lack of contact data. Therefore all samples and medical data used in this study have been irreversibly anonymized to ensure patients privacy.\n\n\nData availability\n\nTP53 sequences for this study has been submitted to Banklt NCBI and had been assigned the accession numbers MH366303 to MH366483. The sequences are available as 4 PopSet entries for exons 5–8:\n\nExon 5: 1472901613\n\nExon 6: 1472901713\n\nExon 7: 1472901809\n\nExon 8: 1472901897\n\nTP53 sequence results were submitted in a zipped file. These sequencing results as received from BGI Company (China) include 50 esophageal cancer patients in this study using the four sets of primers for exon 5, 6, 7 and 8. Sequencing files that needs to be viewed using FinchTV and Notepad file.\n\nDataset 1: Patients information, clinical data, and histological findings of all patients were available in Excel format. 10.5256/f1000research.15534.d21973938", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgments\n\nWe would like to thank Sudan University Research Laboratory staff for their orientation and support, also laboratory staff in all hospitals and clinics in Khartoum, Bahry and Omdurman for their help in sample collection.\n\n\nReferences\n\nZhang Y: Epidemiology of esophageal cancer. World J Gastroenterol. 2013; 19(34): 5598–606. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nUmar SB, Fleischer DE: Esophageal cancer: epidemiology, pathogenesis and prevention. Nat Clin Pract Gastroenterol Hepatol. 2008; 5(9): 517–26. PubMed Abstract \| Publisher Full Text\n\nKeramati MR, Esmaeilzadeh M, Bashashati M: Brain Metastasis from Esophageal Cancer. In: Brain Metastases from Primary Tumors, Elsevier; 2016; 3: 145–54. Publisher Full Text\n\nScarpa M, Valente S, Alfieri R, et al.: Systematic review of health-related quality of life after esophagectomy for esophageal cancer. World J Gastroenterol. 2011; 17(42): 4660–74. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDonehower LA, Harvey M, Slagle BL, et al.: Mice deficient for p53 are developmentally normal but susceptible to spontaneous tumours. Nature. 1992; 356(6366): 215–21. PubMed Abstract \| Publisher Full Text\n\nGeorge P: p53 how crucial is its role in cancer. Int J Curr Pharm Res. 2011; 3(2): 19–25. Reference Source\n\nTP53 tumor protein p53. 2016; [cited 2018 Apr 8]. Reference Source\n\nStavridi ES, Huyen Y, Sheston EA, et al.: The Three-Dimensional Structure of p53. In: The p53 Tumor Suppressor Pathway and Cancer. Springer; 2005; 2: 25–52. Publisher Full Text\n\nel-Deiry WS, Kern SE, Pietenpol JA, et al.: Definition of a consensus binding site for p53. Nat Genet. 1992; 1(1): 45–9. PubMed Abstract \| Publisher Full Text\n\nNadeau JH: Single nucleotide polymorphisms: tackling complexity. Nature. 2002; 420(6915): 517–8. PubMed Abstract \| Publisher Full Text\n\nButler JM: Advanced topics in forensic DNA typing: methodology. Academic Press; 2011; 347–369. Reference Source\n\nRamensky V, Bork P, Sunyaev S: Human non-synonymous SNPs: server and survey. Nucleic Acids Res. 2002; 30(17): 3894–3900. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIARC IA for R on C: TP53 Mutation. Accessed August 2001. Reference Source\n\nMontesano R, Holestein M, Hainaui P: Genetic alterations in esophageal cancer and their relevance to etiology and pathogenesis: a review. Int J cancer. 1996; 69(3): 225–35. PubMed Abstract \| Publisher Full Text\n\nLi LY, Tang JT, Jia LQ, et al.: Mutations of p53 gene exons 4-8 in human esophageal cancer. World J Gastroenterol. 2005; 11(19): 2998–3001. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMakwane N, Saxena A: Study of mutations in p53 tumour suppressor gene in human sporadic breast cancers. Indian J Clin Biochem. 2009; 24(3): 223–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCodonCode aligner. Dedham, USA: CodonCode Corporation. Reference Source\n\nSim NL, Kumar P, Hu J, et al.: SIFT web server: predicting effects of amino acid substitutions on proteins. Nucleic Acids Res. 2012; 40(Web Server issue): W452–W457. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCapriotti E, Altman RB, Bromberg Y: Collective judgment predicts disease-associated single nucleotide variants. BMC Genomics. 2013; 14 Suppl 3: S2. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHepp D, Gonçalves GL, de Freitas TR: Prediction of the damage-associated non-synonymous single nucleotide polymorphisms in the human MC1R gene. PLoS One. 2015; 10(3): e0121812. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAdzhubei IA, Schmidt S, Peshkin L, et al.: A method and server for predicting damaging missense mutations. Nat Methods. 2010; 7(4): 248–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAdzhubei I, Jordan DM, Sunyaev SR: Predicting functional effect of human missense mutations using PolyPhen-2. Curr Protoc Hum Genet. 2013; Chapter 7: Unit7.20. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJia M, Yang B, Li Z, et al.: Computational analysis of functional single nucleotide polymorphisms associated with the CYP11B2 gene. PLoS One. 2014; 9(8): e104311. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDoss CG, Rajith B, Garwasis N, et al.: Screening of mutations affecting protein stability and dynamics of FGFR1-A simulation analysis. Appl Transl Genomics. 2012; 1: 37–43. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCapriotti E, Fariselli P, Casadio R: I-Mutant2.0: predicting stability changes upon mutation from the protein sequence or structure. Nucleic Acids Res. 2005; 33(Web Server issue): W306–W310. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRaghav D, Sharma V: An In Silico Evaluation of Deleterious Nonsynonymous Single Nucleotide Polymorphisms in the ErbB3 Oncogene. Biores Open Access. 2013; 2(3): 206–11. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMahmud Z, Malik SU, Ahmed J, et al.: Computational Analysis of Damaging Single-Nucleotide Polymorphisms and Their Structural and Functional Impact on the Insulin Receptor. Biomed Res Int. 2016; 2016: 2023803. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOu L, Przybilla MJ, Whitley CB: Phenotype prediction for mucopolysaccharidosis type I by in silico analysis. Orphanet J Rare Dis. 2017; 12(1): 125. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVenselaar H, Te Beek TA, Kuipers RK, et al.: Protein structure analysis of mutations causing inheritable diseases. An e-Science approach with life scientist friendly interfaces. BMC Bioinformatics. 2010; 11: 548. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHussain MR, Shaik NA, Al-Aama JY, et al.: In silico analysis of Single Nucleotide Polymorphisms (SNPs) in human BRAF gene. Gene. 2012; 508(2): 188–96. PubMed Abstract \| Publisher Full Text\n\nSchwarz JM, Cooper DN, Schuelke M, et al.: Mutationtaster2: mutation prediction for the deep-sequencing age. Nat Methods. 2014; 11(4): 361–2. PubMed Abstract \| Publisher Full Text\n\nMutation T@ster Documentation. [cited 2018 Apr 7]. Reference Source\n\nSchwarz JM, Rödelsperger C, Schuelke M, et al.: MutationTaster evaluates disease-causing potential of sequence alterations. Nat Methods. 2010; 7(8): 575–6. PubMed Abstract \| Publisher Full Text\n\nShihab HA, Gough J, Cooper DN, et al.: Predicting the functional consequences of cancer-associated amino acid substitutions. Bioinformatics. 2013; 29(12): 1504–10. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nZheng H, Wang Y, Tang C, et al.: TP53, PIK3CA, FBXW7 and KRAS Mutations in Esophageal Cancer Identified by Targeted Sequencing. Cancer Genomics Proteomics. 2016; 13(3): 231–8. PubMed Abstract\n\nShi ST, Yang GY, Wang LD, et al.: Role of p53 gene mutations in human esophageal carcinogenesis: results from immunohistochemical and mutation analyses of carcinomas and nearby non-cancerous lesions. Carcinogenesis. 1999; 20(4): 591–7. PubMed Abstract \| Publisher Full Text\n\nUchino S, Saito T, Inomata M, et al.: Prognostic significance of the p53 mutation in esophageal cancer. Jpn J Clin Oncol. 1996; 26(5): 287–92. PubMed Abstract \| Publisher Full Text\n\nElfaki RM, Abdelaziz MS, Altayb HN, et al.: Dataset 1 in: Molecular and In-Silico Analysis of Single Nucleotide Polymorphism Targeting Human TP53 Gene exon 5-8 in Sudanese Esophageal Cancer Patients. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15534.d219739" }	[ { "id": "40228", "date": "14 Nov 2018", "name": "Carlos Mario Muñetón Peña", "expertise": [ "Reviewer Expertise Human Molecular genetic", "Cancer genetic", "human cytogenetic." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn general it is a good paper, the results are shown in an orderly sequence, but some data of the results are confusing; the discussion is very concrete.\nThe authors performed an interesting computational analysis with a battery of bioinformatics tools for the screen of the SNPs; these analyses are the strongest of the study.\nOther comments specific to the study:\nAccording to the objective of the study, it would have been very interesting if the authors also studied the polymorphism ‘rs104252272 (Pro72Arg)’, identified in different populations and associated with risk to different types of tumors.\n\nRegarding the classification of the variants I would recommend the authors use the ACMG nomenclature throughout the manuscript to homogenize the information, instead of using \"deleterious SNPs, missense SNP”; by pathogenic variant, benign variant, VUS, etc.\n\nThe abstract does not mention the objective of the article or the size of the sample: In the results it appears \"Six out of ten of TP53 gene mutations occurred in exon 5\", but the authors describe that the mutated samples were 20 (40%) and not 10 mutations; exon 5 has more than six mutations. The authors should clarify this confusing data.\n\nThe authors should also describe in Table 2 the mutations identified using the international nomenclature to show exactly the position in which the substitution occurred and the type of mutation (example: c.595 G> T, GGA / TGA.)\n\nTable 2 shows 18 mutations in exon 5, two in exon 6 and one in exon 8, but these data do not agree with the text, and the percentages are made from 10 mutations - these data must be reviewed to agree.\n\nThe description of the mutated SSC cases is not agreed in the manuscript - it is mentioned that there are 15 cases, but 16 cases are described (13 in exon 5, 2 in exon 6 and 1 in exon 8). In this situation, the authors must also present the most clear information.\n\nFinally, in the discussion the authors compare their results of mutations in the TP53 gene only with studies from China - it would be very interesting if they would also compare them with similar studies carried out in other populations (European, American, Hispanic American, etc.).\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "50574", "date": "01 Jul 2019", "name": "Naoise Synnott", "expertise": [ "Reviewer Expertise p53", "breast cancer", "targeted therapies." ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this article the authors carried out TP53 sequencing in 50 esophageal cancers, from the Sudan population. Sequencing identified 40% of samples had a TP53 mutation, with the majority of mutations occurring in the core domain in exon 5. This finding is consistent with the literature. A higher rate of TP53 mutations was found in Ac compared to SCC, however the sample numbers of each cancer type very greatly in this study.\n\nOverall this is a well written article, with novel data on the Sudanese population, with some consistency with previous reports in other geographical locations.\n\nI would suggest including data on the incidence, prevalence and mortality rate of SCC and AC in Sudan in the introduction (GLOBOCAN, 2018).\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "50571", "date": "16 Jul 2019", "name": "Hideo Baba", "expertise": [ "Reviewer Expertise esophageal cancer" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nMajor revision The authors evaluated single nucleotide polymorphisms (SNPs) in TP53 gene exon 5-8, and found that 37.2% of esophageal squamous cell carcinoma (SCC) samples harbored mutation, whereas 57.1% of esophageal adenocarcinoma (AC) samples harbored mutation. These result is interesting. However, this study lacks the information or explanation of SNPs in TP53 gene exon 5-8.\nThe authors need to discuss how SNPs in TP53 gene exon 5-8 influence developing of esophageal cancer.\n\nThe authors should mention the reason why the author focused on exon 5-8 in more detail.\n\nThe author need to show the relationship between SNPs in TP53 gene exon 5-8 and clinicopathological features, including survival information.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1741
https://f1000research.com/articles/7-1739/v1	02 Nov 18	{ "type": "Research Article", "title": "A protocol for extraction of total RNA from finger stick whole blood samples preserved with TempusTM solution", "authors": [ "Basirudeen Syed Ahamed Kabeer", "Sara Tomei", "Valentina Mattei", "Tobias Brummaier", "Rose McGready", "Francois Nosten", "Damien Chaussabel", "Sara Tomei", "Valentina Mattei", "Tobias Brummaier", "Rose McGready", "Francois Nosten" ], "abstract": "Monitoring of blood transcriptional changes during disease or treatment could improve the understanding of cellular mechanisms associated with that particular condition. This can be achieved through serial sampling of small blood volumes. However, molecular analysis of gene expression from low volume samples remains a challenging task. To address this issue, we have developed a set of standard operating procedures (SOP), starting from collection of small volume blood to measurement of gene expression. Previously we published an SOP for the collection of a small volume of blood via finger stick and stabilization of RNA. The aim of this manuscript is to share a modified TempusTM solution based RNA extraction method, developed in our lab, for the extraction of total RNA from low volume whole blood samples collected via finger stick.", "keywords": [ "Transcriptome", "Finger stick", "Whole blood", "RNA extraction", "Gene expression", "TempusTM" ], "content": "Introduction\n\nThe transcriptome is the complete set of transcripts in a specific type of cell or tissue. Transcriptome datasets can be leveraged to understand the genes or pathways associated with particular conditions which will help to develop diagnostic biomarkers and to identify new therapeutic targets1. Although transcriptional profiles of target diseased tissues or cells are ideal biotypes for such analyses, procuring tissue biopsies/cells and extracting a sufficient amount of RNA from these specimens prove often impossible in clinical settings. Therefore, whole blood is often considered as an alternative surrogate tissue in clinical research2,3. Blood plays a crucial role in immunity, inflammation and physiological homeostasis. Blood-based profiles also constitutute a powerful means for exploring basic biology and for approaching the complexity of biological systems.\n\nThe rapid advances in transcriptome profiling technologies, such as microarray and next-generation sequencing, made it possible to measure simultaneously the abundance of RNA on a genome-wide scale. High throughput RT-PCR and NanoString offer the opportunity to profile hundreds of targets at a lower cost than sequencing technologies. Overall, practicality as well as affordability allow studying changes in blood transcript abundance in infection, treatment or specific conditions and enable to maximize information obtained from each patient. Correlating serial blood transcriptome markers with the clinical course of disease has been shown to be a potential approach in diagnostics and for assessment of treatment response4–8.\n\nMinimization of the technical variance in any assay plays a critical role in the measurement of true biological variance. In transcriptome studies, the sources of technical variance can be considerable. In particular, RNA isolation and purification steps greatly influence the results of gene expression profiling, since RNA is a highly unstable molecule that is easily degraded by RNases which are ubiquitous in the environment9. Therefore, extra care has to be taken during this process. Furthermore, the protocol used for the extraction of RNA should (i) provide quantitative recovery of RNA that is intact and free from contaminants and (ii) keep the sample as concentrated as possible for further downstream analysis. There are commercial RNA whole blood collection tubes available in the market; PAXgene™ Whole Blood RNA isolation system (Qiagen, Germany) and Tempus™ Whole Blood RNA isolation system (ThermoFisher Scientific, USA). They have a significant advantage as they lyse whole blood at the time of collection, while simultaneously stabilizing RNA for later purification. However, these collection systems require drawing of 2.5 ml to 3 ml of venous blood at each collection time point, which can be challenging in some settings (e.g. pediatric populations, high frequency collection, home self-collection, field collection).\n\nFinger-stick blood collection is a widely used and safe method for the collection of blood, especially when only small volumes are required10–13. Major advantages of this collection method are that it is less invasive, quicker and can be performed without a trained phlebotomist. Therefore, it is more amenable to field applications and repeated sampling14. A study by Robison et al. found that gene expression measured with venous and finger stick blood collection is comparable15. However, currently available microtainers used for collection of whole blood samples via finger stick methods do not contain any RNA stabilizing solutions. Therefore, this method requires modified protocols for collection of blood, stabilization of samples and extraction of RNA. Recently, we published a detailed a standard operating procedure (SOP) for finger stick blood collection and RNA stabilization16.\n\nThere are few published reports which describe the procedure for extraction of RNA from a small volume of whole blood. A study by Carrol et al. used small volumes of blood (≥300μl) along with a modified PAXgene protocol to obtain high-quality RNA from pediatric samples17. Another study shows the feasibility of an RNA extraction protocol from only 70μl of whole blood collected via finger sticks15. Krawiec et al. successfully extracted RNA from even smaller blood samples from mouses or rats18. However, all these studies used PAXgene based protocols for the purpose of extracting total RNA. Reported yields and quality of RNA stabilized in TempusTM solution was generally greater when compared to PAXgene solution6,19,21–25. When PAXgene™ and Tempus™ were compared by using microarrays as the readout, several known phytohemagglutinin (PHA) inducible genes were only found to be up-regulated when RNA was isolated using the Tempus™ method, but not using the PAXgene™ method22. Considering these factors, we have developed a modified TempusTM solution-based RNA extraction method for the extraction of total RNA from low volume whole blood samples (50μl of whole blood). This method is currently employed in the context of a pregnancy monitory study being conducted on the Thai-Myanmar border (manuscript in preparation). The study aims to assess transcriptional changes in women during pregnancy and in the mother and child post-partum. Overall ~20,000 whole blood samples will be collected from 400 mother and child pairs. The low volume blood sample collection and RNA stabilization method published earlier16 and the related RNA extraction method described below are being shared with an anticipation that they may be of use to others and be improved through comments from reviewers and readers.\n\n\nNarrative of the procedure\n\nThe procedure described in this article can be used for extraction and quality control of total RNA from a small volume of whole blood preserved with TempusTM solution. A narrative is provided in this section, along with general remarks and considerations. A detailed point-by-point SOP follows.\n\nTempusTM spin RNA isolation kits were used for extraction of total RNA from whole blood lysate. The standard protocol recommended by the manufacturer is optimized for 3 ml of whole blood preserved with 6 ml of TempusTM stabilizing solution. Therefore, modifications were made in order to process 50μl of whole blood collected via finger stick and preserved with 100μl of TempusTM solution. Briefly, whole blood lysate is thawed and washed with 1xPBS (phosphate buffer saline) to obtain the RNA pellet. RNA purification is achived by using RNA purification filters and wash buffers. DNase treatment is performed to remove DNA contamination and finally purified RNA is eluted from the column using elution solution. RNA yields and quality are measured on NanoDrop and Fragment Analyzer.\n\n1) Blood sample collection, storage and shipment\n\nThe detailed protocol for collection and storage of whole blood samples for gene expression studies is available in our recent publication16. Briefly, 50μl of blood samples are collected using a plastic capillary straw in a microcentrifuge tube. Following thorough mixing with 100μl of TempusTM RNA stabilizing solution it is stored at -20°C, preferably, or alternatively -80°C. The transcriptional profile is maintained due to effective stabilization of RNA and will accurately reflect the physiological state of the patient at the time of the blood draw. As mentioned above, this sampling protocol is being used for one of our studies which investigates alteration in temporal transcriptional and microbiome trajectories preceding pre-term birth. For this study, blood samples are collected in Thailand and transfered to Qatar on dry ice. RNA yields and integrity reported below indicate that this shipment method permits recovery of nucleic acid in quantities and quality that meet requirements of downstream applications such as RNAseq or PCR.\n\n2) RNA extraction protocol\n\nThis protocol is developed by incorporating the following modifications to the standard Tempus™ Spin RNA isolation protocol for the extraction of total RNA from TempusTM cell lysate:\n\n(a) Washing with PBS\n\nAdd 5μl of 1x PBS to the 150μl of TempusTM lysate, vortex, and centrifuge. After washing, resuspend the RNA pellet in 400μl of RNA Purification Resuspension Solution.\n\n(b) DNase treatment\n\nPerform the DNase treatment as per the recommended protocol. This is an optional step in the manufacturer’s recommended protocol.\n\n(c) Incubation with Nucleic acid purification elution solution\n\nAdd Nucleic Acid Purification Elution Solution to the samples and incubate at 70°C for 1 minute. In our lab, extending the incubation time, for instance to 2–5 minutes, did not help to improve RNA yield.\n\n(d) Elution of RNA\n\nAt the final step, elute the purified RNA with either 25μl or 50μl of elution solution. The concentration of eluted RNA will range from 5 to 20ng/μl with 50μl of elution solution and will be >10ng/μl with 25μl of Elution Buffer.\n\n3) RNA yields and quality\n\nIn a set of 25 whole blood samples >200ng of total RNA could be extracted from 50μl of whole blood (mean±standard deviation: 503±170 ng; Range: 228–861 ng). The RNA integrity numbers (RIN) ranged from 5.9 to 9.2. (mean±standard deviation: 7.5±0.7). These figures are compatible for downstream applications such as RNAseq (input RNA: 150–200 ng), RT-PCR (input RNA for cDNA synthesis - depending on the kit: 10–500 ng) and NanoString (input RNA as little as 10 ng). Samples with an RIN > 5.3 is shown to be sufficient for downstream applications such as RNA-seq26 or RT-PCR27, while degraded RNA is suitable as input for the Nanostring assay (samples with RIN as low as 328. Table 1 and Figure 1 show the quality control analysis of total RNA extracted from selected whole blood samples. It is important to note that this method may not be used for extraction of small RNA such as miRNA since smaller RNAs (i.e. <200nt) are washed off during purification step. Alternative options for RNA extraction methods allowing retention of miRNAs include Norgen (Norgen Biotek Corporation) and MagMAX (ThermoFisher Scientific) RNA extraction kits, as they claim to provide suitable solutions for extraction of all sizes of RNA, from the large mRNA and ribosomal RNA down to microRNA.\n\nBlood samples (50ul) were collected from pregnant women via a finger stick. Total RNA was extracted using Tempus spin RNA extraction kit. At the end, purified RNA was eluted with 25ul or 50ul of elution buffer. A260:280 ratio was measured in Nanodrop. RNA integrity Number (RIN) was measured on Fragment Analyzer.\n\nRepresentative electrophorograms used for RNA quality analysis on Fragment analyzer.\n\n\nMaterials and methods\n\nTempusTM Spin RNA isolation kit (ThermoFisher Scientific, MA, USA; Catalog Nubmer: 4380204; https://www.thermofisher.com/order/catalog/product/4380204)\n\nAbsoluteRNA Wash Solution (Applied Biosystems, CA, USA; Catalog Number: 4305545; https://products.appliedbiosystems.com/ab/en/US/adirect/ab?cmd=catNavigate2&catID=602364)\n\nEppendorf pipettes (10 ul, 100 ul, 200 ul, 1000 ul)\n\nRNase free Axygen Filter tips (10 ul, 100 ul, 200 ul, 1000 ul)\n\nGloves (Cardinal Health Nitrile)\n\nStuart Vortex mixer\n\nThermo Scientific MicroCL 17R centrifuge\n\nEppendorf Thermo Mixer C\n\nNanodrop 8000 Spectrophotometer\n\nFragment Analyzer (Advanced Analytical - AATI)\n\nAATI Standard Sensitivity RNA Anlaysis Kit (DNF-471)\n\n1. Collect 50μl of whole blood into microcentrifuge tube containing 100μl of TempusTM solution. Upon blood collection, mix the samples thoroughly and store at -20°C or -80°C prior to RNA extraction – as described in detail in an earlier publication26.\n\n2. Before RNA extraction, thaw the frozen samples at room temperature.\n\n3. Add 50μl of 1x PBS to each sample and vortex the tubes vigorously for 30 seconds to ensure proper mixing. Then, centrifuge the samples at 4°C at 3,000g for 30 minutes.\n\n4. Pour off the supernatant and leave the tubes on absorbent paper for 1–2 minutes.\n\n5. Add 400μl of RNA Purification Resuspension Solution to the sample tubes and vortex briefly to resuspend the RNA pellet. Keep the RNA pellet on ice during the preparation for the next steps.\n\n6. Insert the RNA purification filter into waste collection tube and pre-wet it with 100 μl of Wash Solution 1.\n\n7. Add the resuspended RNA to the purification filters and centrifuge for 30 seconds at 16,000g.\n\n8. Remove the purification filters and discard the liquid waste. Re-insert the purification filters into the waste tube.\n\n9. Add 500μl of Wash Solution 1 into the purification filters and centrifuge for 30 seconds at 16,000g. Discard the samples’ flow through and re-insert purification filter into the waste tubes.\n\n10. Add 500μl of Wash Solution 2 to the purification filter and centrifuge the sample for 1 minute at 16,000g.\n\n11. Discard the flow through and re-insert the purification filter into the waste tube. Add 100μl of absolute RNA Wash Solution and incubate at room temperature for 15 minutes.\n\n12. Add 500μl of Wash Solution 2 into the purification filter, incubate at room temperature for 5 minutes and centrifuge for 30 seconds at 16,000g.\n\n13. Perform an additional wash with 500μl of Wash Solution 2. Discard the flow through and re-insert the purification filter into a clean collection tube.\n\n14. Add 25μl or 50μl of Elution Buffer to the purification filter. A lower volume of the Elution Buffer may increase RNA concentration while slightly reduce the overall yield.\n\n15. Incubate the sample for 2 minutes at 70°C and then centrifuge for 2 minutes at 16,000g.\n\n16. Measure quantity and quality of total RNA on Nanodrop and Fragment Analyzer. In order to avoid freezing thawing cycles an aliquot for RNA QC (5 μl) is taken at this point before proceeding with freezing.\n\n17. Store the purified RNA elute at -80°C for downstream analysis.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.", "appendix": "Grant information\n\nSupport for this work was provided by the Qatar Foundation.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe thank our study participants and all of the members of SMRU who assisted in this work (ClinicalTrials.gov Identifier NCT02797327). We thank Muna Al Hasmi, Nicola James and Rebecca Mathew for their participation in performing RNA extraction and RNA QC. We thank Darawan Rinchai for assiting in finger stick blood collection.\n\n\nReferences\n\nChaussabel D, Pulendran B: A vision and a prescription for big data-enabled medicine. Nat Immunol. 2015; 16(5): 435–439. PubMed Abstract \| Publisher Full Text\n\nDebey S, Schoenbeck U, Hellmich M, et al.: Comparison of different isolation techniques prior gene expression profiling of blood derived cells: impact on physiological responses, on overall expression and the role of different cell types. Pharmacogenomics J. 2004; 4(3): 193–207. PubMed Abstract \| Publisher Full Text\n\nDePrimo SE, Wong LM, Khatry DB, et al.: Expression profiling of blood samples from an SU5416 Phase III metastatic colorectal cancer clinical trial: a novel strategy for biomarker identification. BMC Cancer. 2003; 3: 3. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPascual V, Chaussabel D, Banchereau J: A genomic approach to human autoimmune diseases. Annu Rev Immunol. 2010; 28: 535–571. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTang BM, McLean AS, Dawes IW, et al.: Gene-expression profiling of peripheral blood mononuclear cells in sepsis. Crit Care Med. 2009; 37(3): 882–888. PubMed Abstract \| Publisher Full Text\n\nBerry MP, Graham CM, McNab FW, et al.: An interferon-inducible neutrophil-driven blood transcriptional signature in human tuberculosis. Nature. 2010; 466(7309): 973–977. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMartinez-Llordella M, Lozano JJ, Puig-Pey I, et al.: Using transcriptional profiling to develop a diagnostic test of operational tolerance in liver transplant recipients. J Clin Invest. 2008; 118(8): 2845–2857. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGaucher D, Therrien R, Kettaf N, et al.: Yellow fever vaccine induces integrated multilineage and polyfunctional immune responses. J Exp Med. 2008; 205(13): 3119–3131. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBecker C, Hammerle-Fickinger A, Riedmaier I, et al.: mRNA and microRNA quality control for RT-qPCR analysis. Methods. 2010; 50(4): 237–243. PubMed Abstract \| Publisher Full Text\n\nHarrison JR, Bevan J, Furth EE, et al.: AccuStat whole blood fingerstick test for Helicobacter pylori infection: a reliable screening method. J Clin Gastroenterol. 1998; 27(1): 50–53. PubMed Abstract \| Publisher Full Text\n\nLaine L, Knigge K, Faigel D, et al.: Fingerstick Helicobacter pylori antibody test: better than laboratory serological testing? Am J Gastroenterol. 1999; 94(12): 3464–3467. PubMed Abstract \| Publisher Full Text\n\nSblendorio V, Palmieri B, Riccioni G: Blood cholesterol concentration measured by CR3000: fingerstick versus venous sampling. Int J Immunopathol Pharmacol. 2008; 21(3): 729–733. PubMed Abstract \| Publisher Full Text\n\nTamborlane WV, Kollman C, Steffes MW, et al.: Comparison of fingerstick hemoglobin A1c levels assayed by DCA 2000 with the DCCT/EDIC central laboratory assay: results of a Diabetes Research in Children Network (DirecNet) Study. Pediatr Diabetes. 2005; 6(1): 13–16. PubMed Abstract \| Publisher Full Text\n\nObermoser G, Presnell S, Domico K, et al.: Systems scale interactive exploration reveals quantitative and qualitative differences in response to influenza and pneumococcal vaccines. Immunity. 2013; 38(4): 831–844. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRobison EH, Mondala TS, Williams AR, et al.: Whole genome transcript profiling from fingerstick blood samples: a comparison and feasibility study. BMC Genomics. 2009; 10: 617. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRinchai D, Anguiano E, Nguyen P, et al.: Finger stick blood collection for gene expression profiling and storage of tempus blood RNA tubes [version 1; referees: 3 approved with reservations]. F1000Res. 2016; 5: 1385. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCarrol ED, Salway F, Pepper SD, et al.: Successful downstream application of the Paxgene Blood RNA system from small blood samples in paediatric patients for quantitative PCR analysis. BMC Immunol. 2007; 8: 20. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKrawiec JA, Chen H, Alom-Ruiz S, et al.: Modified PAXgene method allows for isolation of high-integrity total RNA from microlitre volumes of mouse whole blood. Lab Anim. 2009; 43(4): 394–398. PubMed Abstract \| Publisher Full Text\n\nMejias A, Dimo B, Suarez NM, et al.: Whole blood gene expression profiles to assess pathogenesis and disease severity in infants with respiratory syncytial virus infection. PLoS Med. 2013; 10(11): e1001549. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOswald M, Curran ME, Lamberth SL, et al.: Modular analysis of peripheral blood gene expression in rheumatoid arthritis captures reproducible gene expression changes in tumor necrosis factor responders. Arthritis Rheumatol. 2015; 67(2): 344–351. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nQuerec TD, Akondy RS, Lee EK, et al.: Systems biology approach predicts immunogenicity of the yellow fever vaccine in humans. Nat Immunol. 2009; 10(1): 116–125. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAsare AL, Kolchinsky SA, Gao Z, et al.: Differential gene expression profiles are dependent upon method of peripheral blood collection and RNA isolation. BMC Genomics. 2008; 9: 474. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNikula T, Mykkänen J, Simell O, et al.: Genome-wide comparison of two RNA-stabilizing reagents for transcriptional profiling of peripheral blood. Transl Res. 2013; 161(3): 181–188. PubMed Abstract \| Publisher Full Text\n\nHäntzsch M, Tolios A, Beutner F, et al.: Comparison of whole blood RNA preservation tubes and novel generation RNA extraction kits for analysis of mRNA and MiRNA profiles. PLoS One. 2014; 9(12): e113298. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPankla R, Buddhisa S, Berry M, et al.: Genomic transcriptional profiling identifies a candidate blood biomarker signature for the diagnosis of septicemic melioidosis. Genome Biol. 2009; 10(11): R127. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShen Y, Li R, Tian F, et al.: Impact of RNA integrity and blood sample storage conditions on the gene expression analysis. Onco Targets Ther. 2018; 11: 3573–3581. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFleige S, Pfaffl MW: RNA integrity and the effect on the real-time qRT-PCR performance. Mol Aspects Med. 2006; 27(2–3): 126–139. PubMed Abstract \| Publisher Full Text\n\nSgarilia R, Pisapia P, Nacchio M, et al.: Multiplex digital colour-coded barcode technology on RNA extracted from routine cytological samples of patients with non-small cell lung cancer: pilot study. J Clin Pathol. 2017; 70(9): 803–806. PubMed Abstract \| Publisher Full Text" }	[ { "id": "40836", "date": "05 Dec 2018", "name": "Sunil M. Kurian", "expertise": [ "Reviewer Expertise Organ Transplant", "Genomics", "Biomarkers" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article presents a modified version of a finger stick method of RNA extraction using an already-established Tempus system for RNA. While the study is relevant especially for quick and easy collection of RNA in specific cases such as pediatric samples as well as samples that have to be collected in large populations in a quick and reliable way, I do have some questions and clarifications regarding the study:\nWhile they have been through specific differences between the Tempus and the PaxGene systems for RNA extraction, in this study with the modified methodology there have been no specific attempts to do downstream processing in terms of RNA sequencing or microarray analysis to compare the results. It is important that the authors do not just assume that previous results may replicate despite modifications to the protocol. A quick comparison using a microarray, or a targeted RNAseq method would actually sort out the true differences between the modified method and a standard PaxGene or a PaxGene methodology that has been modified for finger stick purposes.\n\nAs a follow-up to the first question, a downstream processing step to compare RNA quality becomes even more important in the context of having a range of RNA Integrity numbers. Do higher quality numbers translate to better QC when it comes to sequencing or microarrays? These answers will shed more light into the actual quality of RNA from this modified method.\n\nHave the authors tried to do a globin reduction step to remove globin transcripts which are highly abundant and may interfere with the quality of the other expressed RNA transcripts? What is their opinion about performing a globin reduction step? Many protocols suggest this step when dealing with whole blood samples, so it would be nice to get some input about it.\n\nSome comments on the wide range of RNA yields obtained need to be explained. Do the authors expect such a wide range in total RNA to be normal? Is it more of a technical limitation of the methodology that scales down a protocol that normally uses a few mls of whole blood to microliters of blood? What is the experience from their previous studies?", "responses": [] }, { "id": "42795", "date": "05 Aug 2019", "name": "Alexander Tolios", "expertise": [ "Reviewer Expertise Laboratory Medicine", "Platelets", "Genetics." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors present their methodology utilizing TempusTM RNA stabilizing solution and an adapted protocol to be used on very low blood volumes collected via finger stick. The topic is relevant since the easier access to blood and a reduced amount of blood needed is an obvious advantage. Never the less, the article presented has some limitations:\nBy comparing the RNA extracted using the authors methodology from finger stick blood with regularly obtained blood, the authors would be able to confidently evaluate if those two samples are actually comparable. This is currently missing. Following up to that, a statistical evaluation and comparison between those would definitely be relevant to the reader.\n\nAlthough total RNA quantity, RIN and A260/280 are provided, some downstream analysis of the RNA should be performed as well. The authors could for example, evaluate the results with a low-abundance-RNA as well as a high-abundance-RNA and compare those results to RNA extracted from regularly obtained blood.\n\nThe information provided in Table 1 seems superfluous. Although it can be of interest for a reader in some circumstance, I'd suggest adding the table as a supplement and not as table of the main article.", "responses": [] } ]	1	https://f1000research.com/articles/7-1739
https://f1000research.com/articles/7-1736/v1	02 Nov 18	{ "type": "Method Article", "title": "Quantifying steroid hormones in amniotic fluid by ultra-performance liquid chromatography and tandem mass spectrometry", "authors": [ "Noëllie Rivet", "Carole Jamey", "Nathalie Reix", "Pascal Kintz", "Martin Heil", "Kathrin Erdmann", "Lisa M. Körner", "Judith Lawrenz", "Susanne Fröhlich", "Peter Kozlowski", "Gunther Meinlschmidt", "Marion Tegethoff", "Jean-Sébastien Raul", "Noëllie Rivet", "Carole Jamey", "Nathalie Reix", "Pascal Kintz", "Martin Heil", "Kathrin Erdmann", "Lisa M. Körner", "Judith Lawrenz", "Susanne Fröhlich", "Peter Kozlowski", "Gunther Meinlschmidt", "Marion Tegethoff" ], "abstract": "Background: Simultaneous assessment of steroid hormone concentrations in amniotic fluid is of importance for elucidating long-term consequences of intrauterine processes, and of broad scientific and clinical relevance. The objective of the study was to develop sensitive and specific analytical ultra-performance liquid chromatography tandem mass spectrometry (UPLC-MS/MS) methods for determination of steroid hormone concentrations in human amniotic fluid, and to provide proof-of-concept of their applicability. Methods: Methods were validated according to linearity, limit-of-detection, limit-of-quantification, recovery, intra- and inter-assay precision, and applied to 275 amniotic fluid samples. Results: Limits-of-quantification (S/N=10:1) were 0.05 ng/mL for cortisol, dehydroepiandrosterone sulfate (DHEAS), estradiol, estriol, and testosterone, and 0.01 ng/mL and 1.0 ng/mL for cortisone and dehydroepiandrosterone (DHEA), respectively. Good inter- and intra-assay precision were observed. Cortisol, cortisone, DHEAS, estradiol and estriol concentration were quantified in all samples. By lack of sensitivity of the analytical method, DHEA was quantified in 11 samples only. Testosterone was quantified in 119 of 275 samples analyzed. 116 of the quantified testosterone samples were from male offspring (out of 138 male offspring). Conclusion: These specific and sensitive methods offer a simple and non-invasive way to measure cortisol, cortisone, DHEAS, estradiol, estriol, and testosterone concentrations in human amniotic fluid.", "keywords": [ "bioanalytical methods", "hypothalamic-pituitary-adrenal (HPA) axis", "prenatal programming", "steroids", "ultra performance liquid chromatography–tandem mass spectrometry (UPLC–MS/MS)" ], "content": "Introduction\n\nAmniotic fluid, with its nutritive and protective functions, is essential for fetal well-being and development1. It is contained by the amniotic cavity enclosed by the amnion and chorion2. Amniotic fluid originally comes from maternal plasma and passes through the fetal membranes, being fueled by secretions from the fetal respiratory and gastrointestinal tract, from the oral cavity, and from fetal skin before keratinization occurs1,3–5. Furthermore, cells from the amnion layers secrete proteins into the amniotic fluid2. At the beginning of pregnancy, the biochemical composition of human amniotic fluid resembles fetal extracellular fluid, probably due to passage across the still unkeratinized fetal skin6. Given the interplay between amniotic fluid and fetal physiology, amniotic fluid has been used as an important source of potential biomarkers for fetal pathologies7.\n\nFetal endogenous steroids, including cortisol, cortisone, dehydroepiandrosterone (DHEA), dehydroepiandrosterone sulfate (DHEAS), estradiol, estriol, and testosterone are involved in broad and essential physiological functions, including glucose- and energy-metabolism, development, timing of pregnancy, adaptation to the extra uterine environment, and immune processes [e.g. 8,9]. Dysfunction or dysregulation of underlying hormone systems (e.g. the hypothalamic-pituitary-adrenal axis) is seen in major endocrine diseases (e.g. Cushing’s syndrome, Addison’s disease)10,11 and non-endocrine disorders (including mental disorders and stress-related conditions)12–16. Further, information on prenatal steroid hormone concentrations is of relevance for elucidating fetal origins of health, disease, and variation in mental processes17–24. Hence, the assessment of these hormones is of clinical and scientific relevance for a range of different disciplines, including endocrinology, psychiatry, psychobiology, psychology, neonatology, pediatrics, gynecology, and internal medicine [e.g. 22].\n\nTherefore, there is a strong need for powerful and easily applicable analytical procedures for research as well as diagnostic purposes, to assess fetal steroids in amniotic fluid, overcoming previous limitations of sensitivity and specificity. However, as yet, respective methods are scarce. Steroid hormones can be quantified by traditional immunoassays but with several limitations as a lack of sensitivity making the quantification of small amounts of steroids difficult or a lack of specificity due to antibody cross-reactivity resulting in results higher than true concentrations25. The presence of interfering substances such as autoantibodies, exogenous substances, hemolysis or lipaemia in patient samples can alter the measurable concentration of the analyte26. The matrix effect problems also affect immunoassays. Furthermore, a relatively large sample volume is required as immunoanalysis provides only single-analyte assays and this is particularly a problem for precious samples such as amniotic fluid. Mass-spectrometric techniques are an alternative to address immunoassays issues. It allows for highly sensitive and specific simultaneous assessment of multiple biomarkers, for example in the context of metabolic and proteomic profiling in amniotic fluid7,25. Recently, first studies reported on the successful application of different mass-spectrometric approaches, including gas-chromatography mass spectrometry (GC-MS) and liquid-chromatography tandem mass spectrometry (LC-MS/MS), for concomitant assessment of multiple selected steroids in amniotic fluid26–28.\n\nThe main objective of this study was to develop highly sensitive and specific analytical ultra-performance liquid chromatography tandem mass spectrometry (UPLC-MS/MS) methods for determination of cortisol, cortisone, DHEA, DHEAS, estradiol, estriol, and testosterone concentrations in human amniotic fluid, and provide proof-of-principle for the method’s applicability in amniotic fluid samples of pregnant women, collected within obstetric care.\n\n\nMethods\n\n282 pregnant women were recruited via Prenatal Medicine and Genetics, Duesseldorf, Germany, who were undergoing regular amniocenteses for diagnostic reasons with maternal age as the only risk factor. Samples from seven women were not included in the analyses because of problems with either the consent forms or the amniotic fluid samples. Sociodemographic and obstetric characteristics of the 275 women included in the analyses are presented in Table 1. Amniotic fluid samples were collected in the years 2010 to 2012, between week 14 and week 18 of gestation. Samples were collected, using GREINER BIO-ONE (Frickenhausen, GERMANY) tubes CRYO’S 124 and stored at –80°C for later analyses. Biochemical analyses were conducted in 2013 at the laboratory of the Institute of Legal Medicine at the University of Strasbourg, France.\n\n1Percentages may not total 100 due to rounding;\n\nAll participants provided written informed consent that their samples were used in research after genetic examinations. The study protocol was approved by the Ethical Committee of Faculty of Science and Mathematics, Heinrich-Heine-University, Duesseldorf, Germany.\n\nChemicals. Dichloromethane was purchased from Carlo Erba (Val de Reuil, France). HPLC-grade methanol (MeOH), isopropanol, diethyl ether, ammonium hydroxide (NH4OH) and formic acid (HCOOH) were purchased from VWR (Fontenay-sous-bois, France). HPLC-grade acetonitrile (ACN) was obtained from Merck (Darmstadt, Germany) and Oasis HLB (30mg/30µm) solid-phase microextraction plate was obtained from Waters (Milford, USA). Sodium bicarbonate (NaHCO3), dansyl-chloride, cortisol, cortisone, DHEA, testosterone, estradiol and estriol were obtained from Sigma (Saint-Quentin Fallavier, France) while deuterated cortisol (cortisol-d4, 9, 11, 12, 12 D) and DHEAS were obtained from Steraloids (Newport, USA).\n\nSolution preparation. Cortisol, testosterone and estradiol solutions were prepared in MeOH at a final concentration of 2.5, 25 and 250 ng/mL. Cortisone, DHEAS, estriol solutions were prepared in MeOH at a final concentration of 2.5, 25, 250 and 2500 ng/mL. DHEA solutions were prepared in MeOH at a final concentration of 0.025, 0.25, and 2.5 and 25 µg/mL. Deuterated cortisol was prepared in MeOH at a final concentration of 250 ng/mL. These solutions were stable for at least 6 months at 4°C. Sorensen buffer was prepared by adding 38.8 mL KH2PO4 (9.07 g/L) to 6.12 mL Na2HPO4 (11.87 g/L); pH value was adjusted to pH 7.6. NaHCO3 buffer was prepared by adding 1 L of distilled water to 8,4 g of NaHCO3; pH value was adjusted to pH 10.5. Dansyl chloride solution was prepared by adding 20 mL of acetone to 20 mg of dansyl chloride.\n\nCalibration standards and quality control. Distilled water was used for calibration standards. All molecules of interest were thus not detectable in the blank samples. Calibration standards were prepared at concentrations ranging from limit-of-quantification (LOQ) to high concentration for each analyte. Intra- and inter-assay precision were determined by enriching 250 µL of distilled water with the analytes of interest at three concentrations (low, medium and high) for each hormone.\n\nSample preparation for cortisol, cortisone, DHEA, DHEAS and testosterone determination. 0.25 mL of amniotic fluid were mixed in 0.75 mL Sorensen buffer (pH 7.6) in the presence of 20 ng/mL cortisol-d4 as internal standard. For further purification, SPME Oasis® HLB extraction plates were used. Activation was operated with 0.2 mL MeOH, followed by 0.2 mL deionized water. The incubation medium was centrifuged and the supernatant was removed and deposited on the activated plate, then rinsed with 0.2 mL deionized water/MeOH (95:5, v/v). The plate was allowed to dry for 5 minutes at room temperature. Analytes were eluted with 35 µL ACN/isopropanol (40:60, v/v) with 2% of concentrated NH4OH, followed by 35 µL deionized water. Ten µL of this extract were directly injected into the UPLC-MS/MS system.\n\nSample preparation for estradiol and estriol determination. 3 mL of diethylether was added to 0.25 mL of amniotic fluid in the presence of 20 ng/mL cortisol-d4 as internal standard. Samples were agitated 15 minutes. Samples were centrifuged 15 minutes at 3000g and the organic phase was transferred into a new glass tube. The supernatant was evaporated at 40°C under a constant stream of nitrogen until the samples were completely dried. Finally, 50 µL of NaHCO3 buffer (pH 10.5) and 50µL of dansyl-chloride solution were added and the tube was vortexed for 30 second. The mix was heated 3 minutes at 60°C. Ten µL of this extract were directly injected into the UPLS-MS/MS system.\n\nChromatographic and mass spectrometric conditions. A Waters (Milford, USA) Acquity UPLC system with a column heater, autosampler, and a 10µL injection loop was used. Analytes were separated at 30°C on a Waters Acquity UPLC BEH C18 column (1.7 µm, 100 x, 2.1-mm). Separation was achieved by gradient elution with 0.1% HCOOH (pH 2.6) and ACN at a flow rate of 0.4 mL/min (0 to 6 min, 90% HCOOH and 10% ACN; 6 to 9 min, 40% HCOOH and 60% ACN; 9 to 11 min, 90% HCOOH and 10% ACN). The total run time was 11 minutes, including periods required for injection and equilibrating the column before the next injection.\n\nDetection was carried out by a Quattro Premier XE tandem mass spectrometer (MS/MS) (Waters Micromass, Manchester, UK). This mass spectrometer was equipped with an electrospray ionization probe and operated switching between positive and negative ionization mode. The voltage of the capillary was 3.5 kV in positive mode. The ion-source temperature was 120°C and the desolvation gas was heated to 400°C at a flow rate of 800L/h. Quantitative results were obtained in MRM (Multiple Reactions Monitoring) mode after determination of the transition for each glucocorticoid: cortisone m/z 361.1>163.1, cortisol m/z 363.2>121.1, DHEA m/z 289.2>253.2, DHEAS m/z 367.1>96.9 (ES-), testosterone 289.2>97.1, estriol 522.3>170.9 and estradiol 506.3>170.9, against cortisol-d4 m/z 367.2 > 121.1. Qualifier ions were monitored, being m/z 361.1>121.1 for cortisone; m/z 363.2>309.1 and m/z 363.2>327.1 for cortisol; m/z 289.1>213.1 for DHEA; m/z 289.2>109.1 for testosterone, m/z 522.3>155.0 for estriol and m/z 506.3>155.9 for estradiol. Masslynx®4.1 software (Waters) was used for data acquisition.\n\nThe following parameters were tested to validate the method: limit-of-detection (LOD), LOQ, linearity, recovery, intra- and inter-assay precision.\n\nThe LOD and LOQ were determined by analysis of replicate blank samples (n=6) spiked with hormones at various concentrations. The LOD and LOQ were estimated as giving a signal-to-noise ratio greater than 3 and 10 respectively, for each of the quantitative ions transitions monitored.\n\nLinearity was tested by the preparation of calibration curves ranging from the LOQ to 20 ng/mL for cortisol, to 75 ng/mL for cortisone, to 500 ng/mL for DHEA, to 50 ng/mL for DHEAS, to 10 ng/mL for testosterone, to 100 ng/mL for estriol and to 25 ng/mL for estradiol. Three curves were developed for 9 concentration levels of each hormone. Linearity of the method was expressed by the correlation coefficient (r2).\n\nAnalyte recovery of the extraction procedure was determined by analyzing replicate blank samples (n=3) spiked with hormone (LOQ, medium and high concentration) against replicate blank extracts (n=3) spiked at the same levels after extraction.\n\nPrecision was evaluated using three solutions with all hormones (LOQ, medium and high concentration) and expressed as coefficient of variation (CV). The intra-assay precision was assessed by determining these samples on one day (n=10 for each sample), while the inter-assay precision was assessed over 8 days (n=8 for each sample).\n\n\nResults\n\nUnder the UPLC conditions described above, cortisol, cortisone, DHEA, DHEAS and testosterone were sufficiently separated chromatographically (Figure 1). The mean (± standard deviation) retention times were 4.35 ± 0.05 min for cortisol, 4.40 ± 0.04 min for cortisone, 6.35 ± 0.04 min for DHEA, 5.75 ± 0.04 min for DHEAS 6.13 ± 0.04 min for testosterone and 4.38 ± 0.08 min for cortisol-d4. Estradiol and estriol were separated chromatographically under UPLC conditions (Figure 2). The retention times were 8.00 ± 0.04 min for estradiol and 7.16 ± 0.04 min for estriol.\n\nCortisol, cortisone, DHEA, DHEAS, testosterone assessment\n\nLOD, LOQ and linearity\n\nThe LOD and LOQ, as well as the CVs are presented in Table 2, for both cortisol, cortisone, DHEA, DHEAS, testosterone.\n\nAbbreviations: CV, coefficient of variation; DHEA, dehydroepiandrosterone; DHEAS, dehydroepiandrosterone sulfate; LOD, limit-of-detection; LOQ, limit-of-quantification\n\nThe calibration curve obtained showed good linear responses with a r2 of 0.9974, 0.9907, 0.9956, 0.9856 and 0.9940 for cortisol, cortisone, DHEA, DHEAS and testosterone respectively from the range of 0.05 to 20 ng/mL for cortisol, 0.01 to 75 ng/mL for cortisone, 1 to 500 ng/mL for DHEA, 0.05 to 50 ng/mL for DHEAS and 0.05 to 10 ng/mL for testosterone.\n\nExtraction recovery\n\nTable 3 shows the recovery tested for each analyte at three different concentrations (LOQ, medium and high concentrations). The extraction recoveries were > 71%, > 86%, > 73%, > 80% and > 52% for cortisol, cortisone, DHEA, DHEAS and testosterone respectively.\n\nAbbreviations: CV, coefficient of variation; DHEA, dehydroepiandrosterone; DHEAS, dehydroepiandrosterone sulfate\n\nPrecision\n\nIntra- (n=10) and inter-assay (n=8) precision were evaluated for each hormone at 3 concentrations (LOQ, medium, and high). The intra-assay (n=10) precision values were less than 20% for all hormones at LOQ and less than 15% for all hormones at medium and high concentration. Good inter-assay (n=8) values (less than 20% for both analytes at LOQ and less than 15% at other medium and high concentrations) were obtained, as outlined in Table 4.\n\nAbbreviations: CV, coefficient of variation; DHEA, dehydroepiandrosterone; DHEAS, dehydroepiandrosterone sulfate\n\nEstradiol and estriol assessment\n\nLOD, LOQ and linearity\n\nThe LOD and LOQ, as well as the CVs are presented in Table 5, for both estradiol and estriol.\n\nAbbreviations: CV, coefficient of variation; LOD, limit-of-detection; LOQ, limit-of-quantification\n\nThe calibration curves showed good linear responses with a r2 of 0.9967 and 0.9979 for estradiol and estriol respectively, within the range of 0.01 to 25 ng/mL for estradiol and 0.05 to 100 ng/mL for estriol.\n\nExtraction recovery\n\nTable 6 shows the recovery tested for each analyte at three different concentrations (LOQ, medium and high). The extraction recoveries were > 77% and > 92% for estradiol and estriol respectively.\n\nAbbreviation: CV, coefficient of variation\n\nPrecision\n\nIntra- (n=10) and inter-assay (n=8) precision were evaluated for estradiol and estriol at 3 concentrations (LOQ, medium and high). The intra-assay (n=10) values were less than 16% for both estradiol and estriol and good inter-assay (n=8) values (less than 17% for both analytes) were obtained, as demonstrated in Table 7.\n\nAbbreviation: CV, coefficient of variation\n\nApplication on amniotic fluid samples. UPLC-MS/MS analyses were performed on 275 amniotic fluid samples (see Table 8, Figure 3 and Figure 4). Cortisol, cortisone, DHEAS, estradiol and estriol concentrations were quantified in all samples. DHEA was quantified in 11 of the 275 samples analyzed. Testosterone was quantified in 119 of the 275 samples analyzed (samples from pregnancies with male offspring – all twin or higher order pregnancies resulted in male offspring only: quantified in 116 out of 138 samples analyzed; samples from pregnancies with female offspring: quantified in 3 out of 126 samples analyzed). For DHEA, in 35 of the analyzed samples, concentrations were below the LOQ of 1 ng/mL but above the limit-of-detection of 0.75 ng/mL. For testosterone, in 39 of the analyzed samples, concentrations were below the LOQ of 0.05 ng/mL but above the limit-of-detection of 0.02 ng/mL.\n\n1Based on quantification in 11 samples; 2Based on quantification in 119 samples; Abbreviations: DHEA, dehydroepiandrosterone; DHEAS, dehydroepiandrosterone sulfate\n\n\nDiscussion\n\nWe developed an analytical method to simultaneously measure cortisol, cortisone, DHEAS, and testosterone, and an analytical method to simultaneously measure estradiol and estriol in human amniotic fluid, using an LC-MS/MS assay, with good linearity, sensitivity, specificity and accuracy without interferences between the molecules and a low LOQ. Five out of the seven hormones were quantifiable in all amniotic fluid samples, while DHEA was quantifiable only in a few samples and testosterone was primarily quantifiable in samples from women carrying a male fetus.\n\nThe results add to previous findings that indicated that different mass-spectrometric approaches, including gas-chromatography mass spectrometry (GC-MS) and LC-MS/MS, can be applied for assessment of steroids in amniotic fluid27,28.\n\nComparisons of our results on steroid concentrations in human amniotic fluid with data from previous studies are hampered by substantial heterogeneity across previous reports. For example, even when restricting comparisons to findings also based on LC-MS/MS, cortisol concentrations in human amniotic fluid have previously been reported as higher28 as well as substantially lower29 than the here reported values. Reasons for such heterogeneity may include – besides the analytical procedures themselves – differences across studies with regard to i) indication for amniotic fluid collection, ii) gestational age at collection, iii) other sociodemographic or obstetric characteristics related to the pregnancy, iv) procedures applied for amniotic fluid collection and storage, and v) the interval between collection and biochemical analyses of amniotic fluid [cf. 1,29–31]. Overall, the here reported concentrations are compatible in magnitude with previous findings on cortisol, cortisone, DHEA, DHEA-S, and testosterone concentrations in human amniotic fluid, assessed via mass spectrometric approaches27,28, and with previous findings on estradiol and estriol concentrations in human amniotic fluid, assessed via immunoassay-based approaches32,33 (we are not aware of any respective information from previous studies using mass spectrometric approaches).\n\nOur study has several strengths. We developed analytical methods to quantify a range of steroid hormones in human amniotic fluid with low minimal amount of sample and a total analytical run time of 11 minutes, allowing simultaneous assessment of several hormones. Using state-of-the-art approaches, we confirmed high sensitivity and specificity, good accuracy and reproducibility of the analyses. Notably, the UPLC-MS/MS methods show no interference in contrast to many immunoassays. Last but not lest, by providing proof-of-principle for using the methods to quantify hormone concentrations in human amniotic fluid samples, we open the way for their further development for a wide range of potential applications in scientific and clinical settings.\n\nOur study has several limitations. First, for one out of the seven analytes, we were unable to quantify the steroids in all samples and for another one out of the seven analytes, we were unable only in samples from females to quantify the steroids, most likely because concentrations in the respective samples were below our LOQ or even LOD. For DHEA, the analytical method was not sufficiently sensitive to detect a concentration under 0.75 ng/ml. Although recoveries were greater than 80%, detection limits of the UPLC-MS/MS system have been reached for this molecule. The analytical study of steroids in amniotic fluid of midgestation made by Fahlbusch et al. showed DHEA concentrations of 0.64 ± 0.48 ng/mL for male group of fetuses and 0.56 ± 0.36 ng/mL for female group of fetuses28. According to these results, LOQ should be at least 0.1 ng/mL to detect DHEA in 275 samples analyzed. Similarly, the lack of analytical sensitivity also affects testosterone results. According to Fahlbusch et al., testosterone concentrations in amniotic fluid were 0.30 ± 0.15 ng/mL for male group of fetuses and 0.02 ± 0.02 ng/mL for female fetuses30. Kushnir et al. determined serum pediatric references ranges below 0.37 ng/mL for males aged 6 to 24 months and below 0.09 ng/mL for females of similar age34. In the same, Soldin et al. showed testosterone concentrations in serum range from 0.04 to 0.31 ng/mL for males (0–6 years) and range from 0.02 to 0.1 ng/mL for females (0–5 years)35. In view of these studies, the LOQ of our method was not sensitive enough to detect a low concentration of testosterone in amniotic fluid particularly for female fetuses. However, identifying a concentration as being below LOQ may also be of diagnostic value. Second, timing of amniotic fluid collection was not evenly distributed across pregnancy, so we cannot draw conclusions regarding changes in amniotic fluid steroid concentrations throughout gestation. Third, amniotic fluid samples were not immediately assessed after collection, so we cannot exclude degradation of analytes between sample collection and analysis, even though storage conditions would have made substantial degradation rather unlikely. Finally, as amniotic fluid is not routinely collected during every pregnancy, we cannot assume that our sample of study participants is representative with regard to the population of pregnant women. However, we drew our participants from women, undergoing amniotic fluid collection during obstetric care just for the reason that the pregnant women were above 30 years of age. Hence we expect our sample of study participants being largely comparable to a typical sample of women undergoing amniotic fluid testing.\n\nFuture studies should further address the role of storing conditions and time interval between collection and analyses. Moreover, future studies should aim at further increasing the sensitivity of the methods. Finally, further studies should test the functional significance of the methods and provide data on their validity and applicability in clinical and research settings.\n\nThe methods may provide new opportunities for future applications in a range of fields, such as endocrinology and beyond, providing a tool to determine steroid concentrations in pathological conditions characterized by disturbances of steroid hormones and related enzymatic activities.\n\nWe here demonstrated that cortisol, cortisone, DHEAS, estradiol, estriol, and testosterone concentrations are easily detectable and quantifiable in human amniotic fluid, using state-of-the-art and reproducible LC-MS/MS techniques with good linearity, sensitivity, specificity, and accuracy without interferences for the molecules and with a low LOQ. The methods may provide an appropriate approach for a wide range of clinical and research applications, including the field of endocrinology and others.\n\n\nData availability\n\nIndividual (non-aggregated) data cannot be made publicly available, due to ethical restrictions. In order to access this data, data must be requested from the corresponding author. Data requestors will have to provide: i) written description and legally binding confirmation that their data use is within the scope of the study, as outlined in the ethical board request and written informed consent (respective information will be provided by the corresponding author on request); ii) detailed written description and legally binding confirmation of their actions to be taken to protect the data (e.g., with regard to transfer, storage, back-up, destruction, misuse, and use by other parties), as legally required and to current national and international standards (data protection concept); and iii) legally binding and written confirmation and description that their use of these data is in line with all applicable national and international laws (e.g., the General Data Protection Regulation of the EU). In case of requests of hormone data only, less restrictive regulations may apply.", "appendix": "Grant information\n\nGM received funding from the Swiss National Science Foundation (SNSF) under project no. 100014_135328, funding from the Stanley Thomas Johnson Stiftung & Gottfried und Julia Bangerter-Rhyner-Stiftung under project no. PC_28/17, funding from the Forschungsfonds of the International Psychoanalytic University Berlin, and funding from the Swiss Cancer League KLS-4304-08-2017. MT receives funding from the SNSF under project no. PZ00P1_137023, and GM and MT receive funding from the Korea Research Foundation within the Global Research Network Program under project no. 2013S1A2A2035364.\n\nThe funders had no role in study design, in the collection, analysis and interpretation of data, in the writing of the report, and in the decision to submit the paper for publication.\n\n\nReferences\n\nUnderwood MA, Gilbert WM, Sherman MP: Amniotic fluid: not just fetal urine anymore. J Perinatol. 2005; 25(5): 341–348. PubMed Abstract \| Publisher Full Text\n\nParry S, Strauss JF 3rd: Premature rupture of the fetal membranes. N Engl J Med. 1998; 338(10): 663–670. PubMed Abstract \| Publisher Full Text\n\nMoore TR: Amniotic fluid dynamics reflect fetal and maternal health and disease. Obstet Gynecol. 2010; 116(3): 759–765. PubMed Abstract \| Publisher Full Text\n\nBrace RA: Physiology of amniotic fluid volume regulation. Clin Obstet Gynecol. 1997; 40(2): 280–289. PubMed Abstract \| Publisher Full Text\n\nBrace RA, Cheung CY: Regulation of amniotic fluid volume: evolving concepts. Adv Exp Med Biol. 2014; 814: 49–68. PubMed Abstract \| Publisher Full Text\n\nLiu H, Zheng Z, Wintour EM: Aquaporins and fetal fluid balance. Placenta. 2008; 29(10): 840–847. PubMed Abstract \| Publisher Full Text\n\nMichaels JE, Dasari S, Pereira L, et al.: Comprehensive proteomic analysis of the human amniotic fluid proteome: gestational age-dependent changes. J Proteome Res. 2007; 6(4): 1277–1285. PubMed Abstract \| Publisher Full Text\n\nChrousos GP, Gold PW: The concepts of stress and stress system disorders. Overview of physical and behavioral homeostasis JAMA. 1992; 267(9): 1244–1252. PubMed Abstract \| Publisher Full Text\n\nHill M, Parizek A, Cibula D, et al.: Steroid metabolome in fetal and maternal body fluids in human late pregnancy. J Steroid Biochem Mol Biol. 2010; 122(4): 114–132. PubMed Abstract \| Publisher Full Text\n\nFindling JW, Raff H: Cushing's Syndrome: important issues in diagnosis and management. J Clin Endocrinol Metab. 2006; 91(10): 3746–3753. PubMed Abstract \| Publisher Full Text\n\nTen S, New M, Maclaren N: Clinical review 130: Addison's disease 2001. J Clin Endocrinol Metab. 2001; 86(7): 2909–2922. PubMed Abstract \| Publisher Full Text\n\nAnagnostis P, Athyros VG, Tziomalos K, et al.: Clinical review: The pathogenetic role of cortisol in the metabolic syndrome: a hypothesis. J Clin Endocrinol Metab. 2009; 94(8): 2692–2701. PubMed Abstract \| Publisher Full Text\n\nde Kloet CS, Vermetten E, Geuze E, et al.: Assessment of HPA-axis function in posttraumatic stress disorder: pharmacological and non-pharmacological challenge tests, a review. J Psychiatr Res. 2006; 40(6): 550–567. PubMed Abstract \| Publisher Full Text\n\nSchleimer RP: Interactions between the hypothalamic-pituitary-adrenal axis and allergic inflammation. J Allergy Clin Immunol. 2000; 106(5 Suppl): S270–274. PubMed Abstract \| Publisher Full Text\n\nVogelzangs N, Beekman AT, Milaneschi Y, et al.: Urinary cortisol and six-year risk of all-cause and cardiovascular mortality. J Clin Endocrinol Metab. 2010; 95(11): 4959–4964. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nStirrat LI, Reynolds RM: The effect of fetal growth and nutrient stresses on steroid pathways. J Steroid Biochem Mol Biol. 2016; 160: 214–220. PubMed Abstract \| Publisher Full Text\n\nHeil M, Kavsek M, Rolke B, et al.: Mental rotation in female fraternal twins: Evidence for intra-uterine hormone transfer? Biol Psychol. 2011; 86(1): 90–93. PubMed Abstract \| Publisher Full Text\n\nPellkofer J, Jansen P, Heil M: Sex-specific lateralization of event-related potential effects during mental rotation of polygons. Neuroreport. 2014; 25(11): 848–853. PubMed Abstract \| Publisher Full Text\n\nTegethoff M, Raul JS, Jamey C, et al.: Dehydroepiandrosterone in nails of infants: a potential biomarker of intrauterine responses to maternal stress. Biol Psychol. 2011; 87(3): 414–420. PubMed Abstract \| Publisher Full Text\n\nTegethoff M, Greene N, Olsen J, et al.: Inhaled glucocorticoids during pregnancy and offspring pediatric diseases: a national cohort study. Am J Respir Crit Care Med. 2012; 185(5): 557–563. PubMed Abstract \| Publisher Full Text\n\nTegethoff M, Knierzinger N, Meyer AH, et al.: Cortisol awakening response in infants during the first six postnatal months and its relation to birth outcome. Psychoneuroendocrinology. 2013; 38(5): 629–637. PubMed Abstract \| Publisher Full Text\n\nTegethoff M, Pryce C, Meinlschmidt G: Effects of intrauterine exposure to synthetic glucocorticoids on fetal, newborn, and infant hypothalamic-pituitary-adrenal axis function in humans: a systematic review. Endocr Rev. 2009; 30(7): 753–789. PubMed Abstract \| Publisher Full Text\n\nBen Khelil M, Tegethoff M, Meinlschmidt G, et al.: Simultaneous measurement of endogenous cortisol, cortisone, dehydroepiandrosterone, and dehydroepiandrosterone sulfate in nails by use of UPLC-MS-MS. Anal Bioanal Chem. 2011; 401(4): 1153–1162. PubMed Abstract \| Publisher Full Text\n\nJansen-Osmann P, Wiedenbauer G, Heil M: Spatial cognition and motor development: a study of children with spina bifida. Percept Mot Skills. 2008; 106(2): 436–446. PubMed Abstract \| Publisher Full Text\n\nGraca G, Goodfellow BJ, Barros AS, et al.: UPLC-MS metabolic profiling of second trimester amniotic fluid and maternal urine and comparison with NMR spectral profiling for the identification of pregnancy disorder biomarkers. Mol Biosyst. 2012; 8(4): 1243–1254. PubMed Abstract \| Publisher Full Text\n\nHill M, Parizek A, Kancheva R, et al.: Steroid metabolome in plasma from the umbilical artery, umbilical vein, maternal cubital vein and in amniotic fluid in normal and preterm labor. J Steroid Biochem Mol Biol. 2010; 121(3–5): 594–610. PubMed Abstract \| Publisher Full Text\n\nWudy SA, Dörr HG, Solleder C, et al.: Profiling steroid hormones in amniotic fluid of midpregnancy by routine stable isotope dilution/gas chromatography-mass spectrometry: reference values and concentrations in fetuses at risk for 21-hydroxylase deficiency. J Clin Endocrinol Metab. 1999; 84(8): 2724–2728. PubMed Abstract \| Publisher Full Text\n\nFahlbusch FB, Heussner K, Schmid M, et al.: Measurement of amniotic fluid steroids of midgestation via LC-MS/MS. J Steroid Biochem Mol Biol. 2015; 152: 155–160. PubMed Abstract \| Publisher Full Text\n\nBaron-Cohen S, Auyeung B, Nørgaard-Pedersen B, et al.: Elevated fetal steroidogenic activity in autism. Mol Psychiatry. 2015; 20(3): 369–376. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOttolenghi C, Abermil N, Lescoat A, et al.: Gestational age-related reference values for amniotic fluid organic acids. Prenat Diagn. 2010; 30(1): 43–48. PubMed Abstract \| Publisher Full Text\n\nChevy F, Humbert L, Wolf C: Sterol profiling of amniotic fluid: a routine method for the detection of distal cholesterol synthesis deficit. Prenat Diagn. 2005; 25(11): 1000–1006. PubMed Abstract \| Publisher Full Text\n\nBacigalupo G, Saling EZ, Dudenhausen JW: Unconjugated and total estriol in human amniotic fluid--changes in the ratios between the two estriol levels with advancing gestational age. J Perinat Med. 1979; 7(5): 262–269. PubMed Abstract \| Publisher Full Text\n\nFreeman R, Lev-Gur M, Koslowe R, et al.: Maternal plasma and amniotic fluid levels of estradiol, estrone, progesterone, and prolactin in early pregnancy. Obstet Gynecol. 1984; 63(4): 507–510. PubMed Abstract\n\nKushnir M, Blamires T, Rockwood A, et al.: Liquid chromatography-tandem mass spectrometry assay for androstenedione, dehydroepiandrosterone, and testosterone with pediatric and adult reference intervals. Clin Chem. 2010; 56(7): 1138–1147. PubMed Abstract \| Publisher Full Text\n\nSoldin O, Sharma H, Husted L, et al.: Pediatric reference intervals for aldosterone, 17alpha-hydroxyprogesterone, dehydroepiandrosterone, testosterone and 25-hydroxy vitamin D3 using tandem mass spectrometry. Clin Biochem. 2009; 42(9): 823–827. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "46372", "date": "23 Apr 2019", "name": "Brian G Keevil", "expertise": [ "Reviewer Expertise LC-MS/MS method development" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors describe the validation of an LC-MS/MS method for quantification of several steroids in amniotic fluid. The validation is not thorough enough to permit publication and in my view would benefit from extra work:\nPreparation of calibrators is sketchy, it appears that these were not made in protein based buffers. In my experience protein is needed in the buffer to stabilise the steroids and stop losses to plastic and glass surfaces.\n\nThe LLOQ should be decided on a combination of precision (<20% CV) and bias (<20% from target values), the CV was similar for both LOD and LLOQ.\n\nAnalyte recovery should have been performed in amniotic fluid.\n\nPrecision was evaluated in three solutions, how were these solutions made? They should be a similar matrix to amniotic fluid, possibly protein based but not methanol solutions.\n\nHow stable were the samples in amniotic fluid over hours and days?\n\nThere is no attempt to investigate matrix effects.\n\nOnly one internal standard is used with a different retention time to most of the steroids measured.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? No\n\nAre sufficient details provided to allow replication of the method development and its use by others? No\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? No", "responses": [] } ]	1	https://f1000research.com/articles/7-1736
https://f1000research.com/articles/7-1734/v1	01 Nov 18	{ "type": "Software Tool Article", "title": "ranacapa: An R package and Shiny web app to explore environmental DNA data with exploratory statistics and interactive visualizations", "authors": [ "Gaurav S. Kandlikar", "Zachary J. Gold", "Madeline C. Cowen", "Rachel S. Meyer", "Amanda C. Freise", "Nathan J.B. Kraft", "Jordan Moberg-Parker", "Joshua Sprague", "David J. Kushner", "Emily E. Curd", "Zachary J. Gold", "Madeline C. Cowen", "Rachel S. Meyer", "Amanda C. Freise", "Nathan J.B. Kraft", "Jordan Moberg-Parker", "Joshua Sprague", "David J. Kushner", "Emily E. Curd" ], "abstract": "Environmental DNA (eDNA) metabarcoding is becoming a core tool in ecology and conservation biology, and is being used in a growing number of education, biodiversity monitoring, and public outreach programs in which professional research scientists engage community partners in primary research. Results from eDNA analyses can engage and educate natural resource managers, students, community scientists, and naturalists, but without significant training in bioinformatics, it can be difficult for this diverse audience to interact with eDNA results. Here we present the R package ranacapa, at the core of which is a Shiny web app that helps perform exploratory biodiversity analyses and visualizations of eDNA results. The app requires a taxonomy-by-sample matrix and a simple metadata file with descriptive information about each sample. The app enables users to explore the data with interactive figures and presents results from simple community ecology analyses. We demonstrate the value of ranacapa to two groups of community partners engaging with eDNA metabarcoding results.", "keywords": [ "environmental DNA", "data visualization", "citizen science", "community science", "shiny", "metabarcoding", "education", "community ecology" ], "content": "Introduction\n\nThe targeted amplification and sequencing of DNA that living organisms shed into their physical environment, termed “environmental DNA (eDNA) metabarcoding,” is revolutionizing microbiology, ecology, and conservation research (Deiner et al., 2017; Taberlet et al. 2012). Sequencing of eDNA extracted from field-collected soil, water, or sediment samples can yield insight into a range of questions, from profiling the composition of ancient plant and animal communities (Pedersen et al., 2015), to monitoring populations of rare or endangered species (Balasingham et al., 2018). As the cost of eDNA metabarcoding declines and sample collection techniques become more streamlined (e.g. Thomas et al. (2018)), professional research scientists are increasingly using eDNA metabarcoding as a platform to engage a diversity of community partners, including natural resource managers, undergraduate students, and citizen scientists in primary research. However, developing robust and impactful community science programs that engage community partners in all steps of the research process remains a challenge.\n\neDNA metabarcoding-based projects work well for programs that partner researchers with community scientists because non-experts can be quickly trained to collect samples in the field, and because eDNA metabarcoding is an exciting framework for research pertinent to disciplines such as medicine, agriculture, ecology, and geography (Deiner et al., 2017). Community partners in such programs can have heterogeneous backgrounds, ranging from curious members of the public for whom collecting samples in the field is their first scientific research experience (e.g. University of California’s CALeDNA program), to professional natural resource managers who regularly collaborate with research scientists (e.g. Center for Ocean Solutions’ eDNA project). A key ingredient to promote sustained success of such programs is that community partners should be able to engage across multiple stages of the research project, not only in sample collection (European Citizen Science Association, 2015; Pandya, 2012). This can be a challenge for community science programs because although it is relatively easy to train community partners to collect eDNA samples, it is far more challenging to train them to independently visualize and analyze results from these studies. Indeed, learning the bioinformatic tools necessary for managing the large, multidimensional datasets generated in these studies can be difficult for professional researchers (Carey & Papin, 2018), let alone for the non-technical audience of some community science programs.\n\nTo address this challenge, we created the R package “ranacapa”, at the core of which is a Shiny web app that can be used to visualize results from eDNA sequencing studies and perform simple community ecology analyses. ranacapa complements existing visualization platforms (e.g. Phinch (Bik & Phinch Interactive, 2014), Phyloseq-Shiny (McMurdie & Holmes, 2015), QIIME2 Viewer), because in addition to interactive visualizations, ranacapa includes brief explanations of several core analyses used in eDNA studies and includes links to additional educational resources. ranacapa works with community matrices generated via QIIME (Caporaso et al., 2010) or the Anacapa sequence analysis pipeline, the latter being used extensively by the CALeDNA program.\n\nHere, we describe the package and how it is used by two community science partnerships based at the University of California, Los Angeles (UCLA): first, a collaboration between eDNA researchers and resource managers at the National Park Service, and second, a partnership between community ecology researchers and an undergraduate microbiology course at UCLA. As we show in the Use cases, empowering community partners to interact with the data and perform simple but insightful community ecology analyses can help make these collaborations more enriching and valuable to both parties.\n\n\nImplementation\n\nAt the core of ranacapa is a Shiny web app (Chang et al., 2018), which is available at http://gauravsk.shinyapps.io/ranacapa or with ranacapa::runRanacapa(). The package also includes two categories of helper functions (Table 1) that transform user-uploaded taxonomy and metadata tables into R objects that can be visualized and analyzed using the Phyloseq (McMurdie & Holmes, 2013) and Vegan (Oksanen et al., 2018) packages. ranacapa is available for installation from Github or CRAN:\n\n\n\n1 adopted from https://github.com/pmartinezarbizu/pairwiseAdonis (GPL-3 License)\n\n2 adopted from https://github.com/mahendra-mariadassou/phyloseq-extended (GPL-3 License)\n\nThe ranacapa Shiny app allows users to interact with eDNA results through statistical summaries and interactive plots, displayed in the following tabs:\n\n• Sequencing depth: Introduces the potential for variation in sequencing depth among samples and explains the basic logic behind rarefying samples in metagenomics studies (Figure 1). Users can rarefy the dataset to a sampling depth, or can proceed through the rest of the app without rarefying samples. The documentation acknowledges recent disagreement regarding the value of rarefying in metabarcoding and eDNA sequencing studies (McMurdie & Holmes, 2014).\n\nThe online version of this figure is interactive.\n\n• Taxonomy heatmap: Shows the taxon-by-sample matrix as an interactive heatmap made using heatmaply::heatmaply() (Galili et al., 2018), where the color of each cell represents the number of times a given taxon was sequenced in a sample (Figure 2). Users can filter the taxon list by selecting or deselecting specific taxa.\n\nTaxonomy is shown at the Order level in this figure; in the app, users can choose the taxonomic level to show in the heatmap. Users can also select or deselect individual taxa to be shown in the heatmap. The online version of this figure is interactive.\n\n• Taxonomy barplot: Shows the taxonomy-by-sample matrix as an interactive barplot (Figure 3).\n\nTaxonomy is shown at the Order level in this figure; in the app, users can choose the taxonomic level to show in the barplot. The online version of this figure is interactive.\n\n• Alpha diversity plots: Introduces the concept of alpha diversity as the local diversity measured in a single habitat or sample. Users can plot alpha diversity as observed taxon richness or as Shannon diversity per sample, or can group samples according to a variable in the metadata file (Figure 4).\n\nUsers can select the X-axis variable using a dropdown menu in the app. The online version of this figure is interactive.\n\n• Alpha diversity statistics: Allows users to choose a variable from the metadata, and generates an alpha diversity ANOVA table according to the user-selected variable. The tab also shows the output from a post-hoc Tukey test.\n\n• Beta diversity plots: Introduces the concept of beta diversity as the turnover in species composition across habitats (or samples). The tab includes an ordination plot generated by phyloseq::plot_ordination(), which in turn uses an ordination object made with phyloseq::ordinate(., method = \"PCoA\"). Points on the PCoA plot are colored according to a user-selected metadata variable (Figure 5).\n\nUsers can select the grouping variable with a dropdown menu in the app. The online version of this figure is interactive.\n\nThe beta diversity plots tab also includes a dendrogram that groups sites based on Ward’s cluster analysis (stats::hclust(distance_object, method = \"ward.d2\")), where distance_object is made using phyloseq::distance(). For both figures, users can toggle between using Jaccard and Bray-Curtis dissimilarity.\n\n• Beta diversity statistics: Shows results from two statistical tests of species turnover across sites. The first test is a multivariate ANOVA of taxon turnover across sites, implemented with vegan::adonis(). The second statistical test, which is implemented with vegan::betadisper(), is of heterogeneity of variances among samples. This test compares the degree of sample-to-sample variation within habitats (or within other user-selected groups).\n\n\nOperation\n\nranacapa depends on Bioconductor v 3.7, which in turn relies on R v 3.5.0. The Shiny app has been tested on Chrome and Firefox on Windows, Mac-OSX, and Ubuntu.\n\nThe ranacapa Shiny app requires two input files. The first requirement is a taxon-by-sample matrix, uploaded either as a rich, dense .biom table, or as a tab-separated .txt file. Qiime2-generated .qza files generated by QIIME2 are not immediately suitable for ranacapa, as they do not contain full taxonomy information. If the site-by-species matrix is uploaded as a .txt file, the file should match the specifications of the output files from the Anacapa eDNA sequence analysis pipeline. In Anacapa output, each row represents a taxonomic identification, and each column (save one) represents the number of times that taxon appears in each sequenced sample. One column, named sum.taxonomy must contain the taxonomic identification, with taxonomic rank separated by a semicolon, e.g. “Chordata;Actinopteri;Chaetodontiformes;Chaetodontidae;Chaetodon;Chaetodon reticulatus .” A valid input file is structured as follows:\n\n\n\nThe second requirement is a tab-separated .txt file that contains sample metadata. The first column in the metadata file should match the sample names in the taxonomy table; the remaining columns contain sample information for each of the samples in the taxon-by-site matrix. The metadata should contain categorical variables with two or more categories per variable. A valid metadata file for the taxonomy table above is structured as follows:\n\n\n\nThe ranacapa function validate_input_files() verifies that both the taxonomy table and the metadata files match structural requirements, which are documented in the function help files.\n\n\nUse cases\n\nWe expect that researchers with expertise in bioinformatics will use the sequence analysis pipeline of their choice to assign taxonomy to eDNA datasets, and generate clean taxonomy and metadata files that can be visualized in ranacapa. Researchers can share these files with their partners, and emphasize the analyses or visualizations most appropriate to their use case. We now show how ranacapa can facilitate authentic communication between researchers and community partners in two settings.\n\nA team of UCLA researchers partnered with resource managers at the Channel Islands National Park Service to assess the potential for eDNA as a biodiversity monitoring tool to supplement time-intensive visual biodiversity surveys in the Southern California Channel Islands (Deiner et al., 2017; Lessios, 1996; Usseglio, 2015). For this partnership, resource managers collected and filtered 30 unique one-liter water samples for eDNA analysis at permanent monitoring sites inside and adjacent to protected areas, and research scientists at UCLA performed eDNA sequencing of the mitochondrial 12S (Miya et al., 2015) and CO1 (Leray et al., 2013) genes, targeting bony fishes, elasmobranches, and invertebrate taxa. The researchers processed sequences and assigned taxonomy using the Anacapa pipeline, and shared results with the resource managers using the ranacapa Shiny app.\n\nThe taxonomy heatmap of species detected using the 12S and CO1 metabarcodes (Figure 2) was the most valuable visualization to this collaboration, because it allowed the resource managers to filter the large observed species list down to a particular set of key taxa that they regularly monitor. The heatmap showed that this pilot study detected 36 of the 70 key metazoans at the species level, and the remaining 34 at the genus, family, or order level. This indicates that eDNA-based studies can provide critical information for ongoing management efforts and provide new insights into the spatial and temporal distributions of these species. The value of ranacapa in this scenario was to quickly sort through long species lists generated by eDNA sequencing to highlight the strengths and weaknesses in using eDNA to monitor diversity in the Channel Islands. The data from this study are packaged as the demo dataset for the ranacapa Shiny app and are available online (Kandlikar et al., 2018a).\n\nA team of community ecology and environmental DNA researchers in the CALeDNA program collaborated with instructors of a research-based environmental microbiology course at UCLA (Shapiro et al., 2015), in which students used eDNA metabarcoding to study the impact of a local wildfire on the plant and soil microbial community. The goal of this twenty-week course was to provide undergraduate students an authentic experience in basic microbiology and microbial community ecology research. Over the first ten weeks, eDNA researchers on the instructional team sequenced the ITS2 (Gu et al., 2013) and 16S SSU RNA (Caporaso et al., 2012) metabarcoding regions from student-collected soil samples and used the Anacapa pipeline to generate taxon-by-sample tables.\n\nThe course instructors used the ranacapa Shiny app to introduce students to the structure of eDNA sequencing results. The students were encouraged to explore data and perform the statistical analyses most pertinent to the hypotheses they had formed at the beginning of the course. A key benefit of using ranacapa was that despite having no prior bioinformatics experience, students could begin exploring the biodiversity in their samples in a matter of minutes by using the online instance of the Shiny app. This allowed the instructors to focus classroom time on biological questions rather than on troubleshooting bioinformatics problems, as had been the case in previous sessions of the course. The course instructors noted that visualizing eDNA data in ranacapa helped students understand the relationships between taxon-by-site matrices and the various metadata they had collected in the field. By significantly reducing the time and difficulty in visualizing basic biodiversity patterns, ranacapa helped students develop and pursue more sophisticated analyses during the remainder of the course, using tools such as STAMP (Parks et al., 2014) and PICRUSt (Langille et al., 2013). The taxonomy tables and metadata files used in this course are available online (Kandlikar et al., 2018b).\n\n\nSummary and future directions\n\nMetabarcode sequencing of environmental DNA is becoming a key tool in a wide variety of ecological studies, and results from these studies are of interest to a broad audience. Our R package and Shiny app ranacapa helps users conduct exploratory analyses and visualizations on eDNA datasets, and is a step toward more fully engaging participants in all phases of eDNA sequencing-based community science projects.\n\nWe propose three avenues for future work with ranacapa. First, we plan to use ranacapa as the primary tool to present eDNA results from hundreds of samples sequenced by the CALeDNA community science program. Second, ranacapa is being integrated into the upcoming undergraduate curriculum module “Pipeline for Undergraduate Microbiome Analysis”, which will be an open-source, comprehensive suite of analysis and data visualization tools for undergraduate researchers. Finally, in the long-term, we believe there is great promise in linking ranacapa with packages that connect with APIs of online biodiversity databases (e.g. Taxize (Chamberlain & Szöcs, 2013), rinat (Barve & Hart, 2017)). This will help users explore a much wider range of biodiversity questions, for example, by programmatically asking whether their samples include invasive species that are absent from other nearby sites. In sum, tools like ranacapa that allow non-technical audiences to easily interact with results from eDNA sequencing studies have great potential to engage community partners with a wide range of backgrounds and interests in primary research.\n\n\nSoftware availability\n\n• A Shiny app, including a dataset generated for demonstrations, is available at https://gauravsk.shinyapps.io/ranacapa\n\n• Source code is available from GitHub: https://github.com/gauravsk/ranacapa\n\n• Archived source code at time of publication: http://dx.doi.org/10.5281/zenodo.1464285 (Kandlikar & Cowen, 2018)\n\n• Software license (GPL-3)\n\n\nData availability\n\nDatasets used for the Use cases are available from Figshare:\n\nDataset 1: Taxon table and metadata file for Channel Islands eDNA samples (mitochondrial 12S and CO1 metabarcodes sequenced) https://doi.org/10.6084/m9.figshare.7199477.v1 (Kandlikar et al., 2018a)\n\nDataset 2: Taxon table and metadata file for Santa Monica Mountains eDNA samples (16S and plant-ITS metabarcodes sequenced) https://doi.org/10.6084/m9.figshare.7199510.v1 (Kandlikar et al., 2018b)\n\nBoth datasets are available under a CC-BY 4.0 license", "appendix": "Grant information\n\nGSK and ZJG were supported by the US-NSF Graduate Research Fellowship [DEG No. 1650604]. NJBK was supported the National Science Foundation [DEB-1644641]. EEC, RSM, and the CALeDNA program are supported by the University of California President’s Research Catalyst Award [CA-16-376437].\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nWe thank Sabrina Shirazi, Rachel Turba, Chris Dao, and Keith Mitchell for providing feedback on developmental versions of this package. We also thank Mahendra Mariadassau and Pedro Martinez Arbizu for making the phyloseq-extended and pairwiseAdonis packages openly available with a GPL-3 License.\n\n\nReferences\n\nBalasingham KD, Walter RP, Mandrak NE, et al.: Environmental DNA detection of rare and invasive fish species in two Great Lakes tributaries. Mol Ecol. 2018; 27(1): 112–127. PubMed Abstract \| Publisher Full Text\n\nBarve V, Hart E: Rinat: Access iNaturalist data through apis. 2017. Reference Source\n\nBik, Phinch Interactive: Phinch: An interactive, exploratory data visualization framework for –Omic datasets. bioRXiv. 2014. Publisher Full Text\n\nCaporaso JG, Kuczynski J, Stombaugh J, et al.: QIIME allows analysis of high-throughput community sequencing data. Nat Methods. 2010; 7(5): 335–336. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCaporaso JG, Lauber CL, Walters WA, et al.: Ultra-high-throughput microbial community analysis on the Illumina HiSeq and MiSeq platforms. ISME J. 2012; 6(8): 1621–1624. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCarey MA, Papin JA: Ten simple rules for biologists learning to program. PLoS Comput Biol. 2018; 14(1): e1005871. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChamberlain SA, Szöcs E: taxize: taxonomic search and retrieval in R [version 1; referees: 3 approved]. F1000Res. 2013; 2: 191. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChang W, Cheng J, Allaire J, et al.: Shiny: Web application framework for r. 2018. Reference Source\n\nDeiner K, Bik HM, Mächler E, et al.: Environmental DNA metabarcoding: Transforming how we survey animal and plant communities. Mol Ecol. 2017; 26(21): 5872–5895. PubMed Abstract \| Publisher Full Text\n\nEuropean Citizen Science Association: Ten principles of citizen science. 2015. Reference Source\n\nGalili T, O’Callaghan A, Sidi J, et al.: heatmaply: an R package for creating interactive cluster heatmaps for online publishing. Bioinformatics. 2018; 34(9): 1600–1602. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGu W, Song J, Cao Y, et al.: Application of the ITS2 Region for Barcoding Medicinal Plants of Selaginellaceae in Pteridophyta. PLoS One. 2013; 8(6): e67818. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKandlikar G, Cowen M: gauravsk/ranacapa: First release of ranacapa (Version v1.0.0). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1464285\n\nKandlikar GS, Gold ZJ, Cowen MC, et al.: Taxon table and metadata file for Channel Islands eDNA samples (mitochondrial 12S and CO1 metabarcodes sequenced). Figshare. 2018a. http://www.doi.org/10.6084/m9.figshare.7199477.v1\n\nKandlikar GS, Gold ZJ, Cowen MC, et al.: Taxon table and metadata file for Santa Monica Mountains eDNA samples (16S and plant-ITS metabarcodes sequenced). Figshare. 2018b. http://www.doi.org/10.6084/m9.figshare.7199510.v1\n\nLangille MG, Zaneveld J, Caporaso JG, et al.: Predictive functional profiling of microbial communities using 16S rRNA marker gene sequences. Nat Biotechnol. 2013; 31(9): 814–821. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLeray M, Yang JY, Meyer CP, et al.: A new versatile primer set targeting a short fragment of the mitochondrial COI region for metabarcoding metazoan diversity: application for characterizing coral reef fish gut contents. Front Zool. 2013; 10: 34. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLessios HA: METHODS for quantifying abundance of marine organisms. In: Methods and techniques of underwater research. (eds. Lang, M. & Baldwin, C.). American Academy of Underwater Sciences (AAUS), 1996; 149–157. Reference Source\n\nMcMurdie PJ, Holmes S: phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS One. 2013; 8(4): e61217. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMcMurdie PJ, Holmes S: Waste not, want not: why rarefying microbiome data is inadmissible. PLoS Comput Biol. 2014; 10(4): e1003531. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMcMurdie PJ, Holmes S: Shiny-phyloseq: Web application for interactive microbiome analysis with provenance tracking. Bioinformatics. 2015; 31(2): 282–283. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMiya M, Sato Y, Fukunaga T, et al.: MiFish, a set of universal PCR primers for metabarcoding environmental DNA from fishes: detection of more than 230 subtropical marine species. R Soc Open Sci. 2015; 2(7): 150088. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOksanen J, Blanchet FG, Friendly M, et al.: Vegan: Community ecology package. 2018. Reference Source\n\nPandya RE: A framework for engaging diverse communities in citizen science in the US. Front Ecol Environ. 2012; 10(6): 314–317. Publisher Full Text\n\nParks DH, Tyson GW, Hugenholtz P, et al.: STAMP: statistical analysis of taxonomic and functional profiles. Bioinformatics. 2014; 30(21): 3123–3124. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPedersen MW, Overballe-Petersen S, Ermini L, et al.: Ancient and modern environmental DNA. Philos Trans R Soc Lond B Biol Sci. 2015; 370(1660): 20130383. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShapiro C, Moberg-Parker J, Toma S, et al.: Comparing the Impact of Course-Based and Apprentice-Based Research Experiences in a Life Science Laboratory Curriculum. J Microbiol Biol Educ. 2015; 16(2): 186–197. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTaberlet P, Coissac E, Hajibabaei M, et al.: Environmental DNA. Mol Ecol. 2012; 21(8): 1789–1793. PubMed Abstract \| Publisher Full Text\n\nThomas AC, Howard J, Nguyen PL, et al.: ANDe: A fully integrated environmental DNA sampling system. Methods Ecol Evol. 2018; 9(6): 1379–1385. Publisher Full Text\n\nUsseglio P: Quantifying reef fishes: Bias in observational approaches. In: Ecology of fishes on coral reefs.(ed. Mora, C.). Cambridge University Press, 2015; 270–273. Publisher Full Text" }	[ { "id": "40402", "date": "03 Dec 2018", "name": "Tristan Cordier", "expertise": [ "Reviewer Expertise Molecular ecology", "environmental genomics", "bioinformatics" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript “ranacapa: An R package and Shiny web app to explore environmental DNA data with exploratory statistics and interactive visualizations” is describing an R package to visualize metabarcoding data and perform summary statistics. I think that such application will be very useful for quick results sharing within any project that involves stakeholders with various background. Too many projects still rely solely on a printed (or pdf) report with key chosen figures, orienting the readers towards the main messages. I do believe that such application will provide a nice complement to a project report for people that want to dive into the results, or explore the robustness of the conclusion that are highlighted in a report. I also agree with the authors that such tools are particularly interesting for sharing results with non-expert enthusiasts and students. I appreciate the accompanying text to present the specificity of composition (count) data and the possible flaws when analyzing such data. The website “Gusta Me” (https://mb3is.megx.net/gustame) could maybe be mentioned as an external resource for guiding further the users for the interpretation of results, or to guide any additional analysis. I wonder how such application is intended to be deployed. Do the authors plan to let research scientists deploy the app within their own computational resources or do they plan to make the application accessible (https://gauravsk.shinyapps.io/ranacapa/) opened to the community? I made few comments on the GUI that the authors may or may not follow. The most important point to address would be to allow users to write their PERMANOVA model.\n\nData import:\n\nMaybe add the expected format, with a toy example, indicating the separator character (tab or comma or semi-comma?). This is mentioned in the GitHub page of the project (not the separator though), but I think this would be nice to have it here also. My guess is that many users would not even get to the GitHub page before using the app.\n\nSequencing depth:\n\nWould it be possible to add a “free” input field? As it is now, it is not easy to get a precise value. The slider does not allow this kind of precision, it increases or decreases by 1000 but on the demo example, it is stuck to 10006. Would it be possible to include normalization as an option? At least the relative abundance for a start, because as you mention on the accompanying text, we are far from having reach a consensus on that matter. What means the first field “select the variable”, do you mean to rarefy not samples but something else? I don’t understand.\n\nTaxonomy barplot:\n\nI personally don’t like taxonomic barplots, but I recognize that they can be useful to easily illustrate some obvious differences (or similarities) between experimental treatments to an audience unfamiliar with metabarcoding data. The lowest taxonomic ranks return a (way) too many categories. In the presented example. Even the highest taxonomic rank (phylum), it yields too many categories to easily visualize what is actually in the samples, because colors are too close to each other. There is no solution, I just don’t like these plots. But it is nice to have this option in the app anyway.\n\nTaxonomy heatmap:\nIs there an option to scale differently the X and Y axis? In the example dataset, it is actually a similar problem as for taxonomy barplot, lowest taxonomic rank gives too many rows on the plots. A way to better see is to zoom in, but then the horizontal axis is quite messed up. Maybe it would be nice to have input fields to manage those axes and font size.\n\nAlpha-diversity plots:\nI guess the data behind is the rarefied version of the dataset, but this should be clear. Maybe the box can be there, but in grey to show that it would not make sense to use the raw dataset?\n\nAlpha-diversity stats: I think this is fine.\n\nBeta-diversity plots:\nThe users should be able to refine the ordination plots, its size, the size of each axis. Why the ordination plots configuration (and the cluster analysis) is changing when we select different variables to group the dot shapes and colors?\n\nBeta-diversity analyses:\nWould it be possible to let the user write his own model? As for now, it is not possible to have an interaction term, or a nested model. To be able to select the sample as variable does not make sense.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Partly", "responses": [] }, { "id": "40129", "date": "14 Dec 2018", "name": "Holly M. Bik", "expertise": [ "Reviewer Expertise metabarcoding", "genomics", "data visualization" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article presents a new R software package and shiny app called \"ranacapa\", aimed at producing rapid visualization and exploration of eDNA metabarcoding datasets. Overall I found this software tool to be exciting and user friendly, and the outlined use cases provided good examples of the need for such tools amongst end users including researchers and undergraduate students. The R shiny app was easy to explore using the demo dataset, and the plots were responsive and visually useful.\n\nWhen trying to install ranacapa in RStudio (via install.packages), I am consistently getting an error message stating that \"package ‘ranacapa’ is not available (for R version 3.3.2)\". Furthermore, ranacapa does not appear to be listed as a package currently available in CRAN (https://cran.r-project.org/web/packages/available_packages_by_name.html#available-packages-R). Further investigation suggests that this package requires R 3.5.0 or higher to run, but this is not obvious for end users with older versions of R installed. Is it possible to add back compatibility to older versions of R? Especially for end users using HPC facilities who may be unable to update the base R version on their cluster.\n\"Taxonomy-by-sample matrix\" - since ranacapa uses a non-standard input (essentially an OTU table summarized by taxonomy), it would be useful to provide specific instructions and/or a conversion script where users can take their QIIME BIOM files or tab-delimited OTU tables and convert them into a format that is suitable for uploading into ranacapa. Users with minimal technical expertise may struggle to convert their data in the correct way. Especially useful would be R scripts or package commands (e.g. phyloseq) that will output taxonomy-by-sample-matrixes in the correct format.\nThe R shiny web app does not provide much information regarding file formatting (users must read the F1000 article for specific information). The only documentation online states that the imported files should be in .biom or .txt formats. What format(s) are supported for the header row (commented out vs. plain text)? Are there any limitations on sparse vs. dense or JSON vs. HDF5 BIOM files (e.g. from QIIME 1 or QIIME2)? Detailed information and documentation should be provided on the website itself in addition to the F1000 article.\nFor the metdatata file - can users directly import QIIME-formatted mapping files (e.g. with a header row beginning with #SampleID that has been validated via Keemi)? Or does this file need to be be edited further for ranacapa?\nFor the demo file on the web server - what study did this file come from? It would be useful to have some contextual information about the file, eDNA markers used, questions/hypotheses, and study design so that end users can explore the demo dataset in a more thoughtful way.\n\nSome components of the R Shiny web app do not always seem to function correctly over a slow internet connection - it seems to freeze up and require re-loading, and/or users are presented with an error message. This error message sometimes goes away and the plots pop up as expected after a few minutes. Better error dialogs or \"please wait\" message boxes would be much less frustrating for the end user to figure out what is happening.\nSequencing Depth - the rarefaction depth slider bar is straightforward and should be familiar to most users, however, what impact do multiple rarefactions have on the displayed graphs? Are the rarefaction curves averaged somehow before being displayed? Further information on this parameter would be useful to include in the explanatory text.\nTaxonomy heatmap - when choosing a lower taxonomic level, the taxonomy strings are overlapping and unreadable. Zooming in cuts off some of the categories on the x-axis. It would be better to be able to expand the heatmap along the y-axis (stretch it out) so you can read the taxonomy labels while still being able to view all of the heatmap information.\nTaxonomy heatmap - when selecting a taxon to visualize, users might want to see more detail at lower taxonomic levels, however, selecting a taxon to visualize and then choosing a different taxonomic rank currently removes this filter choice and reverts back to showing all taxa at that level.\nAlpha Diversity Plots - what is the difference between Observed vs. Shannon diversity visualizations? The explanatory text only partially explains the difference between these two options, and could be clarified to explicitly state the important differences in these two calculations. Furthermore, what do high vs. low numerical values mean for each of the two indices, in biological terms (e.g. lower number of species)? These kind of diversity statistics may be familiar for ecologists and microbiome researchers, but are likely to be confusing for citizen scientists, students, or scientists outside of these fields.\nAlpha Diversity Stats - Unless you have a strong statistical background, it is impossible to interpret the numbers and outputs listed in this tab in the ranacapa shiny app. More explanation is needed to guide users in what they should be looking for - what columns are most important in the ANOVA table to see if samples are in fact unequal in terms of their diversity? In the post-hoc tests, should users be looking for P values <0.05 in the last column?\n\nBeta Diversity - the Jaccard and Bray-Curtis indexes could be explained more plainly. Jaccard is the simplest measure of shared taxa/OTUs, while Bray-Curtis is an index focused on the most abundant taxa/OTUs in the dataset. Also some explanation of the PCoA axes (e.g. that they explain variation but do not relate to metadata) would be useful. Adding the Canberra index might also be useful here, as this is another useful index which ignores the most abundant taxa and looks at differences across rare species.\nBeta Diversity Stats - same comment as Alpha Diversity stats (see above). Needs more context and explanation to be accessible to end users without a statistical background.\nUse Case 1 - how was the UCLA eDNA dataset filtered down to the set of key taxa that resource managers were looking for? Was this done via R command line, or via the R shiny app interface? Custom filtering is likely to be a common use case, and it would be valuable to add in more specific information regarding how this type of taxon filtering can be accomplished for any dataset.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [] }, { "id": "40128", "date": "19 Dec 2018", "name": "Niklaus J. Grünwald", "expertise": [ "Reviewer Expertise R programing", "biology", "bioinformatics" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors present the R package rancapa that implements an interactive shiny web app available at https://gauravsk.shinyapps.io/ranacapa for exploration of environmental DNA characterized by amplicon sequencing. This app facilitates many of the standard tests and plots done for metabarcoding data. The app requires two dataset to run, namely a taxonomy-by-sample matrix file and a simple metadata file with descriptive information on the actual environmental samples. This package is particularly useful to anybody interested in exploring biome diversity but lacking extensive computational and bioinformatics skills. We appreciate the focus on open-source code and accessibility, in particular testing the app with Ubuntu and using the color-blind friendly viridis color palette. There are some minor changes in the style of plots and wording we would like to see, but these are a matter of opinion and do not affect the validity of the results. The authors admirably point out flaws and potential pitfalls in the analysis that often go unmentioned (e.g. the drawbacks of rarefaction). Although there is always room for improvement, we feel this app is quite useful as is and it is described well in the associated article. We suggest many improvements that could be made, but most of these are intended as constructive suggestions.\nSpecific comments:\nkeywords: include “R package”\nIntroduction\nThe sentence “A key ingredient to promote sustained success of such programs is that community partners should be able to engage across multiple stages of the research project, not only in sample collection” is a bit awkward. Perhaps something like the following would be more clear: “Such programs are more likely to succeed if community partners are able to be involved in multiple stages of the research project, not only sample collection” End of paragraph 2: “let alone for the non-technical audience of some community science programs.” We suggest “audience” be replaced with “members”, “partners”, or “collaborators”, since they are involved in the project.\nImplementation\nParagraph 1, sentence 1: The command “ranacapa::runRanacapa()” does not work. Should it be “ranacapa::runRanacapaApp()”? Figure 4: It would be nice if the results of the Tukey’s HSD test be included on the graph using letter codes. Sequencing depth, sentence 2: “Users can rarefy the data set to a sampling depth, or can proceed through the rest of the app without rarefying samples.” It's nice to have the option, but if sequences are not rarefied, perhaps they should be converted to proportions? The taxon abundance plots (e.g. the heat map) is biased without some correction for sampling depth and proportions would avoid the loss of data associated with rarefaction. While converting to proportions will not remove sample-size bias for alpha and beta diversity analyses, it will remove the bias for taxon abundance analyses for the most part. The conversion to proportions could also be an option in just the taxon abundance analyses. Beta diversity plots, sentence 1: “Introduces the concept of beta diversity as the turnover in species composition across habitats (or samples).” The word “turnover” seems to imply a change over time, rather than a comparison between communities. Perhaps something like: “Introduces the concept of beta diversity as a way to compare community composition across habitats (or samples).” Beta diversity statistics, sentence 1: “Shows results from two statistical tests of species turnover across sites.” See comment above on “turnover”.\nSuggestions on the software\nExploring ranacapa at https://gauravsk.shinyapps.io/ranacapa shows the interactive tools. We suggest that screenshots from the actual shinyapp be presented as figures as they provide the context for exploring the data. For example under Figure 2, the legend states that users can also select or deselect individual taxa’ but the figure 2 in the manuscript does not show this. This is confusing. We also suggest giving explicit examples for executing the shiny app on a desktop maybe right after the installation instructions.\nData import\nThe proper format for input data is described in the article, but it would be nice if it was repeated in the “data import” section of the app as well. You could also include tips to convert common formats to the needed format here.\nSequencing depth\nSecond paragraph, first sentence: “In eDNA sequencing, such variation in how deeply a given sample is sequenced can happen for a variety for reasons-”. Should the “-” be a “:”? Third paragraph, first sentence: You use double quotes for “deeply” in the first paragraph, but single quotes for ‘rarefy’. Add proportions as an alternative to rarefaction instead of “none”? The rarefied samples plot is a good idea, but hard to read in its current state. Perhaps change the faceting settings so that there is one column of plots or have a option to choose the sample shown? Taxonomy bar pot: The number of similar colors make this hard to read without selecting individual taxa. We know this is a limitation of stacked barcharts, but it might make it more readable to sort the taxa by average abundance and then stagger the colors like so: For example, if there are 30 taxa, then the first 10 get colors 1, 4, 7, 10, 13, 16, 19, 22, 25, 28 (seq(1,30, 30 / 10)). Also have the most abundant taxa be the lowest on the stacked bars. That way adjacent colors can be differentiated and the colors of rare taxa won't be confused with the colors of the abundant taxa.\nTaxonomy heatmap\nHow about faceting by sample metadata or clustering samples by similarity? You could use the results of the Ward’s Hierarchical Clustering like you do in the beta diversity section. Can you make the height of the plot depend on the number of samples plotted? Also, making the “Select taxa to visualize” box taller would be nice.\nAlpha diversity plots\nYou use the term “species richness” without explicitly defining it, although you implicitly define it. It might be a good idea to point out that “observed” is the same as “species richness”. Plotting the results of the Tukey’s HSD test on this plot would be great. Actually this tab could probably be combined with the “alpha diversity stats” tab. The color and the different point shapes are not needed, but that’s just a question of style.\nAlpha diversity stats\nYou say “ANOVAs only work when each group is represented by a few samples”, which is great to point out. How about just not allowing for the ANOVA and Tukey’s HSD if there are not multiple samples per treatment? How about not showing the Tukey’s test result if the ANOVA p-value is more than 0.1? You do say “the following results are meaningful only if the ANOVA table above suggests that sites are in fact unequal in terms of their diversity”, but it's very tempting to look at the results of the Tukey’s HSD test anyway. Alternatively, you could add a warning above the Tukey’s HSD results if the ANOVA p-value is more than 0.1. How about using the DT package to make the Tukey’s HSD results table sortable, like you do for other tables? The first thing most people will want to do is sort by p-value.\nBeta diversity plots\nYou say “samples that are distant in this plot have very different species lists”. That is not exactly true. Samples that are more distant from each other are more different from each other, but they all might be quite similar in terms of the species present; The PCoA will always show differences, no matter how small. An explanation of how to interpret the percent of the variation explained by each axis would be nice. How about confidence interval ellipsis for the sample types? How about allowing exploration of other dimensions (maybe in a 3d plot as well)? In the clustering dendrogram, how about coloring sample names by sample type? Ideally, it would be the same colors used in the PCoA plot, so you would not need a second legend.\nBeta diversity stats\nHow about displaying the results of the Adonis analysis in a table instead of a printed data.frame, like how it is done in the “Homogeneity of Variances” analysis? Tables from the DT package would be nice, so we can sort by p-value.\nData export\nIt would be nice to provide an option to export the data in the taxmap format from the taxa package1. Zach Foster can help implement this.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1734
https://f1000research.com/articles/7-1730/v1	31 Oct 18	{ "type": "Method Article", "title": "Evaluation of broad-spectrum antiviral compounds against chikungunya infection using a phenotypic screening strategy", "authors": [ "Rafaela M. Bonotto", "Glaucia Souza-Almeida", "Soraya Jabur Badra", "Luiz Tadeu Figueiredo", "Carolina B. Moraes", "Lucio H. Freitas-Junior", "Rafaela M. Bonotto", "Glaucia Souza-Almeida", "Soraya Jabur Badra", "Luiz Tadeu Figueiredo", "Carolina B. Moraes" ], "abstract": "Chikungunya fever is an emerging disease and a significant public health problem in tropical countries. Recently reported outbreaks in Brazil in 2015 drew attention to the need to develop prevention and treatment options, as no antiviral chemotherapy or vaccines are currently available for this disease. Two strategies have been proved to accelerate the discovery of new anti-infectives: phenotypic screening and drug repurposing. Phenotypic screening can support the fast interrogation of compounds without the need for a pre-validated drug target, which is not available for the chikungunya virus (CHIKV) and has the additional advantage of facilitating the discovery of antiviral with novel mechanism of action. Drug repurposing can save time and resources in drug development by enabling secondary uses for drugs that are already approved for human treatment, thus precluding the need for several of the mandatory preclinical and clinical studies necessary for drug approval. A phenotypic screening assay was developed by infecting the human hepatoma Huh-7 cells with CHIKV 181/25 and quantifying infection through indirect immunofluorescence. The compound 6-azauridine was used as a positive control drug. The screening assay was validated by testing a commercial library of 1,280 compounds, including FDA-approved drugs, and used to screen a panel of broad-spectrum antiviral compounds for anti-CHIKV activity. A high content assay was set up in Huh-7 cells-infected with CHIKV. The maximum rate of infection peaked at 48 hours post-infection, after which the host cell number was greatly reduced due to a strong cytopathic effect. Assay robustness was confirmed with Z’-factor values >0.8 and high correlation coefficient between independent runs, demonstrating that the assay is reliable, consistent and reproducible. Among tested compounds, sofosbuvir, an anti-hepatitis C virus drug, exhibited good selectivity against CHIKV with an EC50 of 11 µM, suggesting it is a promising candidate for repurposing.", "keywords": [ "Chikungunya", "High content screening", "drug discovery", "antivirals" ], "content": "Introduction\n\nChikungunya virus (CHIKV) is an arthropod-borne virus that belongs to the Alphavirus genus of the Togaviridae family. Alphaviruses are positive-sense, single-stranded RNA viruses that can produce severe encephalitis, such as in the infections caused by Ross River virus (RRV), Western- (WEE), Eastern- (EEE) and Venezuelan-equine encephalitis (VEE) virus. Alphaviruses can also be arthritogenic, such as in the case of CHIKV, Mayaro virus (MAYV), and O’nyong’nyong virus (ONNV)1. CHIKV was responsible for several recent (re)emerging outbreaks in humans2–4. Nowadays, approximately one billion people around the globe, especially in the tropics, are estimated to live in risk areas of CHIKV outbreaks4,5. In the Americas, CHIKV was first detected in 2013, in St. Martin, an island in the Caribbean, and quickly spread to other countries, including Brazil6. CHIKV produces an acute disease with high fever, headache, nausea, vomiting and conjunctivitis. Patients also develop severe joint pain, which eventually evolves into an arthritogenic syndrome that can last from weeks to years7. Recently, CHIKV infection has also been associated with neurological complications8. There are no antiviral drugs or vaccines available for CHIKV, and the supportive care treatment aims at reducing symptoms and include analgesics, anti-inflammatory and antipyretic drugs.\n\nSome anti-CHIKV molecules have been discovered as a result of antiviral screening campaigns, such as a harringtonine, a plant alkaloid that reduced CHIKV replication by interfering with protein translation in vitro9; D-N4-hydroxycytidine (NHC), a nucleoside analogue, that inhibits RNA synthesis by targeting replication complex10; and barberine, abamectin and ivermectin, which all also reduce viral RNA synthesis11. Most assays were based on replicon systems, a classic way to evaluate drugs that interfere with the viral replication phase, but which cannot account for drugs that might inhibit other steps of the viral cycle, such as cell entry or virion assembly and release. Thus, alternative assays that deploy infectious viral particles, such as those that are based on measurement of cellular infection by high-content screening (HCS)12,13, enable the investigation of compounds that may interact with different stages of infection and lead to the to discover of new classes of antivirals.\n\n\nMethods\n\nHuh-7 hepatocellular carcinoma cells were cultivated in DMEM F-12 medium (Sigma-Aldrich) supplemented with 10% fetal bovine serum (FBS) (Life Technologies), 100 units/ml Penicillin (GIBCO) and 100 µg/ml Streptomycin (GIBCO) at 37°C, 5% CO2. The Vero cell line derived from the kidney of an African green monkey were cultivated in DMEM high glucose (GIBCO) supplemented with 10% FBS (Life Technologies), 100 U/ml penicillin (GIBCO) and 100 μg/ml streptomycin (GIBCO) at 37°C, 5% CO2. Cells were kindly provided by Professor Amilcar Tanuri (UFRJ-Brazil).\n\nThe CHIKV 181/25 vaccine strain was used in the present study14. The strain was obtained from the World Reference Center for Emerging Viruses and Arboviruses (WRCEVA) of the University of Texas Medical Branch. The viral stock was propagated in Vero cells. Supernatant of infected tissue cultures was harvested and titrated by plaque assay15. Briefly, Vero cells were seeded in 24-well plate and incubated at 37°C, 5% CO2 for 24 hours. Virus suspension was diluted 10-fold in DEMEN high glucose, and 0.2 µl from each virus dilution was added to infect Vero cells. After 1 hour in 37°C, 5% CO2, inoculum was removed and cells were washed twice with Dulbecco’s Phosphate-buffer saline (DPBS pH 7.4, Sigma-Aldrich). An overlay was added with High-glucose DMEM (GIBCO), 10% FBS (Life Technologies) and 3.5% carboxymethylcellulose (CMC, Sigma-Aldrich) prepared in distillated water. At 3 days after, the overlay was removed. Then, cells were fixed with 4% paraformaldehyde (PFA) diluted in DPBS and stained with 0.5% crystal violet to enable plaque visualization and counting. Virus titers were expressed as plaque forming units (PFU) per milliliter.\n\nMouse hyperimmune sera was obtained from previously prepared stocks16. Briefly, to prepare these stocks mice (Mus musculus) received 4 weekly inoculations of 0.2 ml of brain macerate suspensions from newborn mice infected with CHIKV in PBS, by the intraperitoneal route. At 5 days after the last immunization the animals were anesthetized and underwent intracardiac puncture for blood collection. CHIKV-MHS was obtained from this blood.\n\nThe compounds 6-azauridine (CAS #54251), interferon α2A (CAS #H6041), bafilomycin A1 (CAS # 88899552), chloroquine (CAS #50635), 5-fluorouracil (CAS #51218) and the Library of Pharmacologically Active Compounds (LOPAC), containing 1,280 compounds, were purchased from Sigma-Aldrich. Sofosbuvir and daclatasvir were kindly provided by Microbiológica Química e Farmacêutica (Brazil) and ledispavir was donated by MedChemExpress (USA).\n\nHuh-7 cells were seeded in black polystyrene 384-well assay plates (Greiner Bio-One) at 3,000 cells/well in 40 µl DMEM-F12 supplemented with 10% FBS and incubated overnight. Cells were infected with 10 µl of inoculum of CHIKV 181/25 at different multiplicities of infection (MOIs) of 0.5, 0.05 and 0.01. Plates were fixed at different periods of time (36, 48 and 72 hours) and submitted to the immunofluorescence assay (described below) and images are acquired using an InCell Analyzer 2200 (GE Life Sciences).\n\nA library stock plate containing the aforementioned compounds at 2 mM in DMSO was used to prepare the intermediate plate by a 16.6-fold dilution in DPBS, to a concentration of 60 µM and 3% DMSO. Then, 10 µl of the intermediate plate content was transferred onto the cell-containing plate. The final concentration of library compounds in the assay plate was 10 µM, with 0.5% DMSO. Controls were placed in lateral columns in all plates. Positive controls were infected cells treated with 50 µM of 6-azauridine as well as non-infected cells treated with vehicle (0.5% DMSO in DPBS). Negative controls were infected cells treated with vehicle. Cells were infected by 10 µl of CHIKV 181/25 at MOI 0.05. Plates were incubated for 48 h at 37°C, 5% CO2 under humidified atmosphere, and then fixed with 4% (w/v) PFA for 15 min at room temperature and washed twice with DPBS. Then, plates were incubated with CHIKV-MHS (mouse hyperimmune sera) diluted 1:1500 (v/v) prepared in blocking buffer (DPBS containing 5% FBS) for 30 min. Each plate was washed twice with DPBS, followed by incubation at room temperature for 30 min with the AlexaFluor488-conjugated goat anti-mouse IgG (Cat No. A-11001, Thermo-Scientific) diluted 1:2000 (v/v), and 5 µg/ml of 4’,6-diamidino-2-phenylindole (DAPI, Sigma-Aldrich) in DPBS. Each plate was washed twice with DPBS. All plates were filled up with 50 µl of PBS/well.\n\nImages were acquired using a confocal microscope High-Content System InCell Analyzer 2200 (GE Life Sciences) and processed by InCell Investigator v.1.6.1 software (GE, USA). Four different images were acquired from each well at x20 magnification. Automated image analysis was performed cell-by-cell through a defined mask based on fluorescence signal measurement and cell morphology. Cell segmentation parameters defined the analysis performed by the Investigator software. The nuclei were segmented from the DAPI staining images and each nucleus was determined as a minimum area of 50 µM. Total cells were filtered from the AlexaFlour488 channel and were defined as minimum area of 100 µM. The mean AlexaFluor488 florescence signal of infected cells were defined from the cytoplasm mask, with values based on means signal from fluorescence of wells from infected cells showing a six-times higher value compared to the mean of Alexa488 fluorescence signal of non-infected cells. Images were treated using image analysis program ImageJ v1.51 to set up colors and merge image channels to final visualization. The validation of screening was conducted in two independent experiment.\n\nInfection ratio (IR) was defined as the ratio between (i) the total number of infected cells, and (ii) the total number of cells. Data were normalized with the negative (DMSO-treated, infected cells) and positive (infected cells treated with 50 µM 6-azauridine) controls. Normalized activity was calculated as described by Pascoalino et al.12. Cell survival was expressed as the percentage of the total cell number from test sample divided by the average total cell number from the positive control wells: Cell number test sample/Avg. cell number of positive control) × 100. Normalized activity and cell survival values were processed with the GraphPad Prism software version 7. Representative graphs of sample distribution were obtained by plot data at TIBCO Spotfire 7.0 software. Plates were also submitted to quality control measurement of Z’-factor as described by Zhang et al.17.\n\nFor dose-response curves, drugs were prepared as described12. The initial test concentrations were 100 nM for IFN-α2A, 50 nM for bafilomycin, 120 μM for chloroquine and mycophenolic acid, and 100 μM for sofosbuvir, daclatasvir, ledispavir and 5-fluorouracil. Dose response assay were calculated based on percentage of normalized activity and cells survival for each concentration tested. Data were plotted with GraphPad Prism software version 7. The sigmodal dose-response curve (variable slope) function were used to calculate the effective concentration that inhibited 50% of infection (EC50), and the concentration of compound that presented a 50% reduction in cell number in comparison to the controls (CC50). The ratio between CC50 and EC50 determines the selective index (SI).\n\nTwo-way analysis of variance (ANOVA) with Sidak’s test, a multiple comparison test, was conducted to calculate statistical significance (P < 0.05) of cell numbers from non-infected and infected CHIKV 181/25 at different multiplicity of infection and incubation time experiment. The coefficient of determination (R2) test were using to determinate statically coefficient of variation between screening replicates from normalize activity data. All data were plotted using GraphPad Prism Software version 7.\n\n\nResults\n\nA high-content screening assay was developed to evaluate compounds activity against CHIKV infection in vitro. The first step consisted on defining the cell model to support viral infection. A range of cell lines has been reported as being susceptible to CHIKV infection, such as Vero, human fetal lung fibroblast (MRC-5), baby hamster kidney (BHK), human embryonic kidney 293 (HEK-239T) and Huh-718,19. The Huh-7 cell line was selected as it has desirable features for high content imaging, such as adherent monolayer growth, and is human cell line, meaning it is a more representative in vitro model than would be cells of another species. The second step was determining the optimal multiplicity of infection (MOI) and the necessary period of time for the efficient viral infection in 384-well plates. Cells were infected at three different MOI (0.5, 0.05 and 0.01) and incubated for different periods of time (36, 48 and 72 hours). The total cell number and the IR were determined. When cells were plated and infected concomitantly, even the lowest MOI tested showed high cytopathic effect (data not shown). Thus, cells were plated 24 h before infection (Figure 1A). For the highest MOI, CHIKV infection decreased cell number by 70% at 48 hours and by almost 100% at 72 hours compared with non-infected cells. There was no significant difference in cell number between non-infected cells and infected cells for both 0.05 and 0.01 MOIs at 36 hours. Compared to non-infected cells, a decrease in cell number by 42% and 22% at MOIs 0.05 and 0.01, respectively, was observed at 48 hours (Figure 1A). After 36 hours of incubation, the IR showed an association with the MOI: 0.95±0.04 (MOI 0.5), 0.40±0.19 (MOI 0.05) and 0.12±0.20 (MOI 0.01) (Figure 1A), demonstrating that the assay endpoint was within the dynamic range of the infection. For 48 hours of incubation, the IR reached 0.99 for all MOIs, but the lowest MOI gave high variation in IR between replicate wells. Therefore, with the aim of testing drugs, a 0.05 MOI at 48 hours of incubation was selected for further experiments to achieve the longer time of exposure to drug treatment possible under these conditions, a high ratio of infection with minor variability, and a cell survival rate greater than 50% (compared with non-infected) (Figure 1A). Figure 1B describes the established general scheme of CHIKV high-content assay.\n\n(A) Huh-7 cells were infected with different MOIs (0.5, 0.05 and 0.01) of CHIKV. The cell number and the infection ratio (defined as the ratio between total cell number and number of infected cells per sample), were evaluated after 36, 48 and 72 hours of infection. Error bars represent the standard deviation of 48 wells. Quantification of total cell number was comparable between non-infected and CHIKV 181/25-infected cells (p < 0.05). (B) General scheme of CHIKV high content assay. On day 1, Huh-7 cells were plated onto 384-well plates at 3000 cells/well. Then, after 24 hours (day 2), compounds were added, followed by addition of virus diluted at MOI 0.05. The plates were incubated for 48 hours up to immunofluorescence assay and high content analysis on day 4.\n\nTo validate the assay, high-content screening was run using a commercial library of compounds. Cell infection was determined by indirect CHIKV immunofluorescence detection. Figure 2A shows a raw image and software segmentation analysis of the same image. The 6-azauridine compound was previously reported to have activity against CHIKV20, and was chosen as the reference compound in this assay. The activity of 6-azauridine was assessed using a dose-response curve (Figure 2B). The EC50 of 0.65 µM 6-azauridine and EC100 of 50 µM 6-azauridine were determined against CHIKV. In order to validate the assay reproducibility and robustness, a commercial library composed of 1,280 compounds was tested at a single concentration (10 µM). A good window between positive and negative controls was observed (Figure 2C). As a result, the mean for all plates Z’-factor values were 0.86±0.09, indicating that the established assay is reliable. Additionally, there was a high correlation coefficient between runs (Coefficient of determination R2: 0.86), which was determined using normalized activity of each single well between the first (R1) and the second (R2) screens, including compounds and controls (Figure 2D).\n\n(A) Interface of image processing and analysis. (B) Dose-response curve for reference compound 6-azauridine. Left y-axis: Normalized activity values (black squares and curves); Right y-axis: Cell survival values in percentage (red dots and curves); x-axis: Log of molar compound concentrations. Data points are means, and error bars represent standard deviations from two independent experiments. (C) Left graphic: Representative scatter plot of the infection ratio showing controls separation. Dots represent each single tested well and colors represent different treatments, where: 0.5% DMSO negative control (yellow); non-infected cells (red); 50 µM 6-azauridine positive control (green). Z’-factor value of 0.87 was obtained from total data controls from two independent runs for screening validation, between inter-replicates and intra-replicates plates. The continuous line represents the mean of each control, the dotted line represents 3 standard deviations from the mean of the negative and positive controls. Right graphic: Data correlation (normalized activity) from two screening runs of a small library; dots represent each single tested well and colors represent different treatments, where: 0.5% DMSO negative control (yellow); non-infected cells (red); 50 µM 6-azauridine positive control (green); and tested compounds (gray). The coefficient of determination of R square (R2 0.86) was calculated using GraphPad Prism software.\n\nA set of compounds with known antiviral activity were evaluated against CHIKV. A total of 9 compounds were tested in dose-response curves. Figure 3 lists the name, molecular structures and dose-response curve plots for all compounds. Interferon α2A (IFN-α2A) and mycophenolic acid have reported activity against CHIKV19,20. The HCS assay confirmed their reported activity, demonstrated by EC50 values of 0.7 nM and 0.8 μM and high SI of >14 and 8.25, for IFN-α2A and mycophenolic acid, respectively. In order to evaluate compounds with previously reported activity against CHIKV, bafilomycin A121 and chloroquine22 were tested, giving an EC50 of 0.01 μM and 21 μM, respectively; however, they were cytotoxic in Huh-7 cells, with low SI values (5 and 0.3, respectively). The antiviral activity of daclatasvir, an anti-hepatitis C virus (HCV) drug23, against CHIKV was associated with high cytotoxicity, and it had a low SI value of 1.3. Ledispavir, also an anti-HCV compound24, and 5-fluorouracil, which has reported activity against ZIKV12, did not present anti-CHIKV activity in the HCS assay. Sofosbuvir is an FDA-approved compound against HCV, and has been recently described as an active compound against other flaviviruses, including dengue and Zika25,26. Sofosbuvir demonstrated dose-dependent activity against CHIKV (EC50, 11 μM) with no cytotoxicity in Huh-7 cells. Representative images of sofosbuvir activity, alongside 6-azauridine, are displayed in Figure 4.\n\nDose response curves of interferon α2A, sofosbuvir, daclatasvir, ledispavir, bafilomycin A1, chloroquine, 5-fluorouracil and mycophenolic acid as anti-CHIKV activity (black) or the effect on Huh-7 survival (Red). Values of EC50 means effective concentration of 50% infection inhibition, and SI means selective index (SI) based on CC50 concentration of compounds of 50% cytotoxicity (not showed). ND, non-determined values.\n\nPanel of representative images. First line of images shows nuclei staining with DAPI. Second line of images shows immunofluorescence against CHIKV (AlexaFluor488). Third line of images is the merge between the lines one and two. First column displays non-infected cells, followed by infected cells treated with 0.5% DMSO, 50 µM 6-azauridine, 12.5 µM sofosbuvir and 20 µM sofosbuvir.\n\n\nDiscussion\n\nCurrently, most assays available for drug screening against CHIKV are based on cell viability methodologies, which evaluate the compounds capacity to prevent cell lysis27–30. Such approaches have the advantage of being of lower in cost and higher-throughput than image-based phenotypic assays. However, background noise interference in quality and usage of counter-screening assays to assess compound cytotoxicity and support conclusions should be considered. Conversely, HCS assays provide multi-parametric evaluation of both viral infection and cytotoxicity in same assay28,31. Therefore, in the present study we propose the development of a reproducible, phenotypic HCS for CHIKV, in order to trial drugs with antiviral activity.\n\nDifferent approaches have been described to assess drugs in a high-throughput screening (HTS) format against CHIKV, including measurement of cell viability using a resazurin assay27, replicon-based assay using a Renilla luciferase reporter10,11,32,33, which targets only replication-process-interfering compounds, or a HCS assay using BHK cells9. In this study, we opted for the Huh-7 cell line as this has been used for HCS for antiviral discovery by our group and others for hepatitis C34, dengue13 and Zika12,35. Moreover, it is reported that Huh-7 cells are permissive to CHIKV infection. The CHIKV viral cycle usually happens in a short period of time, between 8 and 16 hours, following high cytopathic effect36. In this manner, we opted to use a relatively low MOI (0.05) to prevent high cell lysis (Figure 1A), and it can be expected that multiple infection cycles happen during the assay duration (48 h). Thus, all potential targets during the whole viral cycle can be exposed to the compounds. The developed assay also proved to be robust and reproducible.\n\nIFN-2αA, bafilomicyn A1, chloroquine and mycophenolic acid had all been previously reported as active against CHIKV in vitro21,22,37, and their antiviral activity was confirmed under the conditions used in this study. However, bafilomycin A1 and chloroquine were cytotoxic, resulting in low SIs (<5). Bafilomicyin A1, an inhibitor of mammalian vacuolar-type H(+)-ATPase, prevents the acidification of the endosomal compartment, where the low pH allows the fusion of viral capsid followed by the entrance in the cytoplasm, thus preventing a crucial early step of the CHIKV virus cycle38. The same inhibition mechanism was observed in vitro during for the infection of sindbis virus, a prototype alphavirus39. Previous studies have reported the cytotoxicity of bafilomicyin A1 in HEK123 cells21, as was also observed here in Huh-7 (Figure 3). The activity of chloroquine, an antimalarial compound, was extensively investigated against CHIKV, although studies in vivo with infected mice showed inefficient activity22,40. In addition, clinical trials comparing double-blinded placebo groups and patients with CHIKV infection group did not present convincing data regarding chloroquine treatment efficacy41. Studies in CHIKV-infected Vero cells suggested that chloroquine exerts antiviral activity by preventing CHIKV internalization. The chloroquine EC50 values observed in this study (21 μM) in Huh-7-infected cells are in accordance with values previously reported for Vero cells (17 μM)22. However, the cytotoxicity of chloroquine seems to vary depending on the cell type or assay conditions, as chloroquine showed greater cytotoxicity in Huh-7 cells (with values of CC50 of 56 µM) than Vero cells (CC50 >100 µM)22. Mycophenolic acid inhibits inosine monophosphate dehydrogenase (IMPDH), an essential enzyme in de novo biosynthesis of guanine, and has been reported to have antiviral activity for both single strand RNA negative and positive viruses, for instance, against influenza virus42, dengue virus43, and the alphavirus VEEV44. CHIKV activity was previously reported to have EC50 values of 0.1 µM in Vero cells37, and here was observed in Huh-7 EC50 of 0.8 µM, confirming values in similar levels (Figure 3) albeit with lower selectivity. Nevertheless, a human cell model, such as the Huh-7 cell line, should be preferred to screen antiviral candidates to promote a more representative values of activity and cytotoxicity, which may diverge when compared to non-human cells lines45.\n\nSofosbuvir, 5-fluorouracil, daclatasvir and ledispavir are all FDA-approved drugs with reported activity against flaviviruses. The nucleoside analog 5-flurouracil is used to treat neoplastic disease46, and we have recently shown its antiviral activity in vitro against ZIKV infection12. However, 5-flurouracil presented no activity against CHIKV in our assay, suggesting selectivity for flaviviruses. Comparable results were obtained for ledispavir, which did not show inhibition against CHIKV, even at the highest concentration tested. Ledispavir, daclatasvir and sofosbuvir are direct-acting antiviral agents and have been successfully used to treat HCV-infected subjects. Those compounds target NS5A and NS5B, two HCV non-structural proteins47. NS5A presents three domains, which are responsible for genome replication, virus assembly through production of infection virus particles, and regulation of viral genome replication, from direct interaction of NS5A domain II with NS5B. NS5B is an RNA-dependent RNA-polymerase (RdRp) that directs the RNA synthesis in the HCV replication cycle48. Ledispavir and daclatasvir are NS5A inhibitors, while sofosbuvir, an uridine nucleoside analog that targets NS5B that is usually administrated in combination with daclatasvir49. Besides HCV, recent studies have demonstrated that sofosbuvir can inhibit infection by other flavivirus in vitro and in mice25,26. Our results demonstrated sofosbuvir elicited a concentration-dependent inhibition of CHIKV infection (Figure 3 and Figure 4), suggesting that this drug might have a broader antiviral spectrum than previously known.\n\nDifferently from HCV, which the genome organization consisted in five non-structural proteins (NS2, NS3, NS4A, NS4B, NS5A50), CHIKV possess four non-structural protein (nsP1, nsP2, nsP3 and nsP4), being the RdRp domain localized at nsP4. Alignment sequence of RdRp has demonstrated highly conserved regions between CHIKV and other flaviviruses. More specifically, the motif B region, which is a functional domain of viral RdRp coding region, and the R1 motif, which has a role in nucleoside triphosphate binding during viral RNA synthesis, are highly conserved. Besides, CHIKV RdRp forms similar structures to the RdRp of other RNA viruses51. The search for direct-target compounds against CHIKV have focused on nsP2, due to its multifunctioning domains, which acts as helicases to form RNA secondary structures, as triphosphates responsible for RNA capping enzyme and removing terminal phosphate from new RNA template, and as proteases responsible for processing non-structural polyproteins52. In addition, its well-known structure makes nsP2 a suitable target for drug design53. However, few studies have focused on the search for compounds that target RdRp for CHIKV54,55. A compound that targets RdRp would be attractive, as RdRp acts on a viral process, is essential for replication of the viral genome and does not affect host cells55.\n\nIn conclusion, the phenotypic high content analysis established herein revealed that sofosbuvir is a promising candidate for use against CHIKV infection. Further studies should be performed in order to elucidate the exact mechanism related to CHIKV RdRp inhibition by sofosbuvir.\n\n\nData availability\n\nDataset 1. All raw data from the present study. Raw data are separated according to the figure in which they are presented; a guide to the data is available as a .docx file. DOI: https://doi.org/10.5256/f1000research.16498.d22190556.", "appendix": "Grant information\n\nThis work has been supported by the Sao Paulo State Research Foundation - FAPESP (Process no. 2016/03780-5).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe would like to thank Dr Paolo Zanotto, who provided the virology laboratory for performance of all experiments at Department of Microbiology, ICB-USP. We thank Denise Pilger for technical assistance, Dr. Amilcar Tanuri for supplying Huh-7 and Vero cells lines, Microbiológica Química e Farmacêutica for supplying sofosbuvir and daclatasvir, MedChemExpress (USA) for supplying Ledispavir and the World Reference Center for Emerging Viruses and Arboviruses for support to obtain the CHIKV 181/25 vaccine strain.\n\n\nReferences\n\nStrauss JH, Strauss EG: The alphaviruses: gene expression, replication, and evolution. Microbiol Rev. 1994; 58(3): 491–562. PubMed Abstract \| Free Full Text\n\nAréchiga-Ceballos N, Aguilar-Setién A: Alphaviral equine encephalomyelitis (Eastern, Western and Venezuelan). Rev Sci Tech. 2015; 34(2): 491–501. PubMed Abstract \| Publisher Full Text\n\nLwande OW, Obanda V, Bucht G, et al.: Global emergence of Alphaviruses that cause arthritis in humans. Infect Ecol Epidemiol. 2015; 5(1): 29853. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPaul BJ, Sadanand S: Chikungunya Infection: A Re-emerging Epidemic. Rheumatol Ther. 2018; 1–10. PubMed Abstract \| Publisher Full Text\n\nPialoux G, Gaüzère BA, Jauréguiberry S, et al.: Chikungunya, an epidemic arbovirosis. Lancet Infect Dis. 2007; 7(5): 319–327. PubMed Abstract \| Publisher Full Text\n\nMorrison TE: Reemergence of chikungunya virus. J Virol. 2014; 88(20): 11644–7. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGanesan VK, Duan B, Reid SP: Chikungunya Virus: Pathophysiology, Mechanism, and Modeling. Viruses. 2017; 9(12): pii: E368. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAgarwal A, Vibha D, Srivastava AK, et al.: Guillain-Barre syndrome complicating chikungunya virus infection. J Neurovirol. 2017; 23(3): 504–507. PubMed Abstract \| Publisher Full Text\n\nKaur P, Thiruchelvan M, Lee RC, et al.: Inhibition of chikungunya virus replication by harringtonine, a novel antiviral that suppresses viral protein expression. Antimicrob Agents Chemother. 2013; 57(1): 155–67. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEhteshami M, Tao S, Zandi K, et al.: Characterization of β-d-N4-Hydroxycytidine as a Novel Inhibitor of Chikungunya Virus. Antimicrob Agents Chemother. 2017; 61(4): pii: e02395-16. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVarghese FS, Kaukinen P, Gläsker S, et al.: Discovery of berberine, abamectin and ivermectin as antivirals against chikungunya and other alphaviruses. Antiviral Res. 2016; 126: 117–24. PubMed Abstract \| Publisher Full Text\n\nPascoalino BS, Courtemanche G, Cordeiro MT, et al.: Zika antiviral chemotherapy: identification of drugs and promising starting points for drug discovery from an FDA-approved library [version 1; referees: 2 approved]. F1000Res. 2016; 5: 2523. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCruz DJ, Koishi AC, Taniguchi JB, et al.: High content screening of a kinase-focused library reveals compounds broadly-active against dengue viruses. PLoS Negl Trop Dis. 2013; 7(2): e2073. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLevitt NH, Ramsburg HH, Hasty SE, et al.: Development of an attenuated strain of chikungunya virus for use in vaccine production. Vaccine. 1986; 4(3): 157–162. PubMed Abstract \| Publisher Full Text\n\nNakao E, Hotta S: Immunogenicity of purified, inactivated chikungunya virus in monkeys. Bull World Health Organ. 1973; 48(5): 559–562. PubMed Abstract \| Free Full Text\n\nFigueiredo LT: [The use of Aedes albopictus C6/36 cells in the propagation and classification of arbovirus of the Togaviridae, Flaviviridae, Bunyaviridae and Rhabdoviridae families]. Rev Soc Bras Med Trop. 1990; 23(1): 13–8. PubMed Abstract \| Publisher Full Text\n\nZhang JH, Chung TD, Oldenburg KR: A Simple Statistical Parameter for Use in Evaluation and Validation of High Throughput Screening Assays. J Biomol Screen. 1999; 4(2): 67–73. PubMed Abstract \| Publisher Full Text\n\nRoberts GC, Zothner C, Remenyi R, et al.: Evaluation of a range of mammalian and mosquito cell lines for use in Chikungunya virus research. Sci Rep. 2017; 7(1): 14641. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWikan N, Sakoonwatanyoo P, Ubol S, et al.: Chikungunya virus infection of cell lines: analysis of the East, Central and South African lineage. PLoS One. 2012; 7(1): e31102. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBriolant S, Garin D, Scaramozzino N, et al.: In vitro inhibition of Chikungunya and Semliki Forest viruses replication by antiviral compounds: synergistic effect of interferon-alpha and ribavirin combination. Antiviral Res. 2004; 61(2): 111–7. PubMed Abstract \| Publisher Full Text\n\nKarlas A, Berre S, Couderc T, et al.: A human genome-wide loss-of-function screen identifies effective chikungunya antiviral drugs. Nat Commun. 2016; 7: 11320. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKhan M, Santhosh SR, Tiwari M, et al.: Assessment of in vitro prophylactic and therapeutic efficacy of chloroquine against Chikungunya virus in Vero cells. J Med Virol. 2010; 82(5): 817–824. PubMed Abstract \| Publisher Full Text\n\nDaclatasvir. Drugs and Lactation Database (LactMed). 2006. PubMed Abstract\n\nLink JO, Taylor JG, Xu L, et al.: Discovery of ledipasvir (GS-5885): a potent, once-daily oral NS5A inhibitor for the treatment of hepatitis C virus infection. J Med Chem. 2014; 57(5): 2033–2046. PubMed Abstract \| Publisher Full Text\n\nXu HT, Colby-Germinario SP, Hassounah SA, et al.: Evaluation of Sofosbuvir (β-D-2′-deoxy-2′-α-fluoro-2′-β-C-methyluridine) as an inhibitor of Dengue virus replication. Sci Rep. 2017; 7(1): 6345. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBullard-Feibelman KM, Govero J, Zhu Z, et al.: The FDA-approved drug sofosbuvir inhibits Zika virus infection. Antiviral Res. 2017; 137: 134–140. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCruz DJ, Bonotto RM, Gomes RG, et al.: Identification of novel compounds inhibiting chikungunya virus-induced cell death by high throughput screening of a kinase inhibitor library. PLoS Negl Trop Dis. 2013; 7(10): e2471. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSmee DF, Hurst BL, Evans WJ, et al.: Evaluation of cell viability dyes in antiviral assays with RNA viruses that exhibit different cytopathogenic properties. J Virol Methods. 2017; 246: 51–57. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMüller JA, Harms M, Schubert A, et al.: Development of a high-throughput colorimetric Zika virus infection assay. Med Microbiol Immunol. 2017; 206(2): 175–185. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMcCormick KD, Liu S, Jacobs JL, et al.: Development of a robust cytopathic effect-based high-throughput screening assay to identify novel inhibitors of dengue virus. Antimicrob Agents Chemother. 2012; 56(6): 3399–3401. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRampersad SN: Multiple applications of Alamar Blue as an indicator of metabolic function and cellular health in cell viability bioassays. Sensors (Basel). 2012; 12(9): 12347–12360. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPohjala L, Utt A, Varjak M, et al.: Inhibitors of alphavirus entry and replication identified with a stable Chikungunya replicon cell line and virus-based assays. PLoS One. 2011; 6(12): e28923. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAshbrook AW, Lentscher AJ, Zamora PF, et al.: Antagonism of the Sodium-Potassium ATPase Impairs Chikungunya Virus Infection. mBio. 2016; 7(3): pii: e00693-16. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKim HY, Li X, Jones CT, et al.: Development of a multiplex phenotypic cell-based high throughput screening assay to identify novel hepatitis C virus antivirals. Antiviral Res. 2013; 99(1): 6–11. PubMed Abstract \| Publisher Full Text\n\nBarrows NJ, Campos RK, Powell ST, et al.: A Screen of FDA-Approved Drugs for Inhibitors of Zika Virus Infection. Cell Host Microbe. 2016; 20(2): 259–270. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSourisseau M, Schilte C, Casartelli N, et al.: Characterization of reemerging chikungunya virus. PLoS Pathog. 2007; 3(6): e89. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKhan M, Dhanwani R, Patro IK, et al.: Cellular IMPDH enzyme activity is a potential target for the inhibition of Chikungunya virus replication and virus induced apoptosis in cultured mammalian cells. Antiviral Res. 2011; 89(1): 1–8. PubMed Abstract \| Publisher Full Text\n\nBernard E, Solignat M, Gay B, et al.: Endocytosis of chikungunya virus into mammalian cells: role of clathrin and early endosomal compartments. PLoS One. 2010; 5(7): e11479. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchuchman RM, Vancini R, Piper A, et al.: Role of the vacuolar ATPase in the Alphavirus replication cycle. Heliyon. 2018; 4(7): e00701. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMaheshwari RK, Srikantan V, Bhartiya D: Chloroquine enhances replication of Semliki Forest virus and encephalomyocarditis virus in mice. J Virol. 1991; 65(2): 992–995. PubMed Abstract \| Free Full Text\n\nDe Lamballerie X, Boisson V, Reynier JC, et al.: On chikungunya acute infection and chloroquine treatment. Vector Borne Zoonotic Dis. 2008; 8(6): 837–839. PubMed Abstract \| Publisher Full Text\n\nTo KK, Mok KY, Chan AS, et al.: Mycophenolic acid, an immunomodulator, has potent and broad-spectrum in vitro antiviral activity against pandemic, seasonal and avian influenza viruses affecting humans. J Gen Virol. 2016; 97(8): 1807–17. PubMed Abstract \| Publisher Full Text\n\nTakhampunya R, Ubol S, Houng HS, et al.: Inhibition of dengue virus replication by mycophenolic acid and ribavirin. J Gen Virol. 2006; 87(Pt 7): 1947–52. PubMed Abstract \| Publisher Full Text\n\nLundberg L, Brahms A, Hooper I, et al.: Repurposed FDA-Approved drug sorafenib reduces replication of Venezuelan equine encephalitis virus and other alphaviruses. Antiviral Res. 2018; 157: 57–67. PubMed Abstract \| Publisher Full Text\n\nVicenti I, Boccuto A, Giannini A, et al.: Comparative analysis of different cell systems for Zika virus (ZIKV) propagation and evaluation of anti-ZIKV compounds in vitro. Virus Res. 2018; 244: 64–70. PubMed Abstract \| Publisher Full Text\n\nGajjar KK, Vora HH, Kobawala TP, et al.: Deciphering the potential value of 5-fluorouracil metabolic enzymes in predicting prognosis and treatment response of colorectal cancer patients. Int J Biol Markers. 2018; 33(2): 180–8. PubMed Abstract \| Publisher Full Text\n\nGötte M, Feld JJ: Direct-acting antiviral agents for hepatitis C: structural and mechanistic insights. Nat Rev Gastroenterol Hepatol. 2016; 13(6): 338–351. PubMed Abstract \| Publisher Full Text\n\nHughes M, Griffin S, Harris M: Domain III of NS5A contributes to both RNA replication and assembly of hepatitis C virus particles. J Gen Virol. 2009; 90(Pt 6): 1329–1334. PubMed Abstract \| Publisher Full Text\n\nKeating GM, Vaidya A: Sofosbuvir: first global approval. Drugs. 2014; 74(2): 273–82. PubMed Abstract \| Publisher Full Text\n\nPenin F, Dubuisson J, Rey FA, et al.: Structural biology of hepatitis C virus. Hepatology. 2004; 39(1): 5–19. PubMed Abstract \| Publisher Full Text\n\nte Velthuis AJ: Common and unique features of viral RNA-dependent polymerases. Cell Mol Life Sci. 2014; 71(22): 4403–20. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRupp JC, Sokoloski KJ, Gebhart NN, et al.: Alphavirus RNA synthesis and non-structural protein functions. J Gen Virol. 2015; 96(9): 2483–500. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSaha A, Acharya BN, Priya R, et al.: Development of nsP2 protease based cell free high throughput screening assay for evaluation of inhibitors against emerging Chikungunya virus. Sci Rep. 2018; 8(1): 10831. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDelang L, Segura Guerrero N, Tas A, et al.: Mutations in the chikungunya virus non-structural proteins cause resistance to favipiravir (T-705), a broad-spectrum antiviral. J Antimicrob Chemother. 2014; 69(10): 2770–84. PubMed Abstract \| Publisher Full Text\n\nWada Y, Orba Y, Sasaki M, et al.: Discovery of a novel antiviral agent targeting the nonstructural protein 4 (nsP4) of chikungunya virus. Virology. 2017; 505: 102–12. PubMed Abstract \| Publisher Full Text\n\nBonotto RM, Souza-Almeida G, Badra SJ, et al.: Dataset 1 in: Evaluation of broad-spectrum antiviral compounds against chikungunya infection using a phenotypic screening strategy. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16498.d221905" }	[ { "id": "40081", "date": "19 Nov 2018", "name": "Rana Abdelnabi", "expertise": [], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article describe the optimization of a high throughput screening assay for CHIKV compounds and its validation using a commercial library of 12800 compounds. The assay was developed in Huh7 cells and the endpoint was detected using indirect immunofluorescence staining with CHIKV specific antibodies. The author showed that the assay is working and they confirmed that using a couple of known CHIKV inhibitors. However, there are some concerns regarding the used assay:\n\nThe readout of this assay depends on IFAs which: i) require several steps of washing, fixation and staining and ii) are expensive compared to the ordinary colorimetric methods (e.g. MTS/PMS) because of the use of AlexaFlour antibody. Therefore, the use of this assay for high throughput screening would be time and money consuming. It will be more beneficial and technically sound to optimize high throughput assays using reporter CHIKV. This will eliminate the need for IFAs, which is the main drawback of this assay. It is not addressed clearly in the manuscript what is the added value of the developed assay over the already established CHIKV assays. Furthermore, the authors did not mention whether they found any new hits in the screened library (12800 compounds) using their developed assay. Were all the tested compounds inactive or there are some new hits? This part needs clarification.\n\nIs the rationale for developing the new method (or application) clearly explained? Partly\n\nIs the description of the method technically sound? No\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? No source data required\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? No", "responses": [] }, { "id": "40085", "date": "26 Nov 2018", "name": "Mukesh Kumar", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript entitled \"Evaluation of broad-spectrum antiviral compounds against chikungunya infection using a phenotypic screening strategy” by Bonotto and colleagues evaluated an image-based phenotypic assay for high-throughput screening of anti-CHIKV compounds. The screening assay was validated by testing a commercial library of 1,280 compounds, including FDA-approved drugs. The research topic is interesting and has potential significance. However, there are major gaps in the depth of the information reported in this manuscript that make publication of the findings in its current form problematic. Moreover, results in this manuscript are poorly presented and inadequate information is provided.\nIn this manuscript, the screening assay was validated by testing a commercial library of 1,280 compounds. Where is the data on these compounds? Any anti-CHIKV activity by new compound? Not sure why these compounds were screened.\nIt is not clear what are the advantages of using this image-based assay over other established assays for drug screening. IFA is expensive and time-consuming. Discussion is more about efficacy of anti-viral compounds tested rather than use of IFA-based screening strategy.\nThe anti-CHIKV activity of Sofosbuvir must be validated by using another gold-standard assay such as plaque-assay.\n\nIs the rationale for developing the new method (or application) clearly explained? Partly\n\nIs the description of the method technically sound? Partly\n\nAre sufficient details provided to allow replication of the method development and its use by others? Partly\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Partly", "responses": [] }, { "id": "43513", "date": "25 Feb 2019", "name": "Paban Kumar Dash", "expertise": [ "Reviewer Expertise Virology" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors have developed high content image-based screening and tested a commercial library of 1,280 compounds, including FDA-approved drugs for antiviral activity against Chikungunya virus. The authors found the drug sofosbuvir to be effective against chikungunya virus and proposed it to be a promising candidate for repurposing.\n\nHigh content screening (HCS) assay developed seems to be of medium throughput due to the cumbersome steps involved. Since this homogeneous cell-based assay involves multiple step additions, therefore number of steps in assay could be minimized either by using reporter virus, labelled virus or fluorophore conjugated antibodies. The assay robustness should be further validated by calculating signal-to-noise ratio, CV, inter-plate, intra-plate and day to day variability. The assay performance should be consistent from day to day; ideally the assay signal should not vary with passage and cell density. The authors should attempt to further validate the activity of Sofosbuvir, using different assays. Further, in this study, the experiments are performed with a low MOI of 0.05, would Sofosbuvir be effective when cells are infected with higher MOI of virus? MOI of 1 and 10 should be evaluated too, though the authors pointed out about highly cytotopathic nature of the virus. As such, the CHIV titre is very high in human patients.\n\nFurthermore, from the figure 3 data (sofosburvir), it might be possible the drug has had a slight proliferative effect on the Huh-7 cells and may have slightly increased the SI (cell viability curve, proliferation may affect infection rates). Please provide evidence indicating that proliferation has not occurred.\n\nThe authors should provide information about the remaining compounds that didn’t work against Chikungunya virus in their assay. At least tabulate the name of compounds; reporting of which will be extremely useful in dissemination of important information among researchers.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Partly\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Partly", "responses": [] } ]	1	https://f1000research.com/articles/7-1730
https://f1000research.com/articles/7-1237/v1	10 Aug 18	{ "type": "Research Note", "title": "Association between intensity of STI screening and development of antimicrobial resistance in N. gonorrhoeae in 12 cities in the USA: An ecological study", "authors": [ "Chris R. Kenyon" ], "abstract": "In this study, we assessed if there was a city-level association between sexually transmitted infection (STI) screening intensity in men who have sex with men and antimicrobial sensitivity in Neisseria gonorrhoeae in the United States, 2007 to 2013. We found positive associations between STI screening intensity and increases in minimum inhibitory concentrations for cefixime and azithromycin, but not ceftriaxone.", "keywords": [ "N. gonorrhoeae", "STI screening", "antimicrobial resistance", "MSM" ], "content": "Introduction\n\nIn the United States (USA) the prevalence of antimicrobial resistance in Neisseria gonorrhoeae has typically been higher in men who have sex with men (MSM) than men who have sex with women (MSW) and women1,2. It has also frequently been noted to be highest in the West and lowest in the South1–3. Resistance has characteristically emerged in the West Coast and Hawaii and then spread eastward1–3. This patterning of spread has led to the view that a primary driver of resistance is the import of resistant gonococci from eastern Asia and other world regions3. In support of this theory, a number of studies have documented travel as a means of import of resistance in the USA4,5. A systematic review of risk factors associated with resistance in N. gonorrhoeae, however, found that a history of sex with partners abroad was associated with resistance in 6 studies and was not associated with resistance in 7 studies6. Furthermore, the evidence that travel plays a seminal role in the emergence of resistance in MSM is not that compelling. An analysis of data from the Gonococcal Isolate Surveillance Project (GISP) 2002 to 2007, for example, found a pronounced increase in ciprofloxacin-resistance in MSM and a smaller and later increase in MSW; the association with recent travel was negative in MSM and borderline positive in MSW7.\n\nAntimicrobial resistance results largely from exposure to antimicrobials8,9. This has been extensively documented in vitro and in vivo but for various reasons antimicrobial pressure at a population level may be more important than at an individual level in determining risk of development of antimicrobial resistance8,10,11. In the case of N. gonorrhoeae, extensive antimicrobial exposure in a population would be predicted to result in a high prevalence of resistance genes in the pharyngeal microbiomes that could then be taken up (via transformation) by N. gonorrhoeae and thereby provide it with a fitness conferring resistant phenotype in the setting of ongoing high antimicrobial consumption12. These insights have provided the rationale for ecological level studies that have generally found strong associations between the intensity of antimicrobial use and the prevalence of resistance to that antimicrobial8,13. A recent study from the USA however found no association between an increase in N. gonorrhoeae minimum inhibitory concentration (MIC) for azithromycin, ceftriaxone, cefixime and ciprofloxacin in the 23 GISP sites and the consumption of antimicrobials in the surrounding county3. A weakness of this study design was the use of total-consumption-of-antimicrobials by the entire county population as the explanatory variable. Since resistance has repeatedly emerged in certain MSM populations, it would be prudent to assess if this emergence is correlated with antimicrobial consumption in this group rather than the entire population. One major driver of antimicrobial consumptions in MSM is sexually transmitted infection (STI) screening. Because most N. gonorrhoeae and Chlamydia trachomatis in MSM are carried asymptomatically in the anorectum and oropharynx, screening for these STIs may result in a large increase in antimicrobial exposure. A modeling study for example found that increasing annual gonorrhea/chlamydia screening in an MSM population from 3 to 50% would result in a 11-fold increase in antimicrobial exposure14. In this exploratory paper we hypothesized that the intensity of STI testing plays a role in the genesis of resistance in MSM via the associated increase in antibiotic exposure.\n\n\nMethods\n\nWe assessed if there was an ecological-city-level-association between the intensity of STI testing in MSM in the USA and the development of antimicrobial resistance in N. gonorrhoeae.\n\nData for STI screening was taken from the 2005, 2008 and 2011 National HIV Behavioral Surveillance MSM (NHBS-MSM) studies. These cross-sectional surveys done in 21 cities asked respondents about STI testing in the preceding 12 months. The 2005 survey (n=10,030) asked if respondents had been tested for syphilis/gonorrhea/another-STI during the preceding 12 months (single question), the 2008 survey (n=8,175) if they had been tested for syphilis in the preceding 12 months, and the 2011 survey (n=8,012) if they had been tested for gonorrhea, chlamydia or syphilis in the previous 12 months (3 questions).\n\nData for the change in city geometric mean N. gonorrhoeae MIC between 2005 and 2013 was taken from GISP data3. Spearman's correlation was used to assess if there was an association between the prevalence of STI testing in each survey and the increase in geometric mean MIC of cefixime, ceftriaxone and azithromycin in N. gonorrhoeae between 2005 and 2013. These three antibiotics were chosen since these were the recommended antibiotics for N. gonorrhoeae therapy since 20071. All analyses were conducted in STATA 13.\n\n\nResults\n\nTwelve cities participated in both the NHBS-MSM and GISP surveys (n=9 for 2005, n=12 for 2008, n=12 for 2011). The intensity of self-reported STI testing in 2005 varied between 27% and 56% (median 43%, IQR 39–49). There was little change in the relative positions of the cities in terms of testing intensity between 2005 and 2008 (rho=0.87, p=0.002) and 2005 to 2011 (rho=0.81, p=0.008). Cities in the West tended to have higher STI testing rates than cities in the South (Figure 1). In 2011, the percent reporting testing for gonorrhea was strongly correlated with the percent reporting testing for chlamydia (rho=0.99, p<0.001) and syphilis (rho=0.98, p<0.001). In general, the N. gonorrhoeae geometric mean MIC for cefixime and azithromycin increased more rapidly than ceftriaxone in all cities (data not shown).\n\nScatter plots of change in the minimum inhibitory concentrations (MICs) between 2007 and 2013 and the percent of respondents reporting in 2005 that they had a bacterial STI test in the prior 12 months for (a) cefixime, (b) ceftriaxone and (c) azithromycin by USA city (data sources detailed in Methods).\n\nIn 2005, significant positive associations were found between STI screening and the increase in MIC of cefixime (rho=0.88, p=0.002), azithromycin (rho=0.93, p<0.001) but not ceftriaxone (rho=0.27, p=0.491; Figure 1). Likewise in 2008, there was a positive correlation between the percent reporting testing for syphilis in the prior 12 months and increase in MIC of cefixime (rho=0.71, p=0.010), azithromycin (rho=0.791, p=0.002) but not ceftriaxone (rho=0.36, p=0.247). A positive association was also found for the percent reporting testing for gonorrhea in 2011 and an increase in MIC for cefixime (rho=0.63, p=0.026) and azithromycin (rho=0.64, p=0.024) but not ceftriaxone (rho=0.56, p=0.062). The results for chlamydia and syphilis testing were similar (data not shown).\n\n\nDiscussion\n\nThere was a roughly two-fold variation in the proportion of MSM in different cities reporting testing for bacterial STIs. The proportion testing for bacterial STIs was associated with an increase of MIC for cefixime and azithromycin but not ceftriaxone. A plausible reason for the lack of association between ceftriaxone and MIC change is that ceftriaxone has been used almost exclusively in combination with azithromycin3,15 and even on its own may be less susceptible to the develop of resistance than cefixime and azithromycin16. These findings are compatible with the theory that screening intensity plays a role in the selection of antimicrobial resistance in N. gonorrhoeae in MSM.\n\nThe findings should however be regarded as tentative due to a number of methodological weaknesses: the sample size was small (n=9 to 12), the outcome variable (increase in geometric MIC) referred to all men sampled in GISP and not just MSM, the explanatory variable was only evaluated at three time points, the explanatory variable measured ‘STI testing’ and not ‘STI screening’ and possible confounders were not controlled for. Increased testing could, for example, be associated with other factors that may be associated with antimicrobial resistance such as greater risk behavior, more frequent travel, HIV-infection and access to medical care. Likewise we did not control for changes over time in the percent of GISP samples derived from MSM which may have influenced the geometric mean MICs.\n\nFuture studies that wish to evaluate the screening-resistance hypothesis could assess if there is an association between bacterial STI screening intensity and resistance in N. gonorrhoeae in bigger samples in the USA or elsewhere. Testing this hypothesis in Europe would be instructive since the proportion of MSM reporting anal screening for bacterial STIs in the prior 12 months in 38 different European countries ranges from 9.1% in Romania to 79.6% in Malta (median 18.5%, IQR 13.5–28.4)17. Furthermore whilst two of these countries that report high STI screening rates in MSM (the United Kingdom17 and the Netherlands17) have found an association between resistance in N. gonorrhoeae and MSM18,19, other countries in Europe have not found this association15.\n\nThe recent emergence of combined azithromycin/ceftriaxone resistant N. gonorrhoeae provides additional motivation to better characterize the underlying determinants of the differential emergence of resistance in MSM and other populations20. If screening intensity is found to play a role then this could be taken into account in development of an optimal STI screening strategy.\n\n\nData availability\n\nData for STI screening: National HIV Behavioral Surveillance MSM (NHBS-MSM) studies\n\nGISP data: Technical appendix of 3", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nKirkcaldy RD, Harvey A, Papp JR, et al.: Neisseria gonorrhoeae Antimicrobial Susceptibility Surveillance - The Gonococcal Isolate Surveillance Project, 27 Sites, United States, 2014. MMWR Surveill Summ. 2016; 65(7): 1–19. PubMed Abstract \| Publisher Full Text\n\nKirkcaldy RD, Zaidi A, Hook EW 3rd, et al.: Neisseria gonorrhoeae antimicrobial resistance among men who have sex with men and men who have sex exclusively with women: the Gonococcal Isolate Surveillance Project, 2005-2010. Ann Intern Med. 2013; 158(5 Pt 1): 321–8. PubMed Abstract \| Publisher Full Text\n\nKirkcaldy RD, Bartoces MG, Soge OO, et al.: Antimicrobial Drug Prescription and Neisseria gonorrhoeae Susceptibility, United States, 2005-2013. Emerg Infect Dis. 2017; 23(10): 1657–1663. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIverson CJ, Wang SA, Lee MV, et al.: Fluoroquinolone resistance among Neisseria gonorrhoeae isolates in Hawaii, 1990-2000: role of foreign importation and increasing endemic spread. Sex Transm Dis. 2004; 31(12): 702–708. PubMed Abstract \| Publisher Full Text\n\nBauer HM, Mark KE, Samuel M, et al.: Prevalence of and associated risk factors for fluoroquinolone-resistant Neisseria gonorrhoeae in California, 2000-2003. Clin Infect Dis. 2005; 41(6): 795–803. PubMed Abstract \| Publisher Full Text\n\nAbraha M, Egli-Gany D, Low N: Epidemiological, behavioural, and clinical factors associated with antimicrobial-resistant gonorrhoea: a review [version 1; referees: 2 approved]. F1000Res. 2018; 7: 400. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGoldstein E, Kirkcaldy RD, Reshef D, et al.: Factors related to increasing prevalence of resistance to ciprofloxacin and other antimicrobial drugs in Neisseria gonorrhoeae, United States. Emerg Infect Dis. 2012; 18(8): 1290–1297. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGoossens H, Ferech M, Vander Stichele R, et al.: Outpatient antibiotic use in Europe and association with resistance: a cross-national database study. Lancet. 2005; 365(9459): 579–587. PubMed Abstract\n\nJakobsson HE, Jernberg C, Andersson AF, et al.: Short-term antibiotic treatment has differing long-term impacts on the human throat and gut microbiome. PLoS One. 2010; 5(3): e9836. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLipsitch M, Samore MH: Antimicrobial use and antimicrobial resistance: a population perspective. Emerg Infect Dis. 2002; 8(4): 347–354. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBell BG, Schellevis F, Stobberingh E, et al.: A systematic review and meta-analysis of the effects of antibiotic consumption on antibiotic resistance. BMC Infect Dis. 2014; 14(1): 13. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKenyon C, Schwartz IS: A combination of high sexual network connectivity and excess antimicrobial usage induce the emergence of antimicrobial resistance in Neisseria gonorrhoeae. Emerg Infect Dis. 2018; 24(7).\n\nBruyndonckx R, Hens N, Aerts M, et al.: Exploring the association between resistance and outpatient antibiotic use expressed as DDDs or packages. J Antimicrob Chemother. 2015; 70(4): 1241–1244. PubMed Abstract \| Publisher Full Text\n\nBuyze J, Vanden Berghe W, Hens N, et al.: Current levels of gonorrhoea screening in MSM in Belgium may have little effect on prevalence: a modelling study. Epidemiol Infect. 2018; 146(3): 333–338. PubMed Abstract \| Publisher Full Text\n\nCole MJ, Spiteri G, Jacobsson S, et al.: Overall Low Extended-Spectrum Cephalosporin Resistance but high Azithromycin Resistance in Neisseria gonorrhoeae in 24 European Countries, 2015. BMC Infect Dis. 2017; 17(1): 617. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChisholm SA, Mouton JW, Lewis DA, et al.: Cephalosporin MIC creep among gonococci: time for a pharmacodynamic rethink? J Antimicrob Chemother. 2010; 65(10): 2141–2148. PubMed Abstract \| Publisher Full Text\n\nThe EMIS Network: EMIS 2010: The European Men-Who-Have-Sex-With-Men Internet Survey. Findings from 38 countries. Stockholm: European Centre for Disease Prevention and Control; 2011. Reference Source\n\nIson CA, Town K, Obi C, et al.: Decreased susceptibility to cephalosporins among gonococci: data from the Gonococcal Resistance to Antimicrobials Surveillance Programme (GRASP) in England and Wales, 2007-2011. Lancet Infect Dis. 2013; 13(9): 762–768. PubMed Abstract \| Publisher Full Text\n\nde Vries HJ, Van der Helm JJ, Schim van der Loeff MF, et al.: Multidrug-resistant Neisseria gonorrhoeae with reduced cefotaxime susceptibility is increasingly common in men who have sex with men, Amsterdam, the Netherlands. Euro Surveill. 2009; 14(37): pii: 19330. PubMed Abstract \| Publisher Full Text\n\nYasuda M, Hatazaki K, Ito S, et al.: Antimicrobial Susceptibility of Neisseria gonorrhoeae in Japan from 2000 to 2015. Sex Transm Dis. 2017; 44(3): 149–153. PubMed Abstract" }	[ { "id": "37099", "date": "20 Aug 2018", "name": "Edward Goldstein", "expertise": [ "Reviewer Expertise Infectious Disease Epidemiology" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe paper under review studies the association between the frequency of STI screening in MSM and changes in geometric mean MICs for certain antibiotics in N. gonorrhoeae samples in select US cities. An important potential source of bias in such analysis could be reverse causality, namely places with higher prevalence of drug resistance for gonorrhea, as well as syphilis in MSM may initiate more STI screening efforts. Below are some questions/suggestions to the author.\n\n1. Please define geometric mean MIC. Also, since one refers to geometric mean, it might be more reasonable to examine fold changes (increase or decreases) in geometric mean MIC (or logarithm of thereof) in the correlation analysis.\n\n2. Nine cities reported STI screening information in 2005 (Figure 1), and 12 cities reported STI screening information in 2013. The reviewer would suggest to correlate the frequency of STI screening in 2013 with\n\nGeometric mean MICs in 2013 (Fold) change in geometric mean MICs between 2009 and 2013 (Fold) change in geometric mean MICs between 2005 and 2013\n\nFor the 2005 screening frequencies, it would be interesting to correlate them with geometric mean MICs (in 2013, and possibly 2005), and not only changes in thereof.\n\n3. In ref. 3, no association was found between rates of antibiotic prescribing and geometric mean MICs. Perhaps the results of the paper under review are affected by reverse causality, namely places with higher prevalence of drug resistance for gonorrhea, as well as syphilis in MSM may initiate more STI screening efforts. For example, the California sites had both high geometric mean MIC for ciprofloxacin in 2005 (ref. 3), and high frequency of screening (see also1). Currently, this possibility is not mentioned in the Discussion. The reviewer was also wondering if there is a correlation between the fold change in screening frequency and the fold change in geometric mean MICs between 2005 and 2013.\n\n4. Are other variables in NHBS-MSM surveys (possibly rate of change of sexual partners, or a number of recent sexual partners) correlated with geometric mean MICs/changes in thereof?\n\n5. In the Discussion, it is mentioned that “the sample size was small (9 or 12)”. Currently, only 9 cities are used in the correlation analysis, as far as the reviewer could see.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4009", "date": "27 Sep 2018", "name": "Chris Kenyon", "role": "Author Response", "response": "The paper under review studies the association between the frequency of STI screening in MSM and changes in geometric mean MICs for certain antibiotics in N. gonorrhoeae samples in select US cities. An important potential source of bias in such analysis could be reverse causality, namely places with higher prevalence of drug resistance for gonorrhea, as well as syphilis in MSM may initiate more STI screening efforts. Below are some questions/suggestions to the author. 1. Please define geometric mean MIC. Also, since one refers to geometric mean, it might be more reasonable to examine fold changes (increase or decreases) in geometric mean MIC (or logarithm of thereof) in the correlation analysis. Reply: The geometric mean MIC has now been defined in the methods section.2. Nine cities reported STI screening information in 2005 (Figure 1), and 12 cities reported STI screening information in 2013. The reviewer would suggest to correlate the frequency of STI screening in 2013 with Geometric mean MICs in 2013 (Fold) change in geometric mean MICs between 2009 and 2013 (Fold) change in geometric mean MICs between 2005 and 2013 For the 2005 screening frequencies, it would be interesting to correlate them with geometric mean MICs (in 2013, and possibly 2005), and not only changes in thereof. Reply:The third NHBS survey was in 2011 (not 2013) and thus we have assessed Spearman's correlation between percent reporting screening for any STI in the previous 12 months and MIC for the three antimicrobials in the following year (2012): Azithromycin: Rho 0.45; P=0.141Cefixime: Rho 0.31; P=0.325Ceftriaxone: Rho=0.64; P=0.026 We also assessed Spearman's correlation between percent reporting screening for any STI in the previous 12 months (2011 survey) and the fold change in geometric mean MIC for the three antimicrobials between a) 2005 and 2013:Azithromycin: Rho 0.77; P=0.003Cefixime: Rho 0.82; P=0.001Ceftriaxone: Rho=0.61; P=0.034 b) 2009 and 2013:Azithromycin: Rho 0.33; P=0.299Cefixime: Rho -0.50; P=0.100Ceftriaxone: Rho=0.58; P=0.047 3. In ref. 3, no association was found between rates of antibiotic prescribing and geometric mean MICs. Perhaps the results of the paper under review are affected by reverse causality, namely places with higher prevalence of drug resistance for gonorrhea, as well as syphilis in MSM may initiate more STI screening efforts. For example, the California sites had both high geometric mean MIC for ciprofloxacin in 2005 (ref. 3), and high frequency of screening (see also1). Currently, this possibility is not mentioned in the Discussion. The reviewer was also wondering if there is a correlation between the fold change in screening frequency and the fold change in geometric mean MICs between 2005 and 2013. Reply:Thank you for this interesting suggestion which has been added to the discussion.4. Are other variables in NHBS-MSM surveys (possibly rate of change of sexual partners, or a number of recent sexual partners) correlated with geometric mean MICs/changes in thereof? Reply:Whilst we agree this would be interesting to investigate, we did not assess these correlations in this small study. 5. In the Discussion, it is mentioned that “the sample size was small (9 or 12)”. Currently, only 9 cities are used in the correlation analysis, as far as the reviewer could see. Reply:The sample size for 2005 was 9 and therefore in Figure 1, where one of the variables is from 2005, only 9 data points can be used. There were however 12 data points for the other 2 surveys and these were used in the analyses using these surveys. References: 1. Kirkcaldy RD, Bartoces MG, Soge OO, Riedel S, Kubin G, Del Rio C, et al. Antimicrobial Drug Prescription and Neisseria gonorrhoeae Susceptibility, United States, 2005-2013. Emerg Infect Dis. 2017;23(10):1657-63. Epub 2017/09/21. doi: 10.3201/eid2310.170488. PubMed PMID: 28930001; PubMed Central PMCID: PMCPMC5621530." } ] }, { "id": "38243", "date": "24 Sep 2018", "name": "Ellen Stobberingh", "expertise": [ "Reviewer Expertise My expertise includes microbiology", "bacteriology", "and antimicrobial susceptibility" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript describes the association between STI screening and antimicrobial resistance development of N. gonorrhoeae. Although of interest there are several questions which need to be answered:\nMaterial and Methods:\nThe number of participants reduced over time: 10,030 in 2005 and 8,012 in 2011. Also there was a difference in participation rate in the different cities.\n\nWhat is the influence of the decreased participation rate over time and the variation in the participation in the different cities on the interpretation of the data?\n\nIn 2008 (only syphilis) and in the two other years in addition N. gonorrhoeae, other STI or Chlamydia were tested. What was the reason for this difference?\n\nHow many N. gonorrhoeae isolates were included in the different years?\n\nWas the microbiological method to isolate, identify and antibiotic susceptibility testing over testing similar over time? Which method was used for susceptibility testing?\n\nPlease provide range and GM MIC values of the three antibiotica testing in the different years,\n\nFigure 1:\nThe horizontal axis mentioned an increase in MIC. What was the reference MIC?\n\nDicussion:\nWas there a change in therapy over time among the participants. All three antibiotica are mentioned in the guidelines, but is there information concerning the therapy prescribed over time in the different cities / participants?\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate? I cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4008", "date": "27 Sep 2018", "name": "Chris Kenyon", "role": "Author Response", "response": "The manuscript describes the association between STI screening and antimicrobial resistance development of N. gonorrhoeae. Although of interest there are several questions which need to be answered:Material and Methods: The number of participants reduced over time: 10,030 in 2005 and 8,012 in 2011. Also there was a difference in participation rate in the different cities. What is the influence of the decreased participation rate over time and the variation in the participation in the different cities on the interpretation of the data? Reply:Indeed both the decline in the absolute number of participants and the variation by city could introduce biases. This additional limitation has been added to the discussion. In 2008 (only syphilis) and in the two other years in addition N. gonorrhoeae, other STI or Chlamydia were tested. What was the reason for this difference? Reply:These differences reflect differences in the questions asked in the various surveys. In 2008, for example, respondents were asked if they had been tested for syphilis in the preceding 12 months, whereas in the 2011 survey they were asked if they had been tested for gonorrhea, chlamydia or syphilis in the previous 12 months. This is made clear in the methods section. How many N. gonorrhoeae isolates were included in the different years? Reply:Between 2005 and 2013, 44,144 isolates were tested. The report does not detail the breakdown of number tested by year. Was the microbiological method to isolate, identify and antibiotic susceptibility testing over testing similar over time? Which method was used for susceptibility testing? Reply:There were small changes in the laboratory protocol used. The protocol used and these changes are detailed in the Kirkcaldy et al report as follows [1]:\"Gonococcal isolates collected at each sentinelclinic are subcultured at the clinic’s local public health laboratory on supplemented chocolate medium and frozen in Trypticase soy broth containing 20% glycerol. Isolates are shipped monthly to one of the regionalreference laboratories, where they are tested for Beta-lactamase production and susceptibility to azithromycin, penicillin, tetracycline, ciprofloxacin, spectinomycin, cefixime, and ceftriaxone using the agar dilution method. Isolates were inoculated on Difco GC medium base supplemented with1% IsoVitaleX enrichment (Becton-Dickinson Diagnostic Systems,Sparks, MD). During 2005 to 2007, the lowest azithromycin concentrationtested was 0.008 ug/ml; this increased to 0.03 ug/ml in 2008. The routine testing range for azithromycin extended to 16.0 ug/ml during2005 to 2013. Laboratories were asked to conduct agar dilution testing to identify an endpoint for isolates with an MIC of 16.0ug/ml on the initialtesting run. Testing to an endpoint was not conducted on three isolatescollected during 2005 to 2013. In the absence of Clinical and LaboratoryStandards Institute (CLSI) breakpoints for gonococcal azithromycin susceptibility or resistance, we defined reduced azithromycin susceptibility for this analysis as an MIC of 2.0 ug/ml. Quality assurance processes are described in detail in the GISP protocol.To ensure accuracy and reproducibility of antimicrobial susceptibilityresults from the regional reference laboratories, a set of seven control N. gonorrhoeae strains with known MICs of various antimicrobials are included with each susceptibility run. In addition, reference laboratories test a CDC-provided panel of 15 unidentified strains twice yearly to compare results and ensure consistency among laboratories. The results obtained from the testing of control strains and CDC-provided panels are used for internal quality assurance.\" Please provide range and GM MIC values of the three antibiotica testing in the different years, Reply:This data is not provided in the paper that reports the GISP results [1]. In the online Technical Appendix file of this paper the following data is however reported: https://doi.org/10.3201/eid2310.170488 The geometric MIC of cefixime (Table 1), ceftriaxone (Table 2) and azithromycin (Table 3) by site and year. Table 8 presents the median and interdecile range of the 3 antimicrobials by site.Figure 1: The horizontal axis mentioned an increase in MIC. What was the reference MIC? Reply:The horizontal axis is the change in MIC between 2007 and 2013.Discussion: Was there a change in therapy over time among the participants. All three antibiotica are mentioned in the guidelines, but is there information concerning the therapy prescribed over time in the different cities / participants? Reply:This information is not provided in the report.References: 1. Kirkcaldy RD, Bartoces MG, Soge OO, Riedel S, Kubin G, Del Rio C, et al. Antimicrobial Drug Prescription and Neisseria gonorrhoeae Susceptibility, United States, 2005-2013. Emerg Infect Dis. 2017;23(10):1657-63. Epub 2017/09/21. doi: 10.3201/eid2310.170488. PubMed PMID: 28930001; PubMed Central PMCID: PMCPMC5621530" } ] } ]	1	https://f1000research.com/articles/7-1237
https://f1000research.com/articles/7-1729/v1	31 Oct 18	{ "type": "Research Article", "title": "Vacuum evaporation and nitrogen-assisted deodorization affects the antioxidant capacity in the olein fraction of red palm oil and its emulsion products", "authors": [ "Bohari Bohari", "Muhammad Muhadir", "Anton Rahmadi", "Bohari Bohari", "Muhammad Muhadir" ], "abstract": "Background: Deodorization of the olein fraction of red palm oil (OFRP) determines not only the taste of a multivitamin emulsion but also its antioxidant capacity. The emulsion product was formulated from OFRP, pumpkin juice (PJ), and dragon fruit juice (DFJ). This study aimed to optimize vacuum evaporation and nitrogen-assisted deodorizations of OFRP, observing levels of β-carotene, α-tocopherol, inhibition percentage of ABTS reduction, and ferric reducing antioxidant power (FRAP) activity. Methods: The deodorizations observed were vacuum evaporation in four conditions: (1) 90°C, 80±5 mmHg, (2) 100°C, 80±5 mmHg, (3) 90°C, 100±5 mmHg, (4) 100°C, 100±5 mmHg, and nitrogen-assisted in two flow durations: (1) 15 min and (2) 30 min. β-carotene, α-tocopherol, and butylated hydroxytoluene (BHT) were employed as standards. Results: The deodorized OFRP had fewer than 2% free fatty acids (FFA), lower than 3% peroxide value (PV), and lower than 4% acidic value (AV). Fluctuations of the β-carotene and α-tocopherol concentrations were observed in the deodorized OFRP. The final emulsion product had β-carotene of 259.9±1.4 to 271.7±2.4 ppm and α-tocopherol of 36.36±0.20 to 39.12±0.20 ppm. The total betacyanin of the emulsions were ±25% than DFJ. The emulsions had 22.93 to 32.11% of ABTS reduction inhibitory activity of the BHT activity and FRAP activity of 16.54±0.19 to 17.69±0.67 mM FeSO4•7H2O. Conclusions: The best vacuum evaporation optimized at 90 °C, 100±5 mmHg, 60 RPM for 1 hour. The best nitrogen-assisted deodorization was at 85±3°C and 1 l/minute of nitrogen for 15 minutes. The deodorization process affected the antioxidant activity of OFRP and emulsions.", "keywords": [ "α-tocopherol", "antioxidant capacity", "β-carotene", "deodorization", "olein fraction of red palm oil" ], "content": "Introduction\n\nThe attempt to produce multivitamins locally sourced from the emulsion of the olein fraction of red palm oil (OFRP), pumpkin juice (PJ), and dragon fruit juice (DFJ) was hampered by the strong odor and aftertaste of OFRP1. The stripping of vapor components from OFRP is theoretically achievable through heating at low pressure or flowing inert gas as a carrier of the evaporated odor compounds2. However, application of heat to reduce the off-taste compounds in OFRP may result in the formation of trans-fatty acids and the occurrence of lipid oxidation. Oxidation is a major concern to edible oil quality, deteriorating its chemical, sensory and nutritional properties. The application of heat is also limited the desire to conserve the minor antioxidant components (e.g. tocopherols compounds)3.\n\nMaintaining healthy contents of natural antioxidants and poly-unsaturated fatty acids (PUFA) are key indicators of the quality of refined and deodorized edible oils4. In this regard, natural antioxidants protect oils from hydrolysis, oxidation, polymerization, isomerization, and cyclization5. Previous deodorization efforts to reduce the off odor of red palm oil resulted in α-tocopherol being a better marker for maintaining overall antioxidant contents. An important finding was the need to maintain the critical time of the deodorization, which the destruction of tocol groups of the OFRP was in linear correlation to the deodorization time6.\n\nVacuum evaporation and steam-assisted deodorization are common practices in oil refinery. Applying vacuum pressures of 10-5 to 1 mbar enables heat application to reach 300°C in edible oil processing7. Steam-assisted deodorization requires good contact between the edible oils and the steam as stripping medium4. A nitrogen medium is preferable in comparison to steam medium, due to the inert property of the gas and the effectiveness of thermal breakdown of off-taste precursors in the presence of the nitrogen carrier inside the deodorization chamber8. Based on various application reports9–12, both methods are capable of deodorizing off-odor compounds in the production of better-tasting and rich-in-antioxidant OFRP.\n\nConcerning the challenges of OFRP deodorization, this stage should be revisited and further developed to produce a higher antioxidant capacity in the final OFRP, PJ, and DFJ emulsion products. This study aims to optimize the vacuum evaporation and nitrogen-assisted deodorization method of OFRP, observed using the levels of β-carotene, α-tocopherol, percentage inhibition of ABTS reduction, and ferric reducing antioxidant power (FRAP) activity.\n\n\nMethods\n\nThe crude palm oil (CPO) was obtained from PT Rea Kaltim, Indonesia. The olein fraction of red palm oil (OFRP) was prepared from CPO which had been tested for free fatty acid (FFA) levels and then carried out neutralization and gum removal by adding 100 ml warm water (80–90°C). Shuffling of the sample was carried out for 1 minute in a separating funnel before removal of residual water. About 400 μl of 10% NaOH (Emsure® Merck Cat no. 106498, USA) was added. Rinsing with 50 ml of warm water was carried out repeatedly until two phases of mixture were formed. The OFRP was produced with yields ranging from 60 to 70% of CPO (Rahmadi et al., 2015).\n\nDeodorization of the OFRP was prepared by two methods, namely vacuum and nitrogen-assisted evaporation. The liquid fraction was deodorized using a rotary evaporator (Büchi R-200, Switzerland) at a combination of temperature and pressure: (1) olein fraction processed with vacuum evaporation 1 (OFV1): 90°C, 80±5 mmHg, (2) OFV2: 100°C, 80±5 mmHg, (3) OFV3: 90°C, 100±5 mmHg, (4) OFV4: 100°C, 100±5 mmHg, and speed of 60 RPM for 1 hour. The OFVs that has been obtained was then stored in a closed container in the refrigerator (4±2°C) for further processing. Deodorization of the olein fraction by the nitrogen-assisted method was prepared by addition of 100 ml OFRP in a closed container. The sample had added to it 2% (w/v) pharmaceutical-grade activated carbon (Norit, Japan), then pure-grade nitrogen gas (Samator Gas, Indonesia) was flowed using a valve for 15 minutes (OFN1) and 30 minutes (OFN2) at a temperature of 85±3ºC, and a flow rate of 1 l/minute.\n\nPreparation of pumpkin and red dragon fruit juice was carried out according to previously described method13. A total of 1 kg of each peeled fruit was cut into pieces of approximately 3–5 cm3 and washed using clean water. Then, the pieces of fruit were mashed using a commercial juicer. After that, the juice was put separately in glass bottles and pasteurized at 80°C for 10 minutes, then filtered using a clean filter cloth. The filtered fruit juices were then stored at 4±2°C before being utilized.\n\nA total of 30 ml of the deodorized OFRP was added to 70 ml pumpkin juice, giving a total volume of 100 ml. Food grade carboxymethylcellulose, xhantan gum, and cinnamon powder were added at concentrations of 2% (w/v), 2% (w/v), and 0.5% (w/v), respectively. After homogenization, the product samples had added to them 75% (v/v) of red dragon fruit juice in warm water (80–90°C) to a volume of 400 ml, then fructose syrup sweetener, citric acid and raspberry flavoring were added at 10% (v/v) and 0.25% (w/v), and 0.7% (v/v), respectively. The mixture was homogenized with a blender at low speed for 3 minutes. The samples were filtered and placed in a dark glass bottle. The bottles containing the emulsion product were then pasteurized at 70°C for 15 minutes.\n\nFFA levels, peroxide values (PV), and acidic values (AV) were determined as previously described14.\n\nCarotenoids were measured via measurement of β-carotene by modification of the Palm Oil Research Institute (PORIM) method15, with changes in wavelength 446 to 443 nm (Rayleigh model UV 2601, China) and solvent changed from n-hexane to absolute methanol (Fulltime cat. 6501-04, China). Quantitative determination of β-carotene levels in samples was obtained based on calibration curves with β-carotene standards (Sigma-Aldrich cat. no. C9750-10G, UK).\n\nThe determination of α-tocopherol was accomplished by interpolating the absorbance of the sample with a standard calibration curve of α-tocopherol at a wavelength of 291 nm (Rayleigh model UV 2601, China). The α-tocopherol standard (Sigma-Aldrich cat. no. T3251-25G, UK) was prepared in absolute ethanol (Smartlab cat. no. A1035, Indonesia) with various concentrations of 50. 75. 100. 125 and 150 mg/L in 10 mL of absolute ethanol. Blank sample was prepared without containing the active substance of α-tocopherol.\n\nTotal betacyanin quantification was carried out as described16, and calculated using established equation17. A total of 1 g of sample was macerated with 5% HCl solution (Merck cat. no. 109063, USA) (1:10) in a dark bottle, then stored at 4±2°C for 24 hours. The mixture was vacuum-filtered with Whattman's filter paper no. 4. Next, 5 ml filtrate was diluted with 95% ethanol solution: 1.5 N HCl (85:15) to 10 ml. The absorbance was measured at a wavelength of 535 nm (Rayleigh model UV2601, China).\n\nABTS. For the preparation of antioxidant measurement by the ABTS method, solution A was prepared by dissolving 7.1015 mg of 2.2'-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid) (ABTS) (Sigma-Aldrich cat. no. A1888-1G, UK) in 5 ml distilled H2O, and solution B was prepared by dissolving 3.500 mg K2S2O8 (Sigma-Aldrich cat. no. 216224-100G, UK) in 5 ml of distilled H2O. Solution A and B were separately incubated in dark bottles for 12 hours. After incubation, solution A and B were mixed in a dark room and added with absolute ethanol to produce 25 ml of the final solution. Measurement of absorption of ABTS blank solution was carried out at a wavelength of 750 nm (Eppendorf Single Beam BioSpectrometer®, Germany). Measurement of free radical-binding activity was carried out by mixing the emulsion sample with ABTS solution at a concentration of 100 ppm. The sample was measured at maximum absorbance wavelength. butylated hydroxytoluene (BHT) was used as a standard (Sigma-Aldrich cat. no. W218405, UK). The percentage of antioxidant activity was determined based on the calculation of the difference between the blank and the sample absorbance divided by the blank absorbance.\n\nFRAP. FRAP reagent consists of 0.1 M acetate buffer (pH 3.6) (Merck cat. no. 107827, USA), 10 mM 2.4.6-tripydyl-s-triazine (TPTZ) (Sigma-Aldrich cat. no. 93285-5G, UK), and 20 mM FeCl3•6H2O (Sigma-Aldrich cat. no. 157740-100G, UK) in a ratio of 10:1:1 in 40 mM of HCl. Standard solutions of FeSO4•7H2O (Sigma-Aldrich cat. no. 215422, UK) were prepared at concentrations of 10, 20, 30, 40 and 50 µmol/l. The maximum absorbance of FeSO4•7H2O was determined, which was at wavelength of 594-597 nm (Eppendorf Single Beam BioSpectrometer®, Germany). A total of 0.1 ml emulsion sample was prepared and added to 3 ml FRAP reagent in a test tube. The absorbance of the mixture was then read at 574 nm. Quantitative determination of total antioxidants in the sample was obtained based on calibration curve of FeSO4•7H2O.\n\n\nResults and discussion\n\nFFA indicates the palatability acceptance of the OFRP and its emulsion-derived products; a higher presence of FFA contents results in a more pronounced sharp taste. FFA standard for fresh palm oil is less than 2%, while maximum FFA content for processed palm oil is set to 5%18. The vacuum evaporation of OFRP produced a higher concentration of FFA in comparison to nitrogen-assisted deodorization. However, the final products had less than 2% of FFA (Figure 1).\n\nPV indicates the presence of hydrogen peroxide in edible oils. The standards for the maximum PV in edible oils are determined based on the product and the processing technology used19. For example, the standard PV for virgin olive oil is not more than 20 mEq O2/Kg, while for processed olive oil is not more than 10 mEq O2/Kg20. Initial OFRP had 1.45±0.35 mEq O2/Kg. Neither vacuum-evaporated nor nitrogen-assisted deodorized OFRP had PV of more than 3 mEq O2/Kg. The final emulsion products had PV of less than 2 mEq O2/Kg (Figure 1).\n\nOFRP and the derived emulsion products had an AV of less than 4 mEq KOH/g (Figure 1). Pure palm oil had AV on average of 10 mEq KOH/g. in acidic environment21. Crude Palm Oil (CPO) produced from small scale palm oil mills had an average AV of 18 mEq KOH/g22,23. From the two deodorization approach, in terms of values of FFA, PV, and AV, nitrogen-assisted deodorized OFRP emerged as the better deodorization process.\n\nThe PORIM method15 allows for quick determination of the β-carotene content in CPO. Based on the procedure and to ensure high repeatability, it is necessary to recheck the peak wavelength of β-carotene standard, which was at 443 nm, and to generate a β-carotene standard curve (Figure 2a). The contents of β-carotene in OFRP, PJ, DFJ and derived products are given in Table 1. The relative contents of β-carotene to OFRP were highlighted in Figure 2b. As expected, DFJ had the lowest percentage of β-carotene in comparison to OFRP, while the content of β-carotene in PJ was 60% of the content in OFRP. The β-carotene contents for the emulsion made using vacuum-deodorized OFRP (EV) was 271.7±2.4 ppm, while 259.9±1.4 ppm of β-carotene was found in the emulsion made of nitrogen-assisted deodorized OFRP (EN).\n\nOFRP, olein fraction of red palm oil; DFJ, dragon fruit juice; PJ, pumpkin juice; OFV1, OFRP treated at 90°C, 80±5 mmHg, 60 RPM, for 1 hour; OFV2, OFRP treated at 90°C, 100±5 mmHg, 60 RPM, for 1 hour; OFV3, OFRP treated at 100°C, 80±5 mmHg, 60 RPM, for 1 hour; OFV4, OFRP treated at 100°C, 100±5 mmHg, 60 RPM, for 1 hour; OFN1, 85±3 °C, flow rate of 1 l/minute of nitrogen for 15 minutes; OFN2, 85±3°C, flow rate of 1 L/minute of nitrogen for 30 minutes; EV, emulsion made of OFV3. PJ and DFJ; EN, emulsion made of OFN1. PJ and DFJ; n.t., not tested.\n\nThe use of higher temperatures in vacuum evaporation (OFV2 and OFV4) slightly reduced β-carotene content in OFRP. A temperature increase without an increase in head pressure resulted in a lower accumulation of β-carotene, as observed in OFV4 vs OFV2 and OFV3 vs OFV1. This phenomenon is in line with that previously reported24, stating that a temperature increase while reducing head pressure resulted in lower efficiency of soybean oil neutralization and distillation. Nitrogen-assisted deodorization helped to concentrate β-carotene content in OFRP. The emulsion products contained around 50% of β-carotene in OFRP. Nitrogen was used as a stripping medium substitute for steam, owing to its inert and easy-to-remove properties of non-reacting. This highlights the possibility of neutralizing distillation by stripping with nitrogen in physical refining to obtain refined oils within specified end product's acidity and flavour24.\n\nTo provide a repeatable measurement of the content of α-tocopherol, peak wavelength of α-tocopherol standard was rechecked and occurred at 291 nm and α-tocopherol standard curve was produced as in Figure 3a. Table 1 displays the α-tocopherol content in OFRP, PJ, DFJ and derived products, while Figure 3b highlights the relative contents of α-tocopherol to OFRP. All three sources (OFRP, DFJ and PJ) contained a reasonable concentration of α-tocopherol, (70.61±0.59, 37.02±0.33 and 32.95±0.04 ppm, respectively). The final emulsion products contained between 36.36±0.20 and 39.12±0.20 ppm of α-tocopherol.\n\nPreviously, we reported that processing time was critical to maintain α-tocopherol levels in OFRP deodorization6. While the deodorization time was fixed at 1 hour, a combination of temperature and vacuum pressure (OFV1 to OFV4) had slight effect in reducing the content of α-tocopherol in OFRP The phenomenon was similar to that previously reported24, which found that a 10°C temperature increase resulted in slightly reduced α-tocopherol content in soybean oil neutralization and distillation. The best treatment for vacuum deodorization was at 100 °C and 100±5 mmHg (OFV3).\n\nThe presence of betacyanin in dragon fruit has led to the use of the color for dye in various applications, from food to solar cells16,25. DFJ juice contains 2.18 ppm of betacyanin, while the contents of betacyanin in the emulsion products were 22–26% of the content in DFJ (Figure 4). This indicated that the function of DFJ was to provide pleasant color while masking the after taste of OFRP in the final emulsion product1.\n\nThe antioxidant activity of OFRP, PJ, DFJ, and the derived emulsion products were estimated with ABTS and FRAP assays (Figure 5). The two methods were suitable to measure antioxidant activity in oil based products26. Products containing high concentrations of carotenes, xanthophylls and tocopherols are expected to have moderate-to-strong antioxidant potency, comparable to BHT. In comparison to BHT, the emulsion products had 22.93±0.10 to 32.11±0.04% of free-radical-scavenging activity, as measured by ABTS method. There was no significant difference of percentages of antioxidant activity of OFRP, PJ, DFJ, and the derived emulsion products against the BHT standard. Based on the ABTS and FRAP assays, it was concluded that OFRP, PJ, DFJ, and the derived emulsion products exhibited moderate-to-strong antioxidant activity.\n\nWhile acting as a putative chain-breaking antioxidant, β-carotene was capable of scavenging peroxyl radicals27. Tocopherols are strong antioxidants, but their activity depends on the isomers, reactivity of the tocopheryl radical, surrounding temperature, and the type and viscosity of emulsion28. Oil-based products may contain secondary antioxidants, due to their greater ability to reduce lipid oxidation than acting as free radical scavengers26.\n\n\nConclusion\n\nThe vacuum evaporation of OFRP produced slightly higher FFA and AV in comparison to nitrogen-assisted deodorization, while the final products had less than 2% of FFA, less than 3 mEq O2/Kg of PV, and less than 4 mEq KOH/g of AV. The β-carotene contents for the emulsion containing vacuum-deodorized OFRP was at 259.9±1.4 and the nitrogen-assisted deodorized OFRP was 271.7±2.4 ppm; these values were 51 and 53% of the β-carotene values in the initial OFRP, respectively. The use of higher temperatures in vacuum evaporation slightly reduced the β-carotene content in OFRP, while nitrogen-assisted deodorization concentrated β-carotene content in OFRP. All three sources (OFRP, DFJ, PJ) contained reasonable concentrations of α-tocopherol (70.61±0.59, 37.02±0.33, and 32.95±0.04 ppm, respectively), while the final emulsion products contained 36.36±0.20 to 39.12±0.20 ppm of α-tocopherol. The main purpose of DFJ was to provide pleasant color while masking the after taste of OFRP in the final emulsion product. OFRP, PJ, DFJ, and the derived emulsion products exhibited moderate-to-strong antioxidant activity. The best vacuum evaporation condition was at 100°C, 80±5 mmHg, 60 RPM, for 1 hour, while the best nitrogen-assisted conditions were 85±3°C with flow rate of 1 l/minute of nitrogen for 15 minutes.\n\n\nData availability\n\nDataset 1. All data on the properties of the emulsions produced in the current study. Included in labelled files are all raw data for each variable measured in this study, alongside the processed data. DOI: https://doi.org/10.5256/f1000research.16545.d22173829.", "appendix": "Grant information\n\nAuthors thank the Indonesian Ministry of Research, Technology, and Higher Education, which funded this research in 2018.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nRahmadi A, Setionugroho IB, Yuliani Y, et al.: Dragon Fruit Juice Addition in Palm Oil-Pumpkin Emulsion: Panelist Acceptance and Antioxidant Capacity. Jurnal Teknologi dan Industri Pangan. 2017; 28(2): 122–128. Publisher Full Text\n\nDijkstra AJ: Vacuum stripping of oils and fats. In: Gunstone, F.D., Harwood, J.L., & Dijkstra, A.J. (eds.). The Lipid Handbook, 3rd Edition. Taylor & Francis Group LLC, Boca Raton, FL, 2007; 235–253.\n\nDe Greyt WFJ: Edible oil refining: Current and future technologies. In: W. Hamm, RJ. Hamilton and GH. Calliauw (eds.). Edible Oil Processing. John Wiley & Sons, Chichester, 2013; 127–151. Publisher Full Text\n\nKellens M, De Suray D: Method for vacuum stripping of oils and fats.U.S. Patent No. 7.670.634. Washington, DC: U.S. Patent and Trademark Office. 2010. Reference Source\n\nAlireza S, Tan CP, Hamed M, et al.: Effect of frying process on fatty acid composition and iodine value of selected vegetable oils and their blends. International Food Research Journal. 2010; 17(2): 295–302. Reference Source\n\nRahmadi A, Yanti I, Jannah SM, et al.: Quantitation and Optimation of β-Carotene and α-Tocopherol in Emulsion Prototype with Reversed Phase Chromatography. IJASEIT [submitted]. 2018.\n\nSan Ho DS: Recovery of phytonutrients from oils. U.S. Patent No. 7.544.822. Washington, DC: U.S. Patent and Trademark Office. 2009. Reference Source\n\nHam JS, Kim HY, Lim ST: Antioxidant and deodorizing activities of phenolic components in chestnut inner shell extracts. Ind Crops Prod. 2015; 73: 99–105. Publisher Full Text\n\nKonsoula Z, Liakopoulou-Kyriakides M: Effect of endogenous antioxidants of sesame seeds and sesame oil to the thermal stability of edible vegetable oils. LWT-Food Sci Technol. 2010; 43(9): 1379–1386. Publisher Full Text\n\nCuellar-Bermúdez SP, Barba-Davila B, Serna-Saldivar SO, et al.: Deodorization of Arthrospira platensis biomass for further scale‐up food applications. J Sci Food Agric. 2017; 97(15): 5123–5130. PubMed Abstract \| Publisher Full Text\n\nChen D, Chen X, Chen H, et al.: Identification of odor volatile compounds and deodorization of Paphia undulata enzymatic hydrolysate. J Ocean U China. 2016; 15(6): 1101–1110. Publisher Full Text\n\nAllaf T, Tomao V, Ruiz K, et al.: Deodorization by instant controlled pressure drop autovaporization of rosemary leaves prior to solvent extraction of antioxidants. LWT-Food Sci Technol. 2013; 51(1): 111–119. Publisher Full Text\n\nRahmadi A, Puspita Y, Agustin S, et al.: Penerimaan Panelis dan Sifat Kimiawi Emulsi Labu Kuning dan Fraksi Olein Sawit. Jurnal Teknologi dan Industri Pangan. 2015; 26(2): 201–212. Publisher Full Text\n\nAndarwulan N, Kusnandar F, Herawati D: Analisis Pangan. Dian Rakyat. Jakarta. 2011; 1–41.\n\n[PORIM] Palm Oil Research Institute of Malaysia: Test Methods Carotene Content. Palm Oil Research Institute of Malaysia, Kuala Lumpur. Malaysia. 1995.\n\nWoo KK, Ngou FH, Ngo LS, et al.: Stability of betalain pigment from red dragon fruit (Hylocereus polyrhizus). Am J Food Technol. 2011; 6(2): 140–148. Publisher Full Text\n\nInácio MR, de Lima KM, Lopes VG, et al.: Total anthocyanin content determination in intact açaí (Euterpe oleracea Mart.) and palmitero-juçara (Euterpe edulis Mart.) fruit using near infrared spectroscopy (NIR) and multivariate calibration. Food Chem. 2013; 136(3–4): 1160–1164. PubMed Abstract \| Publisher Full Text\n\nTagoe SMA, Dickinson MJ, Apetorgbor MM: Factors influencing quality of palm oil produced at the cottage industry level in Ghana. International Food Research Journal. 2012; 19(1): 271–278. Reference Source\n\nEldin AK: Methods to determine the extent of lipid oxidation in foods. In: Decker EA, Elias RJ, & McClements DJ. (eds) Oxidation in Foods and Beverages and Antioxidant Applications: Understanding Mechanisms of Oxidation and Antioxidant Activity. Woodhead Publishing. 2010; 181–195. Publisher Full Text\n\nMatthäus B: Oxidation of edible oils. In: Decker EA, Elias RJ, & McClements DJ. (eds) Oxidation in Foods and Beverages and Antioxidant Applications: Understanding Mechanisms of Oxidation and Antioxidant Activity. Woodhead Publishing. 2010; 183–238. Reference Source\n\nHayyan A, Mjalli F, Mirghani M, et al.: Treatment of acidic palm oil for fatty acid methyl esters production. Chem Pap. 2012; 66(1): 39–46. Publisher Full Text\n\nEkop SA, Etuk BA, Eddy NO: Effect of some local additives on the chemical constituent of palm oil. Journal of Applied Sciences and Environmental Management. 2007; 11(1): 85–89. Publisher Full Text\n\nChang AS, Sherazi STH, Kandhro AA, et al.: Characterization of Palm Fatty Acid Distillate of Different Oil Processing Industries of Pakistan. Journal of Oleo Science. 2016; 65(11): 897–901. PubMed Abstract \| Publisher Full Text\n\nGraciani-Constante E, Rodriguez-Berbel F, Paredes-Torronteras A, et al.: Deacidification by distillation using nitrogen as stripper: Possible application to the refining of edible fats. Grasas y Aceites. 1991; 42(4): 286–292. Reference Source\n\nAli RAM, Nayan N: Fabrication and analysis of dye-sensitized solar cell using natural dye extracted from dragon fruit. International Journal of Integrated Engineering. 2010; 2(3): 55–60. Reference Source\n\nChristodouleas DC, Fotakis C, Nikokavoura A, et al.: Modified DPPH and ABTS assays to assess the antioxidant profile of untreated oils. Food Anal Methods. 2015; 8(5): 1294–1302. Publisher Full Text\n\nKiokias S, Varzakas T, Oreopoulou V: In vitro activity of vitamins, flavonoids, and natural phenolic antioxidants against the oxidative deterioration of oil-based systems. Critical reviews in food science and nutrition. 2008; 48(1): 78–93. PubMed Abstract \| Publisher Full Text\n\nKamal-Eldin A, Appelwist LA: The chemistry and antioxidant properties of tocopherols and tocotrienols. Lipids. 1996; 31(7): 671–675. PubMed Abstract \| Publisher Full Text\n\nBohari B, Muhadir M, Rahmadi A: Dataset 1 in: Vacuum evaporation and nitrogen-assisted deodorization affects the antioxidant capacity in the olein fraction of red palm oil and its emulsion products. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16545.d221738" }	[ { "id": "56331", "date": "17 Jan 2020", "name": "Azis Boing Sitanggang", "expertise": [ "Reviewer Expertise food process engineering", "enzymatic membrane reactor", "bioprocess engineering" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe statistical analysis and its appropriate interpretation must be performed on the results to see significant differences between treatments (Fig 1-Fig 5)\n\nThe standard deviations especially in Fig 1 is really high (approx. 50% of the avg. value). By this, the data reproducibility might be questionable.\n\nThe authors should be consistent in presenting the figures. Sometimes the bars are accompanied with standard deviations, and without labels (numbers) on top of them, but sometimes without.\n\nThe drawn conclusion seems to be correct and stems for the discussion being presented. However, if the discussion accompanied with statistical analysis, this would make the study more scientifically sound.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] }, { "id": "61274", "date": "17 Apr 2020", "name": "Maria Liakopoulou-Kyriakides", "expertise": [ "Reviewer Expertise medicinal chemistry", "Natural products chemistry and engineering", "food and environmental biotechnology", "microbial biotechnology" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article reports on optimisation of deodorization methods for the olein fraction of red palm oil by using vacuum evaporation and vacuum evaporation nitrogen- assisted. It is a piece of interesting work, the authors though are focused on the application of the above methods without giving the state of the art in deodorization methods applied so far and what are the advantages of the ones they are reporting. They should add something in the introduction. The conclusions must be rewritten and commented based on their results and efficiency of the methods. It seems that Nitrogen-assisted vacuum evaporation at 85 0C is slitghtly better in terms of b-carotene, tokopherol and all the other analytical data compared to solely vacuum evaporation at higher temperature, which is expected. Antioxidant activity of the final emulsion and content in b-carotene etc are not very encouraging in case of application. They should discuss it. The paper can be accepted after minor revision.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/7-1729
https://f1000research.com/articles/7-1722/v1	31 Oct 18	{ "type": "Research Article", "title": "Managing hundreds of improvement teams", "authors": [ "M Rashad Massoud", "Leighann E. Kimble", "Victor Boguslavsky", "Maina Boucar", "Jorge Hermida", "Donna Jacobs", "Esther Karamagi", "Nigel Livesley", "Mirwais Rahimzai", "M Rashad Massoud", "Victor Boguslavsky", "Maina Boucar", "Jorge Hermida", "Donna Jacobs", "Esther Karamagi", "Nigel Livesley", "Mirwais Rahimzai" ], "abstract": "Recognizing the notable scale of USAID Applying Science to Strengthen and Improve Systems (ASSIST) Project activities and sizable number of improvement teams, which in some cases is close to 1,000 improvement teams managed in one country at a point in time, we sought to answer the questions: How do we manage hundreds of improvement teams in one country alone? How do we manage more than 4,000 improvement teams globally? The leaders of our improvement programs manage such efforts as though they are second-nature, without pointing to the specific skills and strategies needed to manage thousands of teams. This paper was developed to capture the lessons, considerations, and insights shared in discussions with leaders on the USAID ASSIST Project, including country Chiefs of Party and Regional Directors. More specifically, this paper seeks to describe what is involved in scaling up and managing large numbers of improvement teams. Through focus group discussions and individual interviews, participants discussed the key skills, strategies, and lessons needed to successfully manage large numbers of teams on the USAID ASSIST Project. We concluded that the six key components in managing large numbers of teams are 1) leadership; 2) management structures and capacities; 3) clear and open communication; 4) shared learning, collaboration, and support; 5) ownership, engagement, and empowerment; and 6) partnerships. We further analyzed these six components as being interrelated to one another based on the relationship between culture, strategy, and technique in implementing quality improvement activities.", "keywords": [ "Leadership", "quality improvement", "improvement", "teams" ], "content": "Background\n\nThe USAID Applying Science to Strengthen and Improve Systems (ASSIST) Project is a USAID-funded global mechanism for improving healthcare in low- and middle-income countries. The work under the USAID ASSIST Project is the continuation of the efforts of predecessor projects, USAID Health Care Improvement (HCI) and Quality Assurance Project (QAP) I-III. Over the life of the project, the USAID ASSIST Project was working in 3,111 facilities and 2,006 communities, supporting 4,004 quality improvement teams working in these facilities and communities (Figure 1). These numbers include both facilities and communities that received both direct and indirect support from USAID ASSIST. Some of the teams were shared between the facilities and communities. In addition to the improvement teams working in the facilities and communities, the ASSIST Project worked with 159 government and implementing partners to improve health outcomes and strengthen health systems globally. Government partners included country Ministries of Health, Ministries of Social Affairs, Ministries of Gender, National AIDS and HIV Control Programs, and other governmental and parastatal institutions. In many countries, our partnerships also included district and provincial health management teams and leaders.\n\nThe estimated total number people served by the health facilities benefitting from the efforts of these QI teams was 112 million people. USAID ASSIST works with host country national and local level counterparts in providing technical assistance to meet their aims for improving healthcare outcomes in priority areas. This work is done by forming teams among the service delivery providers and managers in counterpart institutions to work on improving the quality of care delivered to patients. Over the life of the USAID ASSIST Project, USAID Country Missions and USAID Offices (such as Health Systems, HIV/AIDS, and Maternal, Child, Health, and Nutrition) have played an instrumental role in facilitating our ability to collaborate and coordinate with partners to expand the reach of our work.\n\nIn Uganda alone, at the end of June 2017, ASSIST was working with 928 quality improvement teams to strengthen the health system; prevent HIV; improve care and treatment for people with HIV; and improve family health related to maternal, newborn, and child health, family planning, nutrition, tuberculosis, and malaria. At the same time, in Tanzania, ASSIST was working with 781 quality improvement teams, including spread and scale-up of preventing mother-to-child transmission and antiretroviral therapy activities. Working with many teams is not unfamiliar to the USAID ASSIST Project. ASSIST and its predecessor projects, HCI and QAP, have supported thousands of different improvement teams to accomplish their improvement aims. The USAID series of quality improvement contract mechanisms have been at the heart of a global movement in taking successful improvements to scale1,2. It became evident that successful spread depends on the adaptation of key principles to context and using “hybrid models” based on lessons learned from successful scale-up efforts3.\n\nThe progression between the predecessor projects and the work on ASSIST today reflects the impact of successful scale-up of quality improvement. Starting with QAP II’s initial large-scale spread efforts in Tula and Tver in the Russian Federation4, the foundations for large-scale spread efforts were developed. These efforts were continually evolved over QAP III, HCI, and ASSIST. Today this is particularly the case in India, Uganda, and Tanzania, which each started off with a small number of teams. The success of improvement at a small scale in each of these countries gradually spread to other health centers and districts throughout the country, leading to a larger number of improvement teams able to reach a larger catchment area.\n\nA brief overview of the work in each of these countries as well as a summary on spread can be found in Supplementary File 1. More information can be found in the FY17 Country Reports for each country, as well as on the USAID ASSIST website (www.usaidassist.org).\n\n\nIntroduction\n\nBeyond the use of improvement methodology and spread methods in country activities, there are several interrelated actions and communications that must take place to not only manage current improvement activities, but to scale up improvement activities that are yielding positive results5. This paper was developed to capture the lessons, considerations, and insights shared in discussions with leaders on the USAID ASSIST Project, including country Chiefs of Party and Regional Directors. More specifically, this paper seeks to describe what is involved in scaling up and managing large numbers of teams6,7.\n\nRecognizing the notable scale of USAID ASSIST Project activities and sizable number of improvement teams, with in some cases close to 1,000 improvement teams at a point in time, we sought to answer the questions: How do we manage 100s of improvement teams in one country alone? How do we manage more than 4,000 improvement teams globally? The leaders of our improvement programs manage such efforts as though they are second-nature, without pointing to the specific skills and strategies needed to manage thousands of teams.\n\n\nMethods\n\nThe underlying objective for this research began with a question from the USAID ASSIST Project Agreement Officer’s Representative (AOR) in a USAID ASSIST Quarterly Review Meeting (QRM). In response to a report on the large number of improvement teams in Tanzania and Uganda on the USAID ASSIST Project, the AOR asked the question, “how do you manage 1000 teams?”.\n\nBecause this topic is specific to the work on the USAID ASSIST Project and involves a small number of individuals, there was no formal sampling method involved. The findings in this work are from the Project Director, Senior Improvement Advisors, Regional Directors, and Chiefs of Party that were available and willing to participate in the focus group discussion via Zoom videoconferencing, one-on-one interviews, and email discussions about the topic of managing large numbers of teams.\n\nThe insights analyzed in this paper were obtained through a series of discussions with the participants. Each of these individuals has many years of experience using quality improvement to yield improved health outcomes in low- and middle-income countries. The group formed encompasses all Regional Directors, Chiefs of Party, and Senior Improvement Advisors available to join the Zoom videoconference and that had experience managing hundreds of teams on the ASSIST Project, in addition to the Project Director. As the experiences are specific to the ASSIST Project, only participants with experience managing on the ASSIST Project were selected for participation in the group discussions and individual interviews. Initial discussions occurred over Zoom videoconferencing meetings and included the authors listed for this paper. The method of data collection began with the use of a larger focus group of participants to explore the topic of how the USAID ASSIST Project manages teams, including lessons learned from the experiences of the participants. The discussion was semi-structured, with the researcher asking questions only to probe deeper discussion or to obtain clarification. The opening question for the focus group stemmed from a previous USAID ASSIST QRM and was the broad question “How do we manage 1000 teams?”. Participants in the focus group and in the interviews and email discussion has the background for this question from attending the USAID ASSIST QRM.\n\nNotes were taken by interviewer throughout these calls, and insights from the meetings documented. The meeting minutes, which were initially a series of quotes from the meeting participants, were then circulated after the group meetings for additional feedback, ideas, and input regarding experiences managing improvement teams.\n\nAdditional one-on-one calls and in-person meetings were held following the group virtual conversations over Zoom to gain more detail from individual participants about what they found important in successfully managing teams in their experience on the USAID ASSIST Project, as well as in their broader experiences working in quality improvement. The input from these one-on-one meetings allowed the interviewer to get a more in-depth account of individual experiences based on the individual’s technical area of expertise, number of years of experience, and country/regional experience in improvement.\n\nThe interviewer for the focus group and individual interviews (LK) is an Improvement Advisor with University Research Co., LLC and the USAID ASSIST Project. Leighann holds a Master of Arts in International Relations and has worked with University Research Co., LLC and the USAID ASSIST Project for over four years. Her credentials and experience working with the USAID ASSIST Project provided her with the background needed to probe discussion and interview questions related to the work of the project. LK was also familiar with the participants, having had several interactions over the years with each of the participants. She has not directly managed large numbers of teams in the field, which allowed her to be objective in asking discussion and interview questions without interjecting her bias, thoughts, or experiences into the conversations with participants.\n\nData from the focus group discussion was initially recorded in the form of quotes from the participants, thereby transcribing the conversation. The interviewer, along with the Project Director, then found common themes between the quotes and categorized the quotes. The updated and categorized list of quotes was then circulated to the participant group to receive feedback, request clarification, and confirm that quotes were placed in the appropriate categories. Several iterations of this process were completed to revise, expand, and clarify each of the categories, while circulating the revisions for group feedback.\n\nThe feedback from the one-on-one meetings was combined with the input from the group discussions to find common themes and lessons learned from conversations on how to manage hundreds of teams. The synthesized ideas were circulated to all participants in the initial group discussions for additional feedback, clarification, and revisions. In the analysis to follow, quotes open each subsection, reflecting direct quotes from the conversations held with USAID ASSIST staff described above. The analysis is derived from a synthesis of inputs, feedback, several rounds of group discussion, one-on-one interviews, and rounds of revision to generate key ideas and lessons learned on how the USAID ASSIST Project has learned to manage hundreds of teams.\n\nOnce the categories were finalized, data analysis involved extracting the meaning behind the quotes in each of the categories and identifying the key lessons in each of the six categories as applied in USAID ASSIST Project activities. With feedback from the Project Director and further feedback from the larger group, the categories were further analyzed based on the theoretical framework of the Quality Management Triangle. Analysis was conducted based on the Quality Management Triangle to draw more depth to the results and conclusions of the discussions, particularly as they relate to the content of the six interrelated categories. Once finalized, the draft version of the paper was circulated to a smaller group on the USAID ASSIST Project for review and several rounds of revisions, prior to sending to USAID for further review and approval.\n\n\nAnalysis\n\nAs depicted in Figure 2, there are three central components for effective quality management: 1) a culture of managing for quality; 2) improvement technique; and 3) strategy for implementation. To effectively lead healthcare quality improvement, the culture, technique, and strategy for implementation must be complementary and adapted in consideration of the context. We highlight the interaction between culture, strategy, and technique due to the overlap between these three factors in quality improvement and in scaling up. As will be reflected throughout the paper, understanding the overlap between these factors and addressing each factor is important because the technique of quality improvement is only successful when applied using the appropriate strategy in the given culture.\n\nQuality improvement requires continual adaptation to context, which is facilitated through constant coordination, review, and feedback, each managed through six interrelated categories of action. When talking about improvement, people most often refer to the technique of applying quality improvement methods to achieve improved health outcomes. The technique encompasses testing and implementing changes to processes of care in a health system to yield improved outcomes. However, as depicted in Figure 2, the technique of quality improvement alone is not sufficient – the culture and strategy surrounding the technique are essential to achieve results. Culture refers to both internal and external factors (i.e., management structures, teamwork, communication environment, etc.) that may support or inhibit quality improvement in the given context. Changes may need to be made to culture to create an environment conducive to quality improvement. Strategy involves the actions needed to meet external needs, in this case improving global health priorities, while making changes to internal processes of care to yield improved outcomes.\n\nSuccess in quality improvement relies on integration within the culture, strategy, and technique specific to the context in which the improvement activity is taking place. This paper provides an overview of how each of these categories has become characteristic of our work, providing examples from activities throughout the USAID ASSIST Project.\n\nThrough discussion with our project leaders in country as well as at headquarters, we distilled the key components to managing our work successfully. These components fall into six interrelated categories:\n\n1. Leadership\n\n2. Management structures and capacities\n\n3. Clear and open communication\n\n4. Shared learning, collaboration, and support\n\n5. Ownership, engagement, and empowerment\n\n6. Partnerships\n\nThe connection between these six categories can be understood within the relationship between culture, strategy, and technique in quality improvement, as will be further described below.\n\nLeadership\n\n“Technical capability is not enough; leadership is needed [for cultural transformation].”\n\nAs highlighted in Table 1, leadership at all levels is key in managing and supporting improvement teams. Local leaders are the individuals that live and work in the communities and facilities in which changes are to be made to yield improvement. The role of the local leader is to support the improvement team in making changes to processes of care and problem-solve any issues that teams cannot fix at their level. Leaders at the district and national levels provide ongoing support and problem-solving beyond the boundaries of the specific facility or community. When multiple local teams are working on the same improvement activity in a district, the role of the district leader becomes more than just supportive – the district leader plays a key role in facilitating collaboration and knowledge sharing between the improvement teams within the district. Likewise, if the improvement team is happening on a national scale, the role of national leadership grows into one that supports collaboration and knowledge sharing between teams across multiple districts.\n\n° Requires a shift in power\n\n° Deliberate recognition of capabilities of individuals at\n\nall levels, not just at the top\n\nWhile ASSIST provides technical assistance in improvement, we recognize that the work of our improvement teams goes beyond building technical capability – leaders must have the capacity to lead improvement and must be engaged and committed to use their capabilities. Within our work, we find that supporting local leadership by creating both the capability and capacity for leading improvement is essential. On one hand, capability involves having the clinical and improvement knowledge and skill needed to improve healthcare results. However, capability on its own is not enough to translate into action. Leaders must be able to use these capabilities to lead improvement. Capacity involves leaders having the empowerment and confidence needed to apply their knowledge and make changes to the processes of care in their health system to yield better results.\n\nMuch of ASSIST’s work in leadership includes engaging leaders in host countries to support the work of the improvement teams and providing them with the knowledge and skills to do so. For example, we encourage leaders to allocate time and resources for teams to meet, participate in reviews of data, attend and present at learning sessions, chair non-ASSIST supported technical meetings where results were presented, facilitate decisions to institutionalize teams’ efforts, present results at regional, national or international meetings, etc. Our understanding of the key roles of leadership for leading improvement teams is highlighted in Box 1.\n\n\n\n1. Set clear goals for all staff.\n\n2. Give all staff the freedom and confidence to figure out how to do their work. Avoid micromanaging.\n\n3. Be available to help with problem solving if things are not going well. Help staff learn from their mistakes rather than blaming and shaming.\n\n4. Focus on the positives and actively look for success and new opportunities.\n\n5. Celebrate the successes and positives.\n\n6. Identify and address failures as system failures, and not individual failures.\n\nBy modeling and encouraging a “culture” that empowers and supports local leadership, ASSIST provides technical assistance to local leaders at various levels of the health system to build their confidence, empowerment, resources, and tools to help them lead local improvement teams. In addition to providing technical assistance related to capacity, or knowledge and “know-how”, for improvement, the resources we provide to leaders include problem-solving skills and ability to develop team dynamics to support improvement. With these skills and knowledge, we give local leaders the sense of accountability for leading improvement teams and reaching improvement aims and goals set by the team. While the overall strategy of quality improvement is to embed quality within the system, the “technique” involves using quality improvement to develop the capacity of individuals to lead the work over and above individual facilities. This too requires engagement and open communication with leaders to ensure they are prepared, capable, and empowered to lead quality improvement. The “strategy” of developing leadership involves providing local leaders the technical assistance they need to be technically capable. Beyond this is the component of empowerment and support of individuals to be capable as leaders accountable and empowered to sustain improvement independent of the project.\n\nManagement structures and capacities\n\n“Most of the successful events that we have had built on prior successes and utilize champions within it (whether or staff or in MOH). [This] requires meticulous management and does not happen without very serious attention to detail and a strong pattern of engagement.”\n\nFor an activity to operate properly, there must be a management structure. In the case of improvement activities, because of the complexity and engagement needed, the “technique” required is the ability to develop structures to support individual quality improvement teams and to allow people to convene and coordinate—not only at a local level but on national, regional, and global levels— and structures to oversee all elements of the work and adapt it as necessary. The USAID ASSIST Project works with existing management, government, and organizational structures, adapting to the context of these existing structures to yield improved results in health system processes.\n\nManagement structures can be facilitated by creating roles for coordinators, points of contact, etc., that are appropriate for the context. For instance, in developing roles, one improvement coach should be identified for every 5–15 facilities, with one supervisor for every 5–10 improvement coaches, and one coordinator for every 5–10 supervisors. These proposed numbers come from many years of experience; too many improvement coaches and supervisors create over-saturation in the facilities, while too few makes it difficult to manage the teams. Improvement coaches convene the improvement teams, supports leaders, and actively builds the improvement teams. Supervisors support the improvement coaches by accompanying improvement coaches when convening improvement teams to provide additional guidance and monitor the progress of the improvement teams and the work of the improvement coaches. Coordinators provide overall oversight of supervisors to ensure consistency across the improvement team activities. The strategy for maintaining a management structure requires accountability and continuity, with attention to detail.\n\nAccountability occurs at multiple levels, with districts holding facilities accountable for improvement, supervisors and coaches holding staff accountable, and staff holding themselves accountable to improvement activities to support the work and progress of their improvement team. In managing improvement activities, improvement teams must have the structure to convene regularly in combination with the accountability to take action in between meetings to make the changes needed for improvement and collect data to keep track of the results related to those changes.\n\nThe culture of a successful management structure requires communication, meticulous planning, and engagement (Table 2). Meticulous planning requires attention to detail in managing teams. Such planning requires specific schedules of activities with clear deliverables, follow-up, timelines, and individuals assigned to keep meetings and activities on track. In the process, schedules and activities must be monitored and adjusted based on reality to make the schedules “living” rather than “static”.\n\nLeadership engagement in activities must involve key members of leadership, improvement coaches, frontline staff delivering healthcare services, and administration, each communicating with one another to work towards a common goal. Communication, as will be described further, should be open, honest, and free of blame. At the heart of management structures lie individuals that can be leveraged as “champions” in supporting and maintaining the structure.\n\nClear and open communication\n\n“Communicating with teams on a regular basis to gather learning and push this information back out to the other teams.”\n\n“Meet with key people at national level, local level, and district level to let them know what the project is about and what it can do and what they can do to support it.”\n\nJust as structures are important, communication with individuals and improvement teams within these structures is essential. Communication allows improvement teams to learn from one another and to ensure that teams have a clear understanding of what and how they are going to improve. Interaction with people at the national, local, and district levels allows the USAID ASSIST Project to articulate what our project is about and how we can work with USAID Missions, Ministries of Health, local entities, and partners to improve healthcare. Our lessons for clear and open communication are detailed in Table 3 and Box 2.\n\n\n\n1. Use internal communication within the project to actively problem-solve and share successes.\n\n2. Identify and discuss solutions for common challenges.\n\n3. Communicate challenges that need support to solve (i.e. internal sharing meetings, internal quality improvement skills-building meetings, case studies, ad hoc meetings, etc.)\n\n4. Develop mechanisms to share learning between facilities and between people managing the program both within ASSIST and in government.\n\nDifferent country contexts involve different scenarios, including: countries where such structures and communication schemes are in place at all levels; countries where efforts are mostly at decentralized levels without formal structures at higher levels; and countries in which there are horizontal approaches in disease and health programs and where such structures ally specific management bodies. Recognizing there are different improvement approaches and terminology, regular communication also helps to avoid confusion or misunderstanding. Clear and open communication includes the communication between ASSIST and improvement teams at learning sessions; communication between coaches and teams during coaching visits; and communication between these sessions. For areas in which internet access was available, communication occurred through emails, WhatsApp and other mobile applications, SMS messaging, and Zoom videoconferencing, as available.\n\nCommunication must occur on a regular basis. Improvement teams keep ongoing record of their work, reporting on this work to their supervisors and to the USAID ASSIST Project. Regularity of interaction encourages improvement teams to problem-solve, learn from improvement progress and challenges, and recognize steps for sustaining and institutionalizing improvement. As will be discussed in the next section, a key to improvement activities is learning. Learning requires open communication between individuals and improvement teams. Open communication means that the culture allows for communication of both successes and failures without fear of blame or penalty. Clear and open communication also assists in setting the stage for shared learning, collaboration, and support in our work, as will be discussed below.\n\nShared learning, collaboration, and support\n\n“Connecting people so that one hospital [or health center] can help another.”\n\nManagement structures must allow for clear and open communication by accepting and acknowledging both failures and successes as learning opportunities. Developing such a management structure requires management actions that support and create opportunities for open communication, collaboration, and engagement of staff. By encouraging staff to communicate and work together, a culture of shared learning and collaboration can be supported. This communication also provides key feedback to management on what is working and what is not, to continually fine-tune strategy and technique.\n\nCommunication is transferred into shared learning, collaboration, and support not only horizontally across health facilities, but vertically between leadership and indirectly with partners. For instance, health facilities may need to communicate and collaborate vertically with local leaders to solve ongoing problems within the facilities, or vice versa. Similarly, shared learning, collaboration, and support occurs when we work with partners with specific strengths and expertise to solve specific problems. In this relationship, we also offer our expertise and resources to support, collaborate with, and problem-solve with our partners and other local entities. Our approaches and lessons for shared learning, collaboration, and support are illustrated in Table 4.\n\nFor instance, using the “technique” of making connections, improvement teams can not only share their experiences with one another, but learn from their successes and failures. Because teamwork is necessary for improvement, shared learning and collaboration sets a tone and “strategy” for scale-up planning and creates a format for people to work together for larger-scale outcomes. This is the “culture” of attention to detail required for such planning and coordination. Through shared experiences, people can come together and work towards a common goal, learning from their experiences along the way.\n\nOwnership, engagement, and empowerment\n\n“Allowing [improvement] work to be given to local authorities so that it can be taken over from us. We need to lead, motivate and encourage them. It is also important to bring them together.”\n\nIn addition to leadership, management structures, and a focus on learning, the sustainability of activities under the USAID ASSIST Project is dependent on creating a “culture” in which local leaders and individuals are engaged and have a sense of ownership and empowerment to continue the work. Engagement requires ongoing negotiations and conversations with the Ministry of Health, local government and authorities, and other stakeholders in the health sector. Conversation itself is not enough. Local entities must feel invested in these activities and must feel encouraged to continue the work.\n\nWhen the USAID ASSIST Project engages in work with a country, initial discussions emphasize that we are in the country to help and describe what the role of ASSIST is and how frontline workers can do their work and collect data so that we may support their improvement activities. We also make sure to engage in conversation with municipalities and districts as their leadership, support, and engagement in the work within their region is necessary for accountability to trickle down to the facility and community levels. The Chiefs of Party for the USAID ASSIST Project play a key role in communicating with the Ministry of Health to create and maintain ownership and accountability. The lessons from our work in increasing ownership, engagement, and empowerment of local actors under the USAID ASSIST Project is highlighted in Table 5. Practical approaches for engagement are described in Box 3.\n\n° Accountability to results and pride in one’s work\n\n° Recognizing one’s role and effect on results\n\n\n\n1. Focus on results that matter to the people who will own the work.\n\n2. Link people new to quality improvement with influential people or groups that have successfully used quality improvement approaches.\n\n3. Support people with good results to talk widely about their results and help them show their results to important figures.\n\n4. Encourage people to talk about the quality improvement methods that they use to achieve good results.\n\nOwnership, engagement and empowerment must exist not only with local government and authorities but must be rooted within improvement teams. Staff in facilities must have ownership over their work and be engaged with the work they are doing in their facility. Leaders within facilities, whether supervisors or coordinators, then engage and empower the staff they are working with, help them achieve their goals, and help them remain accountable to their work. One way that ASSIST encourages ownership, engagement, and empowerment in facility staff is by helping them get results quickly and promoting these results. Achieving positive results requires us to help new quality improvement teams choose aims that are more likely to be attainable. Initial aims must therefore:\n\n1. Be easy to measure without creating new measurement systems\n\n2. Be related to processes that are largely under the control of the improvement team\n\n3. Involve processes that occur frequently so that data can be measured more frequently and changes are more visible (issues of routine care are better than issues of complications management, for instance)\n\n4. Benefit the improvement team if the processes are improved (i.e., the team’s supervisor will be happy with the results, and patients will be happy with the results)\n\nThe “technique” of quality improvement requires multidisciplinary teams and the involvement of individuals at all levels of healthcare delivery. The focus on improvement engages not only “experts” in improvement but empowers others to play a role in quality improvement, engaging them as active members on a team working towards improved health outcomes. By engaging the Ministry of Health, implementing partners, and other local entities into our work, a “strategy” is created in which improvement is built into the system. Engagement of the Ministry of Health should begin at the start of the activity, keeping up-to-date visits and contact with officials. Through this communication, support of the Ministry of Health should be requested and successes celebrated throughout the course of the activity. At the end of the activity, support for the work is to be handed over to the Ministry of Health.\n\nIn some health systems, countries have developed national policy and strategies on quality improvement that have integrated roles and responsibilities for quality in the jobs of staff at all levels. In other instances, countries may wish to develop gradual plans to operationalize quality improvement into their health system. By actively working with government counterparts at local, national, and regional levels, we can ensure that they are engaged in a way that fits improvement into their own structures.\n\nPartnerships\n\n\n\n“Partners are working together with the team, maintaining communication to facilitate shared learning and creating tools that allow teams at the local level [to use them].”\n\nThe role of partners in our work allows us to broaden the range and scope of our work in improving healthcare. Our partners include both government and non-government entities, with host country national (HCN) partners being local organizations. In July 2017, ASSIST worked with 159 government and implementing partners in improving health outcomes. In the case of our partnerships, our “culture” is one of a shared vision and collaboration. Because of this culture, our work in some countries, such as in India, the support of the improvement work of the USAID ASSIST Project occurred primarily through partners. In our work with implementing partners (IPs), the USAID Mission or office that provides funding may decide whether we are working with implementing partners through joint planning of activities, coordination, progress review, reporting, etc. In the cases in which we work with implementing partners, the leadership of USAID leads the work and directs the partnerships. While some of these agreements are formal agreements with signed Memoranda of Understanding between parties or an established partnership established by USAID, many others are informal agreements developed through a relationship built through ongoing collaboration and coordination with the partner entity.\n\nBy working in collaboration with partners, we engage and maintain communication with a larger number of improvement teams, who can share their expertise and experiences so that teams are learning from one another and sharing tools that can improve their work. The support of our partners in the form of technical and human resource support also allows us to make local support available and accessible to our teams locally through universities and local experts. Our lessons in working with partners are detailed in Table 6.\n\n\nConclusion\n\nThe work of the USAID ASSIST Project relies on the success of quality improvement teams worldwide. Improvement teams must have both the technical and leadership capability to carry out and sustain improvements to improve healthcare outcomes. To manage these improvement teams, we have developed a management structure with a “culture” of quality, engagement, and empowerment. With the “technique” of quality improvement methodology at the center, we have used the “culture” of clear, open communication and shared learning and collaboration. By connecting improvement teams, teams can share information and lessons learned from their successes and mistakes, sharing these experiences with one another in a way that promotes learning and teamwork to reach a larger scale. The work of the USAID ASSIST Project on such a large scale and across more than 4,000 improvement teams globally would not be possible without the involvement and engagement of local leaders and partners, who share a vision and dedication to improving healthcare outcomes and strengthening healthcare systems.\n\n\nEthical statement\n\nThe involvement of all meeting participants and those involved in email correspondence for this study was entirely voluntary. Participants involved in meeting discussions, interviews, and email correspondence were made aware that the discussions would be documented for use in this study. Participants were made aware that their participation in the discussions and email correspondence used for this study were voluntary and that they could choose to no longer participate at any time.\n\n\nData availability\n\nThe data provided for this study is in the form of meeting notes (Dataset 1) and email correspondence with participants. Since the email correspondence may contain potentially identifying information, this is not openly available. Those interested in accessing this data may contact Leighann Kimble at lkimble@urc-chs.com with any questions or requests. Data will be accessible under the following conditions <insert conditions>.\n\nF1000Research: Dataset 1. Meeting Notes, 10.5256/f1000research.16099.d2222928", "appendix": "Grant information\n\nThe USAID Applying Science to Strengthen and Improve Systems (ASSIST) Project is made possible by the support of the American people through USAID. The USAID ASSIST Project is implemented by University Research Co., LLC (URC) under the terms of Cooperative Agreement Number AID-OAA-A-12-00101.\n\n\nSupplementary material\n\nSupplementary File 1: ASSIST country examples of managing hundreds of improvement teams.\n\nClick here to access the data\n\n\nFootnotes\n\n1World Health Organization (WHO). An Approach to Rapid Scale-up: Using HIV/AIDS Treatment and Care as an Example. 2004.\n\n2Massoud MR, Nielsen GA, Nolan K, Nolan T, Schall MW, Sevin C. A Framework for Spread: From Local Improvements to System-Wide Change. IHI Innovation Series white paper. Cambridge, Massachusetts: Institute for Healthcare Improvement; 2006. (Available on www.IHI.org)\n\n3Beracochea E. (Ed). “Scaling Up High-Impact Interventions in Health Care”. Improving Aid Effectiveness in Global Health. Ref AID Book ppls P8 AID effectiveness book in a box; Box with approaches P7 AID effectiveness book; Lessons learned box P10 AID effectiveness book; 2015.\n\n4Berwick DM. Lessons from developing nations on improving health care. BMJ 2004;328: 1124–9.\n\n5Langley GJ, Moen RD, Nolan KM, Nolan TW. The Improvement Guide: A Practical Approach to Enhancing Organizational Performance. Second Edition. San Francisco, CA: Jossey-Bass; 2009.\n\n6Massoud MR and Mensah-Abrampah N (2014) A promising approach to scale up health care improvements in low-and middle-income countries: the Wave-Sequence Spread Approach and the concept of the Slice of a System [v1; ref status: indexed, http://f1000r.es/37o] F1000Research 2014, 3:100 (doi: 10.12688/f1000research.3888.1)\n\n7Massoud MR and Mensah-Abrampah N. “Scaling-up of High Impact Interventions”. Ed. Beracochea in Aid Effectiveness in Global Health: How to Realize Global Health Goals. Springer 2015.\n\n8Massoud MR, Kimble L, Boguslavsky V et al.. Dataset 1 in: Managing hundreds of improvement teams: A case study of the USAID Applying Science to Strengthen and Improve Systems Project. F1000Research 2018, (doi: 10.5256/f1000research.16099.d222292)" }	[ { "id": "40061", "date": "16 Nov 2018", "name": "Peter Lachman", "expertise": [ "Reviewer Expertise Patient Safety Quality Improvement Paediatrics" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an interesting report on the management of quality improvement teams by a large donor aid organisation. The focus is on what it takes to set up a successful programme at scale. This is one of the logistical issues to be addressed in any programme. The construct that is offered could be adapted at any level - it could be scaled down as well as scaled up.\nThe framework is useful, and the authors provide a good rationale and explanation of the way to set up the programme. This would need to be tested. The question of successful implementation of projects over time is a reference to generalizability.\n\nIt would be interesting if this is now being tested in a prospective way. Much of what is concluded is not essentially new; what is new is the theory in the context of spread and scale up.\nA few questions need to be answered going forward:\nDo they now use this framework when they start a new programme? Do they use the framework as a continual real time assessment of the implementation of a programme, or is it a theoretical construct?\nThe paper does not provide outcome measures - how do they measure the success of the framework?\nIs there a measurement plan going forward to evaluate the framework? An overview of the outcomes of their programmes would assist in demonstrating how this works. I am sure that there is a range of success and perhaps they could analyse their programmes against the framework linked to programme outcomes.\nOne also needs to consider the issue of sustainability. The authors state that “Improvement teams must have both the technical and leadership capability to carry out and sustain improvements to improve healthcare outcomes. To manage these improvement teams, we have developed a management structure with a ‘culture’ of quality, engagement, and empowerment”. They imply that this will result in sustainable change. It is a laudable aim, but one needs data to support the presumption of sustainability.\nOne can assess this by looking at the differences between those who have followed this framework and those who have not and what happens when the programmes end. One needs to report on whether a difference in sustainability of programmes is there between those who followed this framework and those that did not to determine the outcomes long term, especially after aid and technical support is withdrawn and support has stopped. Experience in UIC is that there is a fall-off once technical support is withdrawn.\nFinally, the word ‘patient’ appears only twice in the paper. I think that it may be implied that partnerships include partnerships with patients. Unfortunately, this is not stated, and I wonder whether another dimension to consider in addition to partnerships with people who deliver and plan care, would be partnerships with patients and families in the improvement process.\nI add a recent review on spread from The Health Foundation which they could consider (Horton et al., 20181).\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4247", "date": "21 Nov 2018", "name": "M. Rashad Massoud", "role": "Author Response F1000Research Advisory Board Member", "response": "Dear Peter,Thank you very much for your thoughtful comments. We would like to respond to some of the questions you posed below, each of which are valid points.The scope of this paper was to present the main points that resulted from discussions among senior staff on the USAID ASSIST project. Although these were recurring themes between the activities and staff working on the project, we did not develop a framework as a result of these discussions. Rather, these points are a synthesis from the discussions that show the common practices in our work. Through the group discussions and in-depth interviews described in this paper, were we able to draw out these key points and present them in this paper.The proof of success in the use of the six interrelated categories presented in this paper is reflected in the outcomes of care that are improved in USAID ASSIST activities, as presented in the country examples in the annex of this paper. In response to your comment on patient-centeredness, the role of the patient is something that we apologize does not come through in this paper. In our work, it is all about the patient. This paper talks about the teams that are serving the patients and delivering care to patients. In many cases, patients are in fact part of the team, as in the case with expert patients as part of the improvement team.Unfortunately, in the scope of this paper, we were unable to address all the points you have presented due to the interest of page limitations and conciseness. We will consider and hope to address these points in future discussions on this topic.Thank you again for your comments." } ] }, { "id": "41222", "date": "14 Dec 2018", "name": "David Galler", "expertise": [ "Reviewer Expertise System theory and practices" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThanks for the opportunity to read this paper which I found fascinating for a number of reasons:\nOutlining the breadth and depth of the US Assist program.\n\nThe complexity of the issues addressed and the methodology of distilling down messages from the many different conversations necessary to deliver the final conclusions.\n\nMaking sense of really complex pieces of work to assist others in being successful is a terrific goal and I appreciate the hard work that has gone into this.\nI would be interested to know given the experience and expertise of the authors whether they were in anyway surprised by the results of those conversations and whether any element of pre-test bias may have influenced the distillation process to deliver the six key components and how that was managed - they seem so intuitively correct. This is always a hard question but I think the process you used may well have overcome that inherent bias if it existed.\nWere any thoughts given to these key components in the setup of the original work?\n\nThe six key components and the descriptions of each of them as they relate to the Quality Management Triangle is neat and may well be of great practical support to those that will subsequently use it.\n\nAs a next step I would like to see this framework tested from the get-go in the establishment of new work.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1722
https://f1000research.com/articles/7-1721/v1	31 Oct 18	{ "type": "Research Article", "title": "DomainBuilder: the knowledge authoring system for SlideTutor Intelligent Tutoring system", "authors": [ "Eugene Tseytlin", "Faina Linkov", "Melissa Castine", "Elizabeth Legowski", "Rebecca S. Jacobson", "Eugene Tseytlin", "Melissa Castine", "Elizabeth Legowski", "Rebecca S. Jacobson" ], "abstract": "One of the major challenges in the development of medical Intelligent Tutoring Systems (ITS) is the development of authored content, a time-consuming process that requires participation of discipline experts. In this publication, we describe the development of software systems called DomainBuilder and TutorBuilder, designed to streamline and simplify the authoring process for general medical ITSs. The aim of these systems is to allow physicians without programming or ITSs background to create a domain knowledge base and author tutor cases in a time efficient manner. DomainBuilder combined knowledge authoring, case authoring, and validation tasks into a single work environment, enabling multiple authoring strategies. Natural Language Processing (NLP) methods were integrated for parsing existing clinical reports to speed case authoring. Similarly, TutorBuilder was designed to allow users to customize all aspects of ITSs, including user interface, pedagogic module, feedback module, etc. Both systems underwent formal usability studies with physicians specializing in dermatology. Open-ended questions assessed usability of the system and satisfaction with its features. Incorporating feedback from usability studies, DomainBuilder and TutorBuilder systems were deployed and used across multiple universities to create customized medical tutoring curriculum. Overall, both systems were well received by medical professionals participating in usability studies with participants highlighting ease of utilization and clarity of presentation. Usability study participants were able to successfully use the system for the authoring tasks. DomainBuilder and TutorBuilder are novel tools that combine comprehensive aspects of content creation, including creation of domain ontologies, case authoring, and validation.", "keywords": [ "Natural language processing", "Intelligent Tutoring Systems", "Concept recognition", "Biomedical terminologies", "ontology development", "Auto-coding", "System evaluation" ], "content": "Introduction\n\nWhile intelligent tutoring systems (ITSs) are becoming more common in medicine and proving to be increasingly effective in educating medical personnel and patients1–3, they are difficult and expensive to build. One of the major challenges in the development of medical ITSs is the creation of authored content, a time-consuming process that requires domain knowledge expertise.\n\nAuthoring systems, computer programs that can allow an instructor to prepare computer-based medical instructional materials without the need to know programming languages or have more than minimal familiarity with the computer hardware, are especially needed in the field of medical resident training, where traditional apprenticeship models of education are costly and time consuming. Previously published reports highlighted the use of various authoring systems for ITSs in non-medical settings, including WETAS4, NetCoach5, and CTAT6. WETAS is a web-based authoring system built to reduce the authoring efforts for constraint-based ITSs. It also functions as a web-based tutoring engine that performs all of the common functions of text-based tutors4. This system is well suited for institutional, web-based tutors like SQL-Tutor7. NetCoach was designed to enable authors to develop adaptive learning courses without programming knowledge5. Cognitive Tutor Authoring Tools (CTAT) is a suite of tools for quick development of cognitive and example tracing tutors6. Authoring systems are commercially available for traditional computer aided instruction (CAI) and multimedia-based training, but these authoring systems lack the sophistication required to build intelligent tutors8, especially for the medical education setting.\n\nThere are many authoring tools that in the past decade lowered the skill threshold for building tutors9–16. For example, Blessing et al. explored multiple approaches to authoring example-based tutors for procedural tasks. They discussed convergence of multiple lines of authoring tools for step-based problem solving tutors toward example-based authoring, which provides an interface where the author builds tutoring content and student support for an individual example or set of examples17. Aleven, Roll, et al. reviewed work on the creation of a knowledge-engineered, rule-based, executable model of help-seeking that can drive tutoring18.\n\nThe Generalized Intelligent Framework for Tutoring (GIFT), developed by the Army Research Laboratory-Human Research and Engineering Directorate, allows researchers to manipulate the computer-based training system components in order to test empirically the effect for different assessments and instruction strategies on learning outcomes19. GIFT is an open-source intelligent tutoring system (ITS) architecture provides authoring, delivering, managing, and evaluating adaptive instruction that was very important for the field in the past decade. Similarly important is the previously published AutoTutor system, which helps students learn by holding a conversation in natural language. AutoTutor is adaptive to the learners’ actions, verbal contributions, and in some systems their emotions20.\n\nIn reviewing authoring tools that have been published in the past two decades, it is important to mention progress in the field of ontology development. An ontology is a general-purpose framework for representing knowledge in various domains. For example, Isotani et al. (2008, 2009) used ontologies to represent strategies in collaborative learning and created systems that can link domain independent knowledge with domain content21,22. Their work, however, is specific to their system for collaborative learning and does not apply to the domain of pathology where an ontology is used to represent concept hierarchies of clinical findings and linking them with one or more pathologic diagnoses to support a model-tracing tutor in the pathology domain.\n\nWhile all of the systems mentioned so far, including those reviewed by Murray8, offer a rich set of features and are designed specifically to simplify and streamline the content authoring process for their respective classes of ITS, none of them are designed for domains that present a user with an interactive virtual slide in the field of dermatopathology. The pathology domain is unique because it relies on the use of annotated virtual slide images for training, which makes general purpose ITS authoring systems unusable. Existing authoring systems (WETAS4, NetCoach5, and CTAT6), developed mainly for the domains of mathematics, physics, and programming, focus on the educational content that is presented to the student in the form of a text question with clear solution paths. A notable disadvantage of these three authoring systems for application in the field of medical education is that solution paths are complex and image recognition systems are required, thus these systems may not be adaptable to an ever-changing medical environment. In our previously published studies, we found that these existing systems are ill-suited for authoring knowledge for the Slide Tutor ITS, a system which was designed to train pathology residents23.\n\nOur group started to address educational challenges in the field of pathology resident education by introducing SlideTutor ITS, a visual classification tutor that relies on expert medical knowledge represented in the form of an ontology24. SlideTutor is a general visual classification tutor that was created to teach microscopic diagnosis of inflammatory dermatological diseases23. SlideTutor is a model-tracing tutor that presents pathology trainees with a virtual whole slide image of dermatopathology and asks them to identify features on the slide in order to come to a differential diagnosis25. For an ITS like SlideTutor, the task requires medical domain expertise, knowledge engineering experience, as well as an understanding of the general ITS architecture. In our previous publications, we have also described ReportTutor, which combines a virtual microscope and a natural language interface to allow students to visually inspect a virtual slide as they type a diagnostic report on the case of melanoma of skin26. A version of SlideTutor was created that explored a student’s metacognition27. Most recently, we have created a simulation tutor geared toward practicing generalist pathologists that combined both SlideTutor and ReportTutor into a single tutoring environment covering both diagnostic and prognostic aspects of solving a case in the melanocytic domain28.\n\nIn our previous attempt to build an authoring system for the SlideTutor ITS, the knowledge base was developed using the Protégé ontology editor, which is a free, open-source ontology editor. The Protégé framework was selected as its plug-in architecture can be adapted to build both simple and complex ontology-based applications29. The case authoring system was a Protégé plug-in that integrated the digital slide viewer, which was adequate for our initial research purposes. The disadvantage of using this framework was that mastering Protégé software required a steep learning curve, since understanding ontology development is not a basic skill that general users and even medical professionals generally possess. The lack of an intuitive interface also made collaboration between the knowledge engineer and medical domain experts difficult, as large amount of time was spent training the domain expert on the use of the complex interface. Thus, the disadvantage of the previous approaches developed to build ITS content in this area was that it was never designed for use by medical professionals and hence was not very intuitive for content creators outside of the core ITS group. Existing systems developed by us and others have been reviewed to inform the design of DomainBuilder30 and TutorBuilder31 systems described in this paper.\n\nWhile the general architecture of tutoring systems developed in our laboratory remained unchanged32, our work over the past decade focused on improving knowledge representation to accommodate different domains in the field of pathology. Each version of the tutor had its own knowledge representation (specific to each domain), as well as an authoring system for tutor cases. The overall aim of our effort was to ensure that new content for our ITS was usable outside of our laboratory with a minimal learning curve on the side of the user. To accomplish this aim, we focused on developing an intuitive knowledge representation system. We ran four major studies in several pathology domains33–36, focusing on various types of tutor feedback, user meta-cognition, and student modeling. During that time, we developed a list of features that an authoring system would need to have, to streamline the creation and deployment of new content for the SlideTutor ITS. The aim of this manuscript is to describe the development of the software systems called DomainBuilder and TutorBuilder that streamline and simplify the authoring process for general medical ITSs.\n\nDomainBuilder and TutorBuilder are innovative in several ways and were designed to fill a serious gap in the current literature on medical ITSs. To our knowledge, our system represents one of the first attempts to build an authoring system for a set of model tracing intelligent tutoring systems in a medical domain. Our approach also explores ways of building intelligent tutors that are adaptable to various healthcare settings and reduces the need to have an input from multiple clinicians (or other domain experts) during design stages. DomainBuilder system uses a unique interface to link virtual slide annotations with diagnostic and prognostic features. Additionally, we used an innovative spreadsheet style interface for rapidly creating complex rules for diagnostic inference and comparing those rules between diagnoses.\n\nThis paper outlines the system descriptions of DomainBuilder and TutorBuilder authoring systems for a medical ITS, as well as describes two usability studies that were conducted to assess the usability of both systems among physicians without programming skills.\n\n\nMethods\n\nTo meet the needs of training pathologists, the SlideTutor ITS architecture has been developed to accommodate tutor customization and pedagogic content. The domain knowledge for SlideTutor is represented in the form of an application ontology. The SlideTutor ITS utilizes two main constructs: findings and diagnoses. A finding is a piece of evidence that a user can identify in a tutor case, while a diagnosis is a solution for a case. Each diagnosis can be described by a set of findings that can distinguish it from other diagnoses. A tutor case is represented by an annotated virtual slide image and a set of findings and diagnoses that are linked to those annotations.\n\nThis system builds on our previous experience with the development of a plug-in to an existing ontology editor, Protégé. During review of domain content, the domain expert (an attending pathologist or fellow) often became overwhelmed with reviewing an entire ontology to provide feedback on the content. Also, substantial effort was needed from a knowledge engineer to rearrange the disease and finding hierarchies within the ontology editor. Another disadvantage to the ontology editor was that there was no way to visualize the entire domain space to see what was lacking. Finally, the ontology editor required that the tutoring domain always be added to the ontology from a top-down manner (e.g. once authoring process began, workflow had to be interrupted to add a missing disease to the knowledge base in a separate tool), which interfered with the workflow of case authoring. Systems described in this paper were designed to address these disadvantages of the existing systems.\n\nBased on our experience with intelligent tutoring systems23,26,32 and a formal usability study, our group developed a design that simplified and streamlined the process of content creation for SlideTutor. Given the unique nature of our ITS and its target group, several key steps were taken to develop DomainBuilder. The first step involved reviewing existing frameworks and previous authoring systems. We utilized the Space Tree System from the University of Maryland as an efficient way to represent concept hierarchies. This provided a simple interface to view the cases and allowed for easy rearrangement of disease and finding hierarchies, as illustrated in Figure 1. SpaceTree is a tree browser that builds on the conventional layout node link diagrams along a single preferred direction37.\n\nThen, a design study/focus group was conducted to identify the most effective methods to represent relationships between “Diagnoses” and “Findings” to the user (a “spreadsheet” design inspired by Excel software format). The “spreadsheet” format allowed for creation and review of domains in small segments that simplified the interactions between the knowledge engineer and the domain expert (Figure 2). The report template layout was inspired by the College of American Pathologists (CAP) cancer protocols to ensure that the layout is consistent with reporting standards and to ensure that each report is completed with the necessary required elements. Google News Cloud and other Word/Tag cloud user interface techniques inspired our validation layout. This layout provided a “birds-eye view” of the domain demonstrating the relationships between findings, diagnoses, and their related cases. Lastly, the authoring layout used to link annotations with domain content, allowed for bottom-up data entry. For instance, DomainBuilder uses NLP to process an existing clinical report and extract the relevant ontology concepts that should be authored in a case. This expedited the authoring process as all that was required was annotating the slide and associating the annotations with ontology concepts by drag and drop feature. In addition, the authoring layout allowed for concepts to be added to the ontology without leaving the authoring environment, which allowed the authoring workflow to be uninterrupted.\n\nTutorBuilder is a system that has been designed to allow authors to create complex tutoring system configurations easily and without prior knowledge of programming or an underlying knowledge representation. Tutor builder shares many functional characteristics with Domain Builder, such as the ability to be used by non-experts. In the development of TutorBuilder, a Wizard interface style has been selected because the task of creating a customized tutoring system was a sequence of decision points. It is important to give users an ability to look at the result of their configuration within the authoring tool itself. Thus, the “Play” feature of TutorBuilder was inspired by common integrated development environments (IDE) such as Eclipse that allows one to immediately see the results of one’s code.\n\nThe goal of both software systems was to support multidisciplinary and cross-disciplinary collaborations among multiple content authors and institutions to improve training in pathology. Specifically, we aimed to support collaboration between knowledge engineers and domain experts by easily visualizing complex relationships between concepts in a domain ontology for medical systems specifically. While developing the software, we used existing resources such as ontologies from BioPortal38, synonyms and definitions from Enterprise Vocabulary Service (EVS)39, and other controlled terminologies such as UMLS40.\n\nOur overarching design goals were consistent with those outlined by Murray et al., which include: decrease the effort (time, cost, and/or other resources) for making intelligent tutors; decrease the skill threshold for building intelligent tutors; help the designer/author articulate or organize her domain or pedagogical knowledge; support good design principles (in pedagogy, user interface, etc.); and enable rapid prototyping of intelligent tutor designs (i.e. allow quick design/evaluation cycles of prototype software)8,41.\n\nDomainBuilder had the following design objectives: (i) to combine knowledge authoring, case authoring, and validation tasks into a single work environment, (ii) to allow users to transition between tasks easily; enabling bottom-up, top-down, and hybrid authoring strategies, (iii) to provide an intuitive graphical user interface (GUI) for authoring that enables any user with domain knowledge to construct domain ontologies and cases without the need to understand the underlying knowledge representation, (iv) to allow authors to create complex diagnostic relationships between diagnoses and findings using familiar tabular representations, (v) to integrate Natural Language Processing (NLP) methods for parsing existing clinical reports to speed case authoring, (vi) to streamline digital slide annotation work flow by enabling simple tagging of image annotations through drag and drop linkage to concepts in the domain ontology, and (vii) to detect inconsistencies between case authoring and diagnostic inference from the domain ontology.\n\nTutorBuilder had the following design objectives: (i) to provide an intuitive graphical user interface (GUI) for authoring that enables any user, even those without the technical abilities, to construct tutoring systems without the need to understand the underlying knowledge representation, (ii) to provide the ability to mix and match different inter-operable components of the tutoring system, (iii) to provide the ability to evaluate the tutoring system as it is currently configured from the authoring tool itself, (iv) to provide the ability to modify the feedback type, as well as its content, and (v) to provide the ability to reproduce previous configurations of the tutoring system that were hand crafted before TutorBuilder’s creation within the authoring tool.\n\nDesign of prototype. Incorporating feedback from the usability studies, DomainBuilder and TutorBuilder systems were deployed and used across multiple universities to create customized medical tutoring curriculum. The use of both systems allowed for several major research studies to be conducted in a short period of time23,26,32,42. One of the best examples of user feedback that drove a change to the prototype was a request to develop a simpler user interface for tutor builder to accommodate the most frequently used tutor configurations.\n\nValidation study design. A total of two separate usability studies were conducted: one for DomainBuilder and one for TutorBuilder. The DomainBuilder authoring system was used to create domain content, while TutorBuilder was used to create customized ITS configurations independent of content. As both systems were developed separately and had a different target user base, two usability studies were required. Considering internal feedback from the researchers that were using both systems during the development steps, as well as feedback from two formal usability studies, there were three major cycles of revisions for both systems.\n\nDomainBuilder usability study. The DomainBuilder usability study was conducted with ten medically trained pathology users with varying levels of postgraduate training and no previous experience with ITSs building. The participants have been recruited through pathology research conferences such as Pathology Informatics APIII, College of American Pathologists (CAP), and American Society for Clinical Pathology (ASCP). Participants were given 32 tasks (Supplementary File 1) to complete, with each task consisting of multiple steps for testing the usability of the interface. Each step had four possible categorical response types: (i) successfully perform the task, (ii) fail to perform the task, (iii) complain about the step, and (iv) report a surprise (unanticipated problem with the interface).\n\nFollowing the usability tasks, participants completed a survey assessing their opinions on various aspects of the system. The questionnaire (Supplementry File 2) included 29 scaled, closed-ended questions on ease of use, understanding, and sufficiency of numerous features of the system and 3 open-ended questions in which users were asked their opinion on what they liked and disliked about the system and suggestions for future tool improvements and modifications.\n\nTutorBuilder usability study. The TutorBuilder usability study followed the same design as the DomainBuilder usability study; however, the tasks were different and specific to the TutorBuilder system. A total of six pathologists with varying levels of postgraduate training and no previous experience with ITSs building participated in the usability study. They were recruited through pathology training programs in the greater Pittsburgh area.\n\nWhile our original intent was for the usability study to follow the model of learning evaluation published by Kirkpatrick43, which focused on user reaction, learning, behavior, and results, we found that this model has been criticized for lack of research. A newer model suggested by Holton focuses on three primary outcomes, including learning, individual performance, and organizational results. These are defined, respectively, as achievement of the learning outcomes, change in individual performance because of the learning being applied on the job, and results at the organization level as a consequence of the change in individual performance. Thus, our final plan for a usability study has been inspired by both, the models suggested by Kirpatrick and Holton44.\n\nEthical approval for the validation studies was obtained from the University of Pittburgh (020348 and 0608117). All participants provided written informed consent prior to participation.\n\nMinimum system requirements for both products. The system was developed in Java and deployed using Java WebStart technology. There was no specialized hardware requirements to run the software on a typical personal computer.\n\nDomain builder description. The primary purpose of DomainBuilder is to author ITS domain content. This content includes a domain ontology in the form of findings, diagnoses, and their relationships. In addition to the ontology, the domain content also includes medical cases to be solved by students, which are generally presented as annotated virtual slide images. The ITS uses domain ontology along with the medical tutor case information to reason about a given domain and provide appropriate feedback to a student.\n\nKnowledge representation and implementation. The SlideTutor ITS utilizes two main constructs: findings and diagnoses. A finding is a piece of evidence that a user can identify in a tutor case, while a diagnosis is a solution for a case. Each diagnosis can be described by a set of findings that can distinguish it from other diagnoses (e.g. the findings vacuolization in basal layer, increased mucin, and mild perivascular lymphocytic infiltrate indicate a diagnosis of dermatomyositis). While those constructs are specific to medical domains, any knowledge representation that distinguishes these two components can be easily adopted within our ITS framework. The Knowledge Base (KB) is implemented using Web Ontology Language (OWL)45. We used Protégé-OWL for KB development and Pellet for Description Logic (DL) reasoning. Figure 2 shows the general KB schema that implements the required ITS framework. All concept entities in our KB are represented using OWL Classes, while OWL Individuals are used to represent tutor case-specific instance information.\n\nKnowledge authoring. The graphical knowledge base creation tool is divided into three interfaces, each of them designed to streamline a specific subtask in the knowledge creation process.\n\nThe Hierarchy builder tab (Figure 1) is focused on building a desired hierarchy of concepts such as findings and diagnoses as well as adding meta-information for each concept, such as the concept’s definition, possible synonyms, linkage to a controlled medical vocabulary and an optional pictorial example. For example, the user would be required to enter the findings vacuolization in basal layer, increased mucin and mild perivascular lymphocytic infiltrate into the findings hierarchy. Similarly, the user would create a concept for the related diagnosis, dermatomyositis, in the disease hierarchy. To facilitate easy and efficient navigation through the ontology, the SpaceTree technology is used46. In addition to adding findings and diagnoses manually, a user can import existing concept hierarchies from any ontology deployed at BioPortal. Having well-structured concept hierarchies in a domain ontology allows the ITS to reason in general and specific terms about features in a domain. For example, if the student chooses a diagnosis of “perivascular skin disease” the tutoring system can provide feedback that the student must find more evidence to choose a more specific diagnosis for the given case. Meta information provided with each finding and diagnosis allows the ITS to provide definitions and examples to a student to aid in the learning of domain. Additional synonymy is used to generate dictionaries for the ITS that requires NLP to process a student’s input for correctness and completeness.\n\nThe Diagnosis builder tab (Figure 2) is used to create relationships between findings and their related diseases. It uses a table layout, where each column represents a diagnosis with rows below representing a disjunction of findings. For example, a user would add a column by choosing the disease, dermatomyositis from the disease hierarchy to create a column. Once the column is created, the user can then click on each cell under the disease name to enter the findings related to that disease (e.g. vacuolization in basal layer, increased mucin and mild perivascular lymphocytic infiltrate). To create more complex relationships, findings can also be negated as well as conjoined together to form a logical OR expression. Also, this interface allows users to compare similarities between several diagnoses by rearranging rows and displaying them side-by-side. These relationships between findings and diagnoses (diagnostic rules) are used by the ITS reasoning module to check if a student’s partial solution is sound and to provide student with possible next steps in a solution path if a hint is requested.\n\nThe Report builder tab is designed for creation of reporting templates that are used by the SlideTutor ITS when teaching students how to write clinical reports. The Report builder interface consists of a list of potential templates, where each template has a list of triggers that determine when this template should be used as well as a list of reportable items that should be mentioned in a student’s clinical report. For instance, the user could create a trigger of dermatomyositis allowing the tutoring system to enforce the defined reportable items (e.g. must report the tissue type of skin, the anatomic location of the biopsy and the biopsy procedure type before reporting the diagnosis) for that given diagnosis. Reporting templates are used by the prognostic ITS that checks students’ input against a standard checklist for accuracy and completeness.\n\nCase authoring. Based on our past experience with authoring domain knowledge, tutor case authoring is the most time-consuming task. A tutor case usually consists of one or more annotated digital slides, a list of findings that can be observed on those slides, one or more diagnoses for that case, as well as an optional expert report associated with that case (Figure 3).\n\nThe Virtual Slide Component is the most visible and the most widely used component of the system, since most of the time during case authoring is spent adding annotations to digital slides. It consists of a virtual slide viewer, a slide manager, and an annotation manager. The virtual slide viewer can load whole slide images from several different vendors and should be familiar to most pathologists. The slide manager is used to switch between several different slides that belong to a single case. Lastly, the annotation manager is used to organize annotations such as free hand polygons, rectangles, ovals and arrows into tagged groupings. These groupings can then be linked to findings and/or diagnoses associated with the case by dragging and dropping the tag name to a finding or diagnosis.\n\nA list of findings from the knowledge base that are found or should be found in the given case are displayed to the user above the virtual slide. Findings can be added in three different ways: (i) manually added by a user (ii) inferred by the system from a diagnosis, or (iii) extracted from NLP processing of the clinical report. Those findings are used as objects that can be associated with annotations on a digital whole slide. DomainBuilder uses the underlying knowledge base to infer appropriate findings and diagnoses. Given a diagnosis, DomainBuilder can add a set of findings to the interface for the user to author in this tutor case. Conversely, given a set of findings, the tool can infer and add one or more relevant diagnoses to the interface for the user to author in the tutor case.\n\nThe expert report panel looks like a regular text panel with predefined report section headings. One of the advantages of DomainBuilder is that it uses Natural Language Processing (NLP) techniques to parse an already existing report and extract the relevant diagnoses and findings. For parsing, it uses the SPECIALIST noun phrase parser47 in combination with a technique similar to IndexFinder’s algorithm48 to achieve accurate results. This functionality helps expedite the case authoring and also aids in validation of the authored tutor case content.\n\nValidation. During the validation process, the knowledge base and authored cases are checked for consistency. To complete the validation process, the user needs to see the entire knowledge base, as well as all authored cases. This information needs to be presented in a compact fashion, so it does not overwhelm the user with large amounts of data and the relationships between elements can be viewed. After evaluating several UI techniques, it was decided to use a tag cloud interface49. This method allows users to quickly explore relationships between elements in a limited amount of screen space.\n\nDescription of domain authoring workflow. Since DomainBuilder encompasses modules for authoring of tutor cases as well as domains, it supports several approaches for content authoring. One option is to use a top-down approach, where knowledge is authored first and cases are authored using this authored knowledge. The second option is to use a bottom-up method, where new knowledge is added to the knowledge base while authoring tutor cases.\n\nThe diagram in Figure 4 demonstrates two potential workflow scenarios. Grey boxes indicate steps that involve the use of the DomainBuilder tool. In the initial step, “Acquire Domain Knowledge”, authors must acquire adequate domain knowledge from a domain expert, medical texts, and other relevant materials. In “Create Domain Knowledge Base”, the DomainBuilder tool is used for knowledge authoring. Domain knowledge is created using the Knowledge Authoring tab, which is further subdivided into Hierarchy Builder, Diagnosis Builder, and Report Builder tabs. In the “Acquire Cases” step, authors must acquire the desired number of medical cases for the target domain, preferably covering most of the knowledge base. For each case, the author should have a set of digital whole slide images and a de-identified expert report from the provider who diagnosed that case. The “Case Authoring” tab in the DomainBuilder tool is used for authoring tutor cases. In the “Check Authoring” step, users are checking the knowledge and case authoring in the tool with the domain expert. Changes to the system can be introduced, when relevant, based on the expert’s feedback.\n\nAfter that, the “Validation” tab in DomainBuilder is used to check for consistency between existing, authored cases and the current knowledge base.\n\nAn alternative process of content creation involves the same steps with one small modification. In this alternative pathway, a user combines steps two and four into a continuous interactive process. “Create Knowledge Base while Authoring Cases” step uses the Case Authoring tab to add new cases. However, since all the required knowledge might not exist during the process of case authoring, the user may add missing concepts to the knowledge base as they are encountered straight from the authoring tab. The advantage of this technique is that it allows users to jump right into case authoring without a need to spend any time developing the underlying knowledge base. The big disadvantage of this technique is that the resulting knowledge base will only fit the set of authored cases, requiring revision of the knowledge base with each newly added case.\n\nTutor builder description The primary purpose of TutorBuilder is to create and customize an ITS configuration. Aspects of the ITS such as type of feedback, user interface as well as pedagogic strategy can be customized using TutorBuilder.\n\nTutor Builder uses a wizard interface to create a configuration and customization for the SlideTutor ITS. SlideTutor is composed of several interlinking modules communicating with each other via a message bus. At any point of the configuration process, TutorBuilder can run a customized version of the ITS directly from the tool to demonstrate results of customization. Figure 5 demonstrates the TutorBuilder interface with advanced configuration options enabled.\n\nThe presentation module presents a tutoring case to a trainee mostly in the form of a virtual slide or a question. The interface module is an interface where a student’s solution to a tutoring case is presented. There are two available interfaces for solving diagnostic problems and one interface for typing a prognostic case report. The reasoning module describes tutor behavior in response to a student’s input. The feedback module allows users to choose the type of feedback given to a student. For example, the tutor can provide immediate feedback to a student’s errors and allow for hints, while delayed feedback displays a student’s errors at the point of submission. The pedagogic module is designed to suggest the next case for a user to solve, while the student module keeps track of a student’s learning progress. The behavior module describes a customizable behavior of the tutoring system when it is interacting with a user. Finally, the protocol module keeps track of all of student and tutor actions and saves them to a database.\n\nThe TutorBuilder workflow consists of the following sequence of steps within a wizard interface, which customizes each aspect of the ITS. A “simple” workflow allows users to create a configuration with minimal customization by selecting from a limited number of pre-configured options, while the “advanced” workflow gives the option to change every aspect of the ITS configuration.\n\n\nResults\n\nDomainBuilder is an authoring system that is designed to be easy-to-use to complete the tasks of knowledge engineering and case authoring. Due to time constraints, the usability study had to be composed of much smaller sub-tasks that could be accomplished by untrained users to test for usability of the system. Good examples of usability problems were linking virtual slide annotations with findings, as well as creating complex rules for inferring diagnoses from a set of findings. Overall, several usability problems focused on small subtasks within the authoring workflow and were targeting users who were not familiar with the system, its architecture, and logic flow. Users were successful in accomplishing the task of completing knowledge engineering tasks. Users demonstrated a general ability to use a highly complex system with no prior instruction and accomplish the tasks given, establishing usability of the system and ability to use it for its intended purposes.\n\nThe DomainBuilder authoring tool usability study had 10 pathologists of varying levels of expertise (4 residents, 3 fellows, 3 practicing pathologists) as participants. Additionally, one of the authors of this manuscript (MC) also informally participated in the study and provided her feedback for system improvement. Participants were given 32 tasks to complete; these tasks were comprised of numerous actions that had to be performed to complete the tasks. Throughout these actions, approximately 95% of the interface was tested. Overall, between 166 and 202 steps were completed within the 32 tasks. A particular action was considered a usability problem for a user if one of the following criteria were met: (i) the elapsed time on task was greater than two minutes, (ii) three or more different actions were attempted for a task, (iii) the user gave up on the task, (iv) the user made a design suggestion, (v) the user showed a negative response to the task, (vi) the user showed surprise, or (vii) the user did not perform the action correctly. An action was considered a major usability problem if two or more users demonstrated a usability problem at that step. A total of 75 actions met the criteria for a usability problem for at least one user. The mean number of usability problems per user was 14 out of an average 187 steps per user (8%) with the highest number of usability problems reported by fellows. There were 30 steps that were reported to have “major” usability problems (at least 2 users with a usability problem).\n\nFollowing the usability tasks, users were given a questionnaire asking their opinions on various aspects of the system. The questionnaire included 29 scaled closed-ended questions on ease of use, understanding, and sufficiency of numerous features of the system and 3 open-ended questions in which users were asked their opinion on what they liked and disliked about the system and suggestions for future additions to the tool.\n\nSeveral actions were taken in response to user feedback. A new feature was added that allowed users to enrich the concept hierarchy with synonyms and definitions from the Unified Medical Language System (UMLS). Another feature that was requested and implemented was the sorting of findings associated with each diagnosis to display differences and similarities between diagnostic parameters.\n\nThe TutorBuilder usability study had a total of six participating pathology fellows. Subjects were given 11 tasks to complete; these tasks were comprised of numerous steps that had to be performed to complete the assignment. The usability problem was defined the same way as with the DomainBuilder study. The average number of usability problems reported per user was 15.3, with 34 steps (34/110 or 31% of steps) that caused usability problems for an average user. There were 24 steps (or 21% of steps) that were “major” usability problems (at least 2 users with a usability problem). Just like with the DomainBuilder system, users were successful in accomplishing the tutor customization tasks and using the system for its intended purposes.\n\nAs part of the user and internal feedback, a simplified user interface was added to the TutorBuilder UI that covered most frequently used tutor configuration options in a single screen, while a more customizable configuration workflow was presented as an “advanced” option.\n\n\nDiscussion\n\nThis paper describes the development of the DomainBuilder and TutorBuilder software systems that streamline and simplify the authoring and medical tutor building process for general medical ITSs. These systems combined knowledge authoring, case authoring, and validation tasks into a single work environment, enabling multiple authoring strategies. While the underlying concepts utilized for the development of these systems have been published in the domain of mathematics and other sciences, DomainBuilder and TutorBuilder are different from these systems in several respects, with consequences for its performance, scalability, extensibility, and customizing capacities for highly specific medical domains.\n\nBoth physical and virtual slide sets are the closest educational resources training pathologists have to learn the domain of pathology. However, these tools are not classified as ITSs as they do not give the user just in time feedback and they do not adequately teach users inference of diagnosis from given set of findings in a case. They also fail to teach the training pathologists how to write complete and accurate prognostic reports.\n\nThe current versions of DomainBuilder and TutorBuilder incorporate the latest technology to ultimately achieve their goals of allowing medical doctors with minimal knowledge engineering training or ITSs background to create a domain knowledge base, to author tutor cases, and to develop novel educational modalities. Between the University of Pittsburgh and the University of Pennsylvania, DomainBuilder has already been used to create more than 20 medical domain ontologies and author more than 500 tutor cases. Two NIH-funded studies have been conducted using the new SlideTutor ITS, which uses cases produced by DomainBuilder42.\n\nSeveral technologies that were developed specifically for DomainBuilder are beneficial for other applications not related to ITSs. The implementation of an independent Ontology API that wraps Protégé OWL and BioPortal 2.0 is used by the ODIE project. Several NLP tools – such as concept coder, reimplementation of string normalization, and negation detection – that were initially developed for DomainBuilder later morphed into the NOBLE project50. The Digital Whole Slide Image server that wraps vendor agnostic OpenSlide library from CMU51,52 is another example of general purpose technologies that stemmed from DomainBuilder development, but could be used anywhere.\n\nThe limitation of the usability studies is the fact that the authoring system has been tested only in the setting of inflammatory dermatological diseases. Future studies will focus on testing the system in other medical domains. An additional limitation of our study is limited number of participants in our studies, which would also be addressed in our future efforts. It is also important to point out that while DomainBuilder and TutorBuilder can cover a wide range of model tracing tutors in the medical domain, they are not as general purpose as some of the other authoring systems reviewed in the introduction. Another limitation of the usability studies outlined in this manuscript is the inability to measure the complete workflow of either DomainBuilder or TutorBuilder systems to assess the quality of user generated content. Because of that, the quality of a user generated tutoring system cannot be assessed.\n\nCurrent versions of DomainBuilder and TutorBuilder accommodate medical domains that require the use of a virtual microscope or another visual representation in cases that require users to identify clinically relevant findings. Instructional materials in other formats are not currently supported by either authoring systems.\n\n\nConclusion\n\nDomainBuilder is a unique authoring framework that creates a systematic workflow to accomplish a highly complex task of content authoring for a medical visual classification ITS. While still in the beta version, it is already used in a research setting, delivering content that is already being used in two major research studies. By allowing instructors to create new curricula for a generic visual classification ITSs, DomainBuilder can address challenges in both the ITSs and medical education fields by delivering systems that can save resources needed for educating highly specialized physicians.\n\nSimilarly, TutorBuilder allows ITSs configuration and customization by non-programmers to fit the educational needs of a teaching curriculum.\n\nAs the domain ontology was developed in the same tool as tutoring cases were being authored, it was possible to author cases without having a complete domain knowledge mapped out and only add representation for components that had cases. Integration of the concept cloud in the validation tab, allowed authors to have a view of the entire domain content in a single screen, which aided in identifying possible authoring errors and generated domain statistics based on the informal impression of the participating authors.\n\nDevelopment of DomainBuilder also contributed to the development of several NLP related technologies, including NOBEL, a flexible concept recognition tool for large-scale biomedical natural language processing50.\n\n\nRecommendations and future work\n\nWe previously reported on the concept of remote pathology consultations with external organizations across the world53. The same virtual slide technology that is a part of DomainBuilder, powered a remote pathology consultation system53. We envision that the systems we describe in this paper as well as their enabling technologies will help with pathology education and potentially pathology consultations across the world. This technology may especially be useful for training medical pathologists focusing on rare diseases or conditions, as traditional apprenticeship based medical educational models may not expose students to important rare disease cases. Between DomainBuilder and TutorBuilder, an educator will be able to create his unique medical visual classification ITSs with his own knowledge representation, cases, tutor feedback mechanism, and message content that will be harnessed to improve pathology education globally.\n\n\nData availability\n\nDataset 1. Results of the usability study and responses to questionnaire. DOI: https://doi.org/10.5256/f1000research.16060.d22185754.\n\n\nSoftware availability\n\nDomainBuilder software is available as a web start application: http://slidetutor.upmc.edu/domainbuilder/.\n\nSource code available from: https://github.com/dbmi-pitt/domainbuilder.\n\nArchived source code at time of publication: https://doi.org/10.5281/zenodo.145895730.\n\nTutorBuilder software is available as a web start application: http://slidetutor.upmc.edu/its.\n\nSource code available from: https://github.com/dbmi-pitt/slidetutor.\n\nArchived source code at time of publication: https://doi.org/10.5281/zenodo.145895931.\n\nLicense: BSD 3-Clause.", "appendix": "Grant information\n\nThe work described in this manuscript has been partially supported in by the National Cancer Institute grant R25CA101959 and Agency for Healthcare Research and Quality grant RR024153. This project was also partially supported by NIH grants 1U24CA180921, R01CA132672, and U54RR023506.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nThe authors would like to thank the project members, students, and medical doctors who participated in this study for their valuable contributions to the design and testing of the systems described in this publication.\n\n\nSupplementary material\n\nSupplementary File 1. DomainBuilder usability tasks.\n\nClick here to access the data\n\nSupplementary File 2. DomainBuilder usability questionnaire.\n\nClick here to access the data\n\n\nReferences\n\nAssociation for Computing Machinery: ACM Digital Library. 2017. Reference Source\n\nVanLehn K: The Relative Effectiveness of Human Tutoring, Intelligent Tutoring Systems, and Other Tutoring Systems. Educ Psychol. 2011; 46(4): 197–221. Publisher Full Text\n\nWolfe CR, Reyna VF, Widmer CL, et al.: Efficacy of a web-based intelligent tutoring system for communicating genetic risk of breast cancer: a fuzzy-trace theory approach. Med Decis Making. 2015; 35(1): 46–59. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMartin B, Mitrovic A: WETAS: A Web-Based Authoring System for Constraint-Based ITS. In: P. De Bra, P. Brusilovsky, R. Conejo (Eds.) Adaptive Hypermedia and Adaptive Web-Based Systems: Second International Conference, AH 2002 Málaga, Spain, May 29–31, 2002 Proceedings, Springer Berlin Heidelberg, Berlin, Heidelberg, 2002; 543–546. Publisher Full Text\n\nWeber G, Kuhl HC, Weibelzahl S: Developing Adaptive Internet Based Courses with the Authoring System NetCoach. In: S. Reich, MM. Tzagarakis, PME. De Bra (Eds.) Hypermedia: Openness, Structural Awareness, and Adaptivity: International Workshops OHS-7, SC-3, and AH-3 Aarhus, Denmark, August 14–18, 2001 Revised Papers, Springer Berlin Heidelberg, Berlin, Heidelberg, 2002; 226–238. Publisher Full Text\n\nAleven V, McLaren BM, Sewall J, et al.: The Cognitive Tutor Authoring Tools (CTAT): Preliminary Evaluation of Efficiency Gains. In: M. Ikeda, KD. Ashley, TW. Chan (Eds.) Intelligent Tutoring Systems: 8th International Conference, ITS 2006, Jhongli, Taiwan, June 26–30, 2006. Proceedings, Springer Berlin Heidelberg, Berlin, Heidelberg, 2006; 61–70. Publisher Full Text\n\nMitrovic A, Martin B, Mayo M: Using Evaluation to Shape ITS Design: Results and Experiences with SQL-Tutor. User Model User-adapt Interact. 2002; 12(2–3): 243–279. Publisher Full Text\n\nMurray T: An Overview of Intelligent Tutoring System Authoring Tools: Updated Analysis of the State of the Art. In: T. Murray, S. Blessing, S. Ainsworth (Eds.) Authoring Tools for Advanced Technology Learning Environments: Toward Cost-effective Adaptive, Interactive, and Intelligent Educational Software Netherlands, Kluwer Academic Publishers, 2003; 491–544. Publisher Full Text\n\nBlessing S, Devasant S, Gilbert S: Evaluation of WebxPST: A browser-based authoring tool for problem specific tutors. In: G. Biswas, S. Bull, J. Kay, A. Mitrovic (Eds.) Proceedings of the 15th International Conference on Artificial Intelligence in Education. LNAI 6738, Springer, Berlin, 2011; 423–425. Publisher Full Text\n\nBlessing SB, Gilbert SB, Ourada S, et al.: Authoring model-tracing cognitive tutors. Int J Artif Intell Educ. 2009; 19: 189–210. Reference Source\n\nMacLellan CJ, Koedinger KR, Matsuda N: Authoring Tutors with SimStudent: An Evaluation of Efficiency and Model Quality. In: S. Trausan-Matu, KE. Boyer, M. Crosby, K. Panourgia (Eds.) Intelligent Tutoring Systems: 12th International Conference, ITS 2014, Honolulu, HI USA, June 5–9, 2014. Proceedings, Springer International Publishing, Cham, 2014; 551–560. Publisher Full Text\n\nMitrovich A, Martin B, Suraweera P, et al.: ASPIRE: An authoring system and deployment environment for constraint-based tutors. Int J Artif Intell Educ. 2009; 19: 155–188. Reference Source\n\nNye BD, Graesser AC, Hu X: AutoTutor and Family: A Review of 17 Years of Natural Language Tutoring. Int J Artif Intell Educ. 2014; 24(4): 427–469. Publisher Full Text\n\nPaquette L, Lebeau JF, Beaulieu G, et al.: Designing a Knowledge Representation Approach for the Generation of Pedagogical Interventions by MTTs. Int J Artif Intell Educ. 2015; 25(1): 118–156. Publisher Full Text\n\nRazzaq L, Patvarczki J, Almeida SF, et al.: The ASSISTment Builder: Supporting the Life Cycle of Tutoring System Content Creation. IEEE Transactions on Learning Technologies Special Issue on Real-World Applications of Intelligent Tutoring Systems. 2009; 2(2): 157–166. Publisher Full Text\n\nSottilare R: Considerations in the development of an ontology for a generalized intelligent framework for tutoring. I3M Defense and Homeland Security Simulation Conference, DHSS, Vienna, Austria, 2012. Publisher Full Text\n\nBlessing SB, Aleven V, Gilbert SB, et al.: Authoring Example-based Tutors for Procedural Tasks. In: R. Sottilare, A. Graesser, X. Hu, K. Brawner (Eds.) Design Recommendations for Adaptive Intelligent Tutoring Systems: Authoring Tools, U.S. Army Research Laboratory, Orlando, FL, 2015; 71–93. Reference Source\n\nAleven V, Roll I, McLaren BM, et al.: Help Helps, But Only So Much: Research on Help Seeking with Intelligent Tutoring Systems. Int J Artif Intell Educ. 2016; 26(1): 205–223. Publisher Full Text\n\nSottilare RA, Brawner KW, Sinatra AM, et al.: An Updated Concept for a Generalized Intelligent Framework for Tutoring (GIFT). In: U.A.R.L.H.R.E.D. (ARL-HRED) (Ed.), US Army Research Laboratory, Orlando, FL, 2017. Publisher Full Text\n\nGraesser AC: Conversations with AutoTutor Help Students Learn. Int J Artif Intell Educ. 2016; 26(1): 124–132. Publisher Full Text\n\nIsotani S, Inaba A, Ikeda M, et al.: An ontology engineering approach to the realization of theory-driven group formation. Int J Comput Support Collab Learn. 2009; 4(4): 445–478. Publisher Full Text\n\nIsotani S, Mizoguchi R: Adventures in the Boundary between Domain-Independent Ontologies and Domain Content for CSCL. In: I. Lovrek, RJ. Howlett, LC. Jain (Eds.) Knowledge-Based Intelligent Information and Engineering Systems: 12th International Conference. KES, 2008, Zagreb, Croatia, September 3–5, 2008, Proceedings, Part III Springer Berlin Heidelberg, Berlin, Heidelberg, 2008; 523–532. Publisher Full Text\n\nCrowley RS, Medvedeva O: An intelligent tutoring system for visual classification problem solving. Artif Intell Med. 2006; 36(1): 85–117. PubMed Abstract \| Publisher Full Text\n\nEl Saadawi GM, Tseytlin E, Legowski E, et al.: A natural language intelligent tutoring system for training pathologists: implementation and evaluation. Adv Health Sci Educ Theory Pract. 2008; 13(5): 709–722. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCrowley R, Legowski E, Medvedeva O, et al.: An ITS for medical classification problem-solving: Effects of tutoring and representations. In: CK. Looi, G. McCalla, B. Bredeweg, J. Breuker (Eds.) Artificial Intelligence in Education, IOS Press, Amsterdam, The Netherlands, 2005; 192–199. Reference Source\n\nCrowley RS, Tseytlin E, Jukic D: ReportTutor - an intelligent tutoring system that uses a natural language interface. AMIA Annu Symp Proc. 2005; 171–175. PubMed Abstract \| Free Full Text\n\nYudelson MV, Medvedeva O, Legowski E, et al.: Mining student learning data to develop high level pedagogic strategy in a medical ITS. Workshop on Education Data Mining at AAAI 2006, AAAI, Menlo Park, CA, 2006. Reference Source\n\nCrowley R, Grzybicki D, Legowski E, et al.: Use of a Medical ITS Improves Reporting Performance among Community Pathologists. In: V. Aleven, J. Kay, J. Mostow (Eds.) Intelligent Tutoring Systems: 10th International Conference, ITS 2010, Pittsburgh, PA USA, June 14–18, 2010, Proceedings, Part I Springer Berlin Heidelberg, Berlin, Heidelberg, 2010; 338–348. Publisher Full Text\n\nNoy NF, Sintek M, Decker S, et al.: Creating Semantic Web contents with Protege-2000. IEEE Intell Syst. 2001; 16(2): 60–71. Publisher Full Text\n\nTseytlin E: dbmi-pitt/domainbuilder: Final Release (Version 1.0). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1458957\n\nTseytlin E: dbmi-pitt/slidetutor: Final Release (Version 1.0). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1458959\n\nCrowley RS, Medvedeva O: A general architecture for intelligent tutoring of diagnostic classification problem solving. AMIA Annu Symp Proc. 2003; 2003: 185–189. PubMed Abstract \| Free Full Text\n\nCrowley RS, Legowski E, Medvedeva O, et al.: Automated detection of heuristics and biases among pathologists in a computer-based system. Adv Health Sci Educ Theory Pract. 2013; 18(3): 343–363. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nEl Saadawi GM, Azevedo R, Castine M, et al.: Factors affecting feeling-of-knowing in a medical intelligent tutoring system: the role of immediate feedback as a metacognitive scaffold. Adv Health Sci Educ Theory Pract. 2010; 15(1): 9–30. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFeyzi-Behnagh R, Azevedo R, Legowski E, et al.: Metacognitive Scaffolds Improve Self-Judgments of Accuracy in a Medical Intelligent Tutoring System. Instr Sci. 2014; 42(2): 159–181. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPayne VL, Medvedeva O, Legowski E, et al.: Effect of a limited-enforcement intelligent tutoring system in dermatopathology on student errors, goals and solution paths. Artif Intell Med. 2009; 47(3): 175–197. PubMed Abstract \| Publisher Full Text\n\nPlaisant C, Grosjean J, Bederson BB: SpaceTree: supporting exploration in large node link tree, design evolution and empirical evaluation. IEEE Symposium on Information Visualization, 2002. INFOVIS. 2002; 57–64. Publisher Full Text\n\nNational Center for Biomedical Ontology: BioPortal. 2017. Reference Source\n\nNational Cancer Institute: Enterprise Vocabulary Services. 2017. Reference Source\n\nU.S. National Library of Medicine: Unified Medical Language System® (UMLS®). 2017. Reference Source\n\nMurray T: Authoring intelligent tutoring systems: An analysis of the state of the art. Int J Artif Intell Educ. 1999; 10: 98–129. Reference Source\n\nTseytlin E, Castine M, Crowley R: DomainBuilder – An Authoring System for Visual Classification Tutoring Systems. In: V Aleven, J Kay, J Mostow (Eds.) Intelligent Tutoring Systems: 10th International Conference, ITS 2010, Pittsburgh, PA USA, June 14–18, 2010, Proceedings, Part II Springer Berlin Heidelberg, Berlin, Heidelberg, 2010; 441–442. Publisher Full Text\n\nKirkpatrick DL: Techniques for Evaluation Training Programs. Journal of the American Society of Training Directors. 1959; 13: 21–26. Reference Source\n\nHolton EF: The flawed four-level evaluation model. Hum Resour Dev Q. 1996; 7(1): 5–21. Publisher Full Text\n\nBechhofer S: OWL: Web Ontology Language. In: L Liu, MT ÖZsu (Eds.) Encyclopedia of Database Systems, Springer US Boston, MA, 2009; 2008–2009. Publisher Full Text\n\nUniversity of Maryland Human Computer Interaction Lab: SpaceTree. SpaceTree is a novel tree browser that builds on the conventional layout node link diagrams along a single preferred direction. 2017. Reference Source\n\nBodenreider O: Lexical, terminological and ontological resources for biological text mining. In: S. Ananiadou, J. McNaught (Eds.) Text mining for biology and biomedicine, Artech House, 2006; 43–66. Reference Source\n\nZou Q, Chu WW, Morioka C, et al.: IndexFinder: a method of extracting key concepts from clinical texts for indexing. AMIA Annu Symp Proc. 2003; 2003: 763–767. PubMed Abstract \| Free Full Text\n\nTag cloud. 2017. Reference Source\n\nTseytlin E, Mitchell K, Legowski E, et al.: NOBLE - Flexible concept recognition for large-scale biomedical natural language processing. BMC Bioinformatics. 2016; 17: 32. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nOpenSlide. 2017. Reference Source\n\nGoode A, Gilbert B, Harkes J, et al.: OpenSlide: A vendor-neutral software foundation for digital pathology. J Pathol Inform. 2013; 4: 27. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRomero Lauro G, Cable W, Lesniak A, et al.: Digital pathology consultations-a new era in digital imaging, challenges and practical applications. J Digit Imaging. 2013; 26(4): 668–677. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTseytlin E, Linkov F, Castine M, et al.: Dataset 1 in: DomainBuilder: the knowledge authoring system for SlideTutor Intelligent Tutoring system. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16060.d221857" }	[ { "id": "40040", "date": "12 Nov 2018", "name": "Ana Marusic", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis study presents the creation of the DomainBuilder, an Intelligent Tutoring System (ITS) tool that allows users to apply a systematic workflow to the complex task of content authoring for a medical visual classification IT such as occurs in pathology. The study presents clearly the state of the art in the field, and describes the preliminary studies of the authors to then present the objectives of the study, as well as its innovativeness. In the Methods section, the authors describe in detail the development process for the individual components and processes of the tool and their implementation. The usability study described in the results clearly present the testing with 10 pathologist.\n\nMinor comment: the authors present the mean number of usability problems per user - if means are used to present quantitative date, the a measure of variability should also be presented.\n\nThe Discussion section presents the summary of the finding, their limitations and possible uses. It will be interesting to follow the success of the tool in actual practice.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "44626", "date": "05 Mar 2019", "name": "Claudia Mello-Thoms", "expertise": [ "Reviewer Expertise Image perception", "observer performance", "technology assessment" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this paper, the authors describe the development of two powerful tools to aid in the implementation of Intelligent Tutoring Systems (ITS) in the medical domain. The first tool, named DomainBuilder, combines knowledge authoring, case authoring and validation tasks into a single environment, whereas the second tool, TutorBuilder, allows users to select and customize different aspects of the design on an ITS, such as the user interface, pedagogic and feedback modules, etc. While the tools, as proposed, are clearly powerful and should be well received by the research community, there are some aspects of the paper that are less clear. For example, in several parts the authors claim that these tools are useful for “general medical ITS”, whereas in other parts they restrict the use of the tools to the domain of Pathology and their SlideTutor ITS. This dichotomy permeates the paper, and it is at times difficult to tell whether the authors are talking about DomainBuilder, TutorBuilder, or SlideTutor. For example, in a subheading reading “Operation”, it seems like the authors are discussing SlideTutor, not the other two systems. In addition, I would have encouraged the authors to include more “Results” in the appropriate section, as the responses to the questionnaire and the usability study are presented as an aside dataset. Finally, several projects (which have apparently benefited from the development of DomainBuilder and TutorBuilder) are named, such as ODIE and NOBLE, but the reader does not benefit from this information as no description is given of these projects.\n\nIs the work clearly and accurately presented and does it cite the current literature? No\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/7-1721
https://f1000research.com/articles/7-1714/v1	29 Oct 18	{ "type": "Research Article", "title": "Prostate cancer awareness at Brigham Young University of Idaho: A cross-sectional study", "authors": [ "Alain Mwamba Mukendi", "Drew Jenks", "Hannah Moore", "Victoria Ackerman", "Drew Jenks", "Hannah Moore", "Victoria Ackerman" ], "abstract": "Background: Prostate cancer is the second leading cause of cancer death in American men, and is very common in older men. Early screenings have been proven to help diagnose prostate cancer sooner. Ignorance about prostate cancer can be a huge problem impeding men from getting screened. Hence, it is important to be aware of the disease and encourage prostate cancer screening by age 50. The purpose of this study was to establish the level of awareness of prostate cancer among college students at Brigham Young University of Idaho (BYU-I). Methods: This survey research was conducted at BYU-I. Questionnaires were sent via email. Responses were received the same way and analyzed using SPSS. Results: The study shows that knowledge about prostate cancer varied greatly among BYU-I students. The level of awareness is poor and is not correlated to gender or age. Conclusion: This study shows a significant lack of awareness of prostate cancer among BYU-I students. Necessary steps should be taken to promote more awareness and early screening for prostate cancer in this setting. Educational opportunities should be offered for recognition of symptoms and to promote screening which will lead to early diagnosis and treatment.", "keywords": [ "prostate cancer", "awareness", "university students" ], "content": "Introduction\n\nProstate cancer screening continues to be a huge controversial topic in our society mostly because of the benefit of diagnosing it versus the risk of overtreating it1. Prostate cancer is not the deadliest type of cancer, but it is very common in older men. Early screenings have been proven to help diagnose prostate cancer sooner1. Ignorance about prostate cancer can be a huge problem impeding men from getting screened. According to The American Cancer Society, each year more than 30,000 men die of this condition1, making it the second deadliest cancer for American men. However, the American Cancer Society also stated that an early diagnosis yields a five-year survival rate of almost 100%. Hence, it is important to be aware of the disease and encourage prostate cancer screening by age 501.\n\nThis study was performed in order to establish the level of awareness of prostate cancer among college students at Brigham Young University of Idaho (BYU-I).\n\n\nMethods\n\nOur study focused on students attending BYU-I and was conducted in February 2016 over a period of nine days. The study received approval from the Brigham Young University Institutional Review Board for Human Subjects (IRB) under reference W16-175.\n\nThe majority of students attending BYU-I have backgrounds from all over the country and most students are under the age of 30 with a few outliers. There were 17,562 students enrolled at BYU-I with 9,278 of those students being women and 8,284 men2.\n\nFor our study, we used Google Forms to create a 7-question survey and used BYU-Idaho Student Research to disperse it to BYU-Idaho students’ emails. If a participant completed and sent the questionnaire back then that was considered to be consent for participation in the study.\n\nWe initially had 12 questions and decided to only use 5 of these, besides age and gender, to make the survey as short and purposeful as possible. There was no exclusion criteria to participate in the survey. Each participant received an email accompanying the survey explaining the purpose of this research and clearly stated that participation was voluntary and that the data would be confidential and anonymous (Supplementary File 1). To avoid any bias, efforts were made to have questions clear and understandable, well-structured, logical and short. Two questions used a scale and the rest required a Yes/No/I don’t know answer. After a little over one week of compiling we received 55 responses.\n\nData collected by the survey was age, gender and answers to the following questions: Have you heard of prostate cancer?; Do you know or have you heard of someone suffering from prostate cancer? How familiar are you with prostate cancer?; Is one of the following symptoms of prostate cancer? (Pain, Trouble urinating, Blood in urine); Is one of the following treatments for prostate cancer? (Surgery, Radiation Therapy, Anti Hormonal therapy).\n\nWe used SPSS for data analysis. We analyzed the data to measure prostate cancer awareness in BYU-Idaho students. Using the data, we used statistics to see how familiar BYU-Idaho students are with prostate cancer. We performed a Pearson correlation coefficient test with the level of significance of 0.001 to see if there was a correlation between gender and prostate cancer awareness. We created bar graphs displaying these statistical analyses. We performed the same test to see if there was a correlation between age and prostate cancer awareness.\n\n\nResults\n\nIn total, 17,562 students were eligible to participate in the survey. Only 55 responses were received, of which 7 had missing data and were not included in the analysis. The final data included 48 participants aged from 18 to 59 (years). 31 participants were women and 17 were men. Our sample matched the age of the BYU-I student population with a mean age of 25.5 years old, and a median of 22 years old. The study shows that knowledge about prostate cancer varied greatly among BYU-I students. Although this data represented the student body at BYU-I well, there was no correlation between age and familiarity with prostate cancer (Table 1).\n\nWe also tested to see whether or not gender and familiarity with prostate cancer would yield a correlation (Table 2). The 0.100 in Table 2 shows the correlation is very weak. Women reported to have more familiarity with prostate cancer than men. The woman who ranked her familiarity with prostate cancer the highest used an 8 on a scale of 1 to 10. The most common answers were 2, 3, and 4. These numbers are very low compared to the amount of prostate cancer that occurs in the population. It was interesting to see that although prostate cancer affects men, in our study fewer men had heard of prostate cancer than women (Figure 1).\n\nGreen, women; blue, men.\n\n\nDiscussion\n\nProstate cancer is a significant public health issue in the United States. The American Cancer Society estimates that in 2011, about 240,890 men were diagnosed with prostate cancer and 33,720 men died from it3.\n\nResearch is battling against the rising issue of prostate cancer. Attempts to increase awareness have yielded successful results. Prostate Cancer Awareness Week is held every third week of September by the Prostate Cancer Education Council, which offers a great opportunity to the community in the US to learn more about prostate cancer and to encourage early detection and diagnosis of the disease4. Prostate cancer awareness has been one of the top priorities in order to cure or treat prostate cancer effectively5. The Journal of Cancer Education has stated that young adults’ knowledge of cancer is not a greatly researched area5. This disease has claimed many lives so far. The aim of this study was to highlight the level of awareness in the population of students at BYU-I, with the hope to increase knowledge of how devastating the disease is and how manageable it is with the proper treatment.\n\nThe present survey showed clear evidence of the lack of prostate cancer awareness among college students at BYU-I. Assuming that most college campuses are similar to BYU-I, the results are most likely representative of prostate cancer awareness in other undergraduate college campuses; however, due to the lack of a substantial number of responses, this data needs further investigation to be representational of under-graduate college campuses throughout the United States. Varying results will occur depending on the size of the university, environment, and whether or not the university is for post-graduates. College majors and student backgrounds may have a huge effect on college students’ awareness of prostate cancer. The awareness of this disease is not high among college aged students attending BYU-I; therefore, necessary steps should be taken to promote more awareness and early screening for prostate cancer in this setting.\n\nThe survey only captured 55 responses out of about 18000 students, so we do not know the experiences of the majority of the student body. Time constraint were a huge factor as this was a university assignment with deadline for submission.\n\n\nConclusion\n\nThis study shows a significant lack of awareness of prostate cancer among Brigham Young University- Idaho students. Necessary steps should be taken to promote more awareness and early screening for prostate cancer. Educational opportunities should be offered for recognition of symptoms and to promote early diagnosis for a potential curative management.\n\n\nData availability\n\nF1000Research: Dataset 1. Survey data on prostate cancer awareness at Brigham Young University of Idaho. , 10.5256/f1000research.16566.d2224596.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nSupplementary material\n\nSupplementary File 1: Questionnaire and key for dataset.\n\nClick here to access the data\n\n\nReferences\n\nHarris R, Lohr KN: Screening for prostate cancer: an update of the evidence for the U.S. Preventive Services Task Force. Ann Intern Med. 2002; 137(11): 917–929. PubMed Abstract \| Publisher Full Text\n\nCrandall B: BYU-Idaho Reaches Largest Enrollment in University History. BYU Idaho newsroom, 2015. Reference Source\n\nBrawley OW: Prostate cancer epidemiology in the United States. World J Urol. 2012; 30(2): 195–200. PubMed Abstract \| Publisher Full Text\n\nCrawford ED: Prostate cancer awareness: Much has changed since '89. Urology Times. 2005. Reference Source\n\nCampbell LC, McClain J: Exploring prostate cancer literacy and family cancer awareness in college students: getting ahead of the curve in cancer education. J Cancer Educ. 2013; 28(4): 617–22. PubMed Abstract \| Publisher Full Text\n\nMukendi AM, Jenks D, Moore H, et al.: Dataset 1 in: Prostate cancer awareness at Brigham Young University of Idaho: A cross-sectional study. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16566.d222459" }	[ { "id": "39974", "date": "05 Nov 2018", "name": "Noor N.P. Buchholz", "expertise": [ "Reviewer Expertise Urology" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nINTRODUCTION:\nThe introduction gives a generally confusing background that doesn’t focus on the object of the study.\n\nThe statement “Prostate cancer is not the deadliest type of cancer” is incorrect. Prostate cancer is in fact the most common cancer in men and the second leading cause of death after lung cancer.\nI would state pragmatically:\nProstate cancer is the most common …….. Early diagnosis lead to ……even if the screening trials have shown conflicting results……overtreatment…… However, the prostate cancer awareness is another important…….it might allow to optimize an opportunistic screening…… The prostate cancer awareness has been spread in the last years by …..month awareness….. However the effective level of awareness in the population is difficult to assess……… This study was performed in order to establish the level of awareness of prostate cancer among college students at Brigham Young University of Idaho (BYU-I).\n\nMETHODS:\n\nIn the methods section we usually specify only the methods we use to reach the results, how we select the population study, the variables, characteristics and statistical method. We don’t speak about numbers which have to be mentioned in the results section:\n“After a little over one week of compiling we received 55 responses”.\n\n“We initially had 12 questions and decided to only use 5 of these, besides age and gender, to make the survey as short and purposeful as possible.”\n\n“To avoid any bias, efforts were made to have questions clear and understandable, well-structured, logical and sh…”\nThe above statements are discussion statements; they are not suitable to be put in the M&M section.\nThe questionnaire includes 5 empiric questions:\nHave you heard of prostate cancer? Do you know or have you heard of someone suffering from prostate cancer? How familiar are you with prostate cancer? Is one of the following symptoms of prostate cancer? (Pain, Trouble urinating, Blood in urine). Is one of the following treatments for prostate cancer? (Surgery, Radiation Therapy, Anti Hormonal therapy).\n\nNo one of these regard screening and early detection, which is the most important target of cancer awareness. Questions 1, 2 and 3 have the same meaning. In particular, question 3 doesn’t allow a ‘yes/no/I don’t know’ answer. Question 4 is completely useless because early stage curable prostate cancer is not symptomatic. Question 5 has to be addressed to healthcare professionals.\n\nRESULTS:\n\nI cannot discuss the results but a statistical analysis is very hard in the people who gave feedback. The sample is very limited.\nThe variables have to be described as mean, range, SD.\n\nDISCUSSION:\n\nThe authors do a dissertation about prostate cancer awareness. I instead would have discussed the results of this study:\nLow number of feedbacks. Methodology chosen to reach the people has to be discussed because it led to very poor results. The statement “It was interesting to see that although prostate cancer affects men, in our study fewer men had heard of prostate cancer than women” has to be discussed. “The present survey showed clear evidence of the lack of prostate cancer awareness among college students at BYU-I”. It is not a clear evidence because these results could be due to an insufficient methodology…\n\nCONCLUSIONS:\n\nThe conclusions are not supported.\n\nIn my opinion this article is not suitable for indexing.", "responses": [ { "c_id": "4167", "date": "06 Nov 2018", "name": "Alain Mwamba Mukendi", "role": "Author Response", "response": "Thank you so much for a thorough review of this research note. I really appreciate your inputs. I will consider every single point you raised to improve the quality of this paper that I did not want to waste as a simple assignment. Would you be able to review the next version?" } ] }, { "id": "39978", "date": "16 Nov 2018", "name": "Nelson Bennett", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThank you for the opportunity to review this manuscript. All comments and suggestions are meant to enhance the quality of this manuscript and future research by your group.\nIt is apparent that a significant amount of time has been spent by the authors in the conception, data collection and authoring of this manuscript.\n\nIntroduction:\nThe introduction should include background information and set up the purpose of the study. While prostate cancer is important to know about as adults, it is not something that college-aged people know much about. This is because the only physicians that this age group has seen are most likely pediatricians. I do not have a clear understanding of what this study is attempting to accomplish.\n\nMethods/Results:\n\nSurvey projects are tricky and rife with pitfalls. The most glaring issue is the very poor response rate. The results are based on a response rate of 0.3%. It is very difficult to accept the results as valid unless the response rate is >30%.\n\nThe other issue with survey projects has to do with the questions asked. If the right questions are not asked in the right way, the results will be invalid. Additionally, if the questionnaire is not formally validated, then the results tend to change even if applied to the same group. This manuscript unfortunately suffers from these issues.\n\nDiscussion:\nIn the first paragraph of the discussion, the author is supposed to summarize the results. In the next 2-3 paragraphs, the author is charged with picking 1-2 important findings from the study and providing further commentary with comparisons to existing literature.", "responses": [ { "c_id": "4239", "date": "16 Nov 2018", "name": "Alain Mwamba Mukendi", "role": "Author Response", "response": "Thank you so much for taking time to review this article. We will address your concerns to improve the quality of the article. " } ] } ]	1	https://f1000research.com/articles/7-1714
https://f1000research.com/articles/7-1704/v1	26 Oct 18	{ "type": "Research Article", "title": "Screening of protease, cellulase, amylase and xylanase from the salt-tolerant and thermostable marine Bacillus subtilis strain SR60", "authors": [ "Bruno Oliveira de Veras", "Yago Queiroz dos Santos", "Katharina Marquez Diniz", "Gabriela Silva Campos Carelli", "Elizeu Antunes dos Santos", "Yago Queiroz dos Santos", "Katharina Marquez Diniz", "Gabriela Silva Campos Carelli", "Elizeu Antunes dos Santos" ], "abstract": "Background: The marine environment harbours different microorganisms that inhabit niches with adverse conditions, such as temperature variation, pressure and salinity. To survive these particular conditions, marine bacteria use unique metabolic and biochemical features, producing enzymes that may have industrial value. Methods: The aim of this study was to observe the production of multiple thermoenzymes and haloenzymes, including protease, cellulase, amylase and xylanase, from bacterial strains isolated from coral reefs Cabo Branco, Paraiba State, Brazil. Strain SR60 was identified by the phylogenetic analysis to be Bacillus subtilis through a 16S ribosomal RNA assay. To screening of multiples enzymes B. subtilis SR60 was inoculated in differential media to elicit the production of extracellular enzymes with the addition of a range of salt concentrations (0, 0.25, 0.50, 1.0, 1.25 and 1.5 M NaCl). Results: The screening showed a capacity of production of halotolerant protease, cellulase, amylase and xylanase and thermostable by the isolate (identified as B. subtilis SR60). Protease, cellulase, amylase and xylanase production were limited to 1.5, 1.5, 1.0 and 1.25 M NaCl, respectively. Conclusions: Bacillus subtilis SR60 was shown in this study be capable of producing protease, cellulase, amylase and xylanase when submitted to a high salinity environment. These data demonstrate the halophytic nature of SR60 and its ability to produce multiples enzymes.", "keywords": [ "Bacteria", "Thermoenzymes", "Haloenzyme", "Enzymes", "Industrial Applications." ], "content": "Introduction\n\nCovering large surface of the Earth's surface, the marine environment is a rich source of biological and chemical diversity; it contains endless habitats that may present adverse conditions of survival. However, these conditions favour the establishment of microorganisms able to produce enzymes that have extraordinary properties, such as salt tolerance, thermostability, pH and temperature variations. These enzymes have many industrial applications, such as the production of detergents, food, feed, pharmaceuticals, leather and biofuel1,2.\n\nThe conditions of the industrial scale activities are related to the maintenance of enzymatic activity in environments with variations in temperature (55°C to 121°C and -2°C to 20°C), pressure (> 500 atmospheres), pH (pH> 8, pH <4) and salinity (1–5 M NaCl or KCl)3. The production of enzymes of bacterial origin is a frequent application of industrial biotechnology; the enzymes produced include hydrolytic thermostable enzymes such as amylases, cellulases, proteases and xylanases for the production of biofuel4. Use of the genus Bacillus is promising for the production of biomolecules, because it is classified by the FDA as being generally recognized as safe and research has revealed the ability of this genus to produce and secrete enzymes with infinite applications5.\n\nThis study aimed to produce multiple thermoenzymes and haloenzymes (protease, cellulase, amylase and xylanase) expressed by Bacillus subtilis strain SR60, a bacterial symbiont isolated from Siderastrea stellate (Verrill, 1868) in a Brazilian coral reefs ecosystem 7°08’50” S; 34° 47’51” W.\n\n\nMethods\n\nThe bacterial strains were obtained from aseptically collected tissues of Siderastrea stellate Verrill, 1868 (Cnidaria, Scleractinia) colonies at Cabo Branco coral reefs, Paraiba State, Brazil (7°08’50” S; 34°47’51” W). For bacterial isolation from the anthozoan, samples were suspended in sterile saline solution, agitated until homogenization was achieved and then spread over marine agar plates (pH 8.0± 0.3) containing 5 g/l peptone; 1 g/l yeast extract; 15 g/l agar diluted in sterile marine water and incubated at 55°C until adequate growth was achieved6. A total of 12 bacterial isolates were obtained, which were analysed for protease, cellulase, amylase and xylanase production capacity, and only the one with the simultaneous production capacity of these enzymes was selected.\n\nFor further screening of enzymatic activity described below, two bacterial colonies, isolated using the above culturing conditions, were inoculated onto each plate. A total of three replicates were performed for each salt molarity.\n\nIn order to identify the isolate, morphophysiological and molecular data were evaluated7. The obtained 16S rRNA gene was sequenced by ATCGene (UFRGS, Porto Alegre, RS, Brazil) using the automated sequencer ABI-PRISM 3100 Genetic Analyzer. The SR60 isolate sequence was compared to sequences deposited in the Genbank database (NCBI). For the local alignment, the BLASTn tool (NCBI) was used. MEGA 6.0 software was used for monitoring multiple sequences and for construction of a dendrogram by the Neighbor-Joining method.\n\nThe isolated bacterial strains were screened production for protease on agar medium comprising 10 g/l gelatine and 20 g/l agar in increasing concentrations of NaCl (0, 0.25, 0.50, 1.0, 1.25 and 1.5 M) pH 8.0± 0.3. The inoculated plates were incubated at 48 h at 55°C and observed for the formation of zone of hydrolysis8.\n\nThe ability of isolate on produce cellulose was tested a plate containing 1 g/l carboxymethylcellulose (CMC); 0.5 g/l NaNO3; 1 g/l K2HPO4; 0.5 g/l MgSO4∙7H2O; 0.001 g/l FeSO4∙7H2O; 1 g/l yeast extract; 15 g/l agar) in increasing molarities NaCl (0, 0.25, 0.50, 1.0, 1.25 and 1.5 M) for 48 h at 55°C on pH 8.0±0.3 and then overlaid with 0.2 g/l potassium iodide for 5 min, bacterial colonies showing clear zones were considered to be cellulase producers9.\n\nAmylolytic activity of culture was screened on starch nutrient agar plates containing: 10 g/l starch; 0.05 g/l NaNO3; 1 g/l K2HPO4; 0.5 g/l MgSO4∙7H2O; 0.001 g/l FeSO4∙7H2O; 1 g/l yeast extract; 15 g/l agar, in increasing molarities of NaCl (0, 0.25, 0.50, 1.0, 1.25 and 1.5 M). After incubation at 55°C pH 8.0±0.3 for 48 h, the zone of clearance was determined by flooding the plates with 0.2 g/l potassium iodide for 5 min10.\n\nXylanase activity was detected using a saline medium containing: (10 g/l xylan; 0.005 g/l NaNO3; 1 g/l K2HPO4; 0.5 g/l MgSO4∙7H2O; 0,001 g/l FeSO4∙7H2O; 1 g/l yeast extract; 15 g/l agar) in increasing molarities of NaCl (0, 0.25, 0.50, 1.0 and 1.5 M) on pH 8.0±0.3. After incubation at 55°C for 48 h, the plates were with 0.2 g/l potassium iodide for 5 min. The clear zones around colonies indicated qualitative xylanase activity11.\n\n\nResults and discussion\n\nThe SR60 isolate was revealed to be a Gram-positive spore-forming bacillus, facultative anaerobe, catalase-positive; it was negative for indole, H2S production and citrate utilization bacterium (Table 1). Those findings led us to consider the isolate belonging to the genus Bacillus which was posteriorly confirmed by the phylogenetic analysis which revealed that the SR60 strain formed a clade with Bacillus subtilis (Figure 1). The nucleotide sequence was deposited in GenBank under accession number MH698455.1.\n\nThe scale bar represents 0.01 substitutions per site. GenBank accession numbers of the sequences are given in parentheses.\n\nIn differential media for the production of different extracellular enzymes, it was observed that conditions of high salinity from 0 to 1.5 M NaCl, a SR60 strain showed proteolytic, cellulolytic, aminolytic and xylanolytic activity, these productions being observed by zones of enzymatic hydrolysis (Table 2). The halo detection for protease and cellulase was observed up to the maximum salinity, 1.5 M NaCl (Figure 2 and Figure 3). Cellulolytic enzymes comprise a group of glycosidic hydrolases, including endoglucanases, exoglucanases and beta-glycosidase. In general, the production of the enzyme group is mainly observed in fungi, actinomycetes and some other bacteria. The use of fungi to produce cellulases has been practiced in the food, textile, fuel and chemical industry, but the growth period for the microorganism does not match the high demand from the industries for production. In an attempt to solve this problem bacteria present rapid growth and high enzymatic production12. Bacterial isolates produced from different environments, such as bovine ruminants, soil and in isolation, were found to produce hydrolases12,13. Biofuel industries that use lignocellulose as the first raw material pre-treatment process for the release of cellulose, making it more accessible to the enzymatic action. During the processing of the cellulose, various compounds containing salts are used, the enzymatic catalysis being reduced or inhibited in this halophilic environment15. The extracellular production of amylase and xylanase reached an upper NaCl concentration limit of 1.0 M and 1.25 M NaCl, respectively (Figure 4 and Figure 5); however, as a bacterial cell growth molecule at the other salt concentrations.\n\nHalos around bacterial colonies are indicative of cellulose degradation.\n\nHalos around bacterial colonies are indicative of cellulose degradation.\n\nHalos around bacterial colonies are indicative of cellulose degradation.\n\nHalos around bacterial colonies are indicative of cellulose degradation.\n\n\nConclusions\n\nThe Bacillus sp. isolate identified in this study, Bacillus subtilis SR60, has the capacity for proteases, cellulases, amylases and xylanases with thermostable and halotolerant characteristics. These products can be used as thermostable enzymes in the production of biofuels in crucial stages of this bioprocess.\n\n\nData availability\n\nThe sequence of the Bacillus subtilis strain SR60 16s RNA gene isolated in this experiment is available from GenBank, accession number MH698455.1: https://identifiers.org/ncbigi/GI:1435753077.\n\nImages of the repeats of the screening for enzymatic activity have been uploaded to Harvard Dataverse, DOI: https://doi.org/10.7910/DVN/J5JCC0. Data are available under the terms of the Creative Commons Zero “No rights reserved” data waiver (CC0 1.0 Public domain dedication.", "appendix": "Grant information\n\nThis work was supported in part by the Federal University of Pernambuco.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nHu Y, Chen J, Hu G, et al.: Statistical research on the bioactivity of new marine natural products discovered during the 28 years from 1985 to 2012. Mar Drugs. 2015; 13(1): 202–221. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIzadpanah Qeshmi F, Homaei A, Fernandes P, et al.: Marine microbial L-asparaginase: Biochemistry, molecular approaches and applications in tumor therapy and in food industry. Microbiol Res. 2018; 208: 99–112. PubMed Abstract \| Publisher Full Text\n\nDumorné K, Córdova DC, Astorga-Eló M, et al.: Extremozymes: A Potential Source for Industrial Applications. J Microbiol Biotechnol. 2017; 27(4): 649–659. PubMed Abstract \| Publisher Full Text\n\nSrivastava A, Verma J, Singh H, et al.: Screening of biologically active microbial strains having therapeutic applications. Indian J Exp Biol. 2018; 56: 244–251. Reference Source\n\nHadjidj R, Badis A, Mechri S, et al.: Purification, biochemical, and molecular characterization of novel protease from Bacillus licheniformis strain K7A. Int J Biol Macromol. 2018; 114: 1033–1048. PubMed Abstract \| Publisher Full Text\n\nDustan P: Distribution of zooxanthella and photosynthetic chloroplast pigment of the reef building coral Montastrea annularis Ellis and Solander in relation to depth on a West Indian coral reef. Bull Mar Sci. 1979; 29(1): 79–95. Reference Source\n\nHogg JC, Lehane MJ: Identification of bacterial species associated with the sheep scab mite (Psoroptes ovis) by using amplified genes coding for 16S rRNA. Appl Environ Microbiol. 1999; 65(9): 4227–4229. PubMed Abstract \| Free Full Text\n\nFogarty WM, Kelly CT: Microbial Enzymes and Biotechnology. In: Microbial Enzymes and Biotechnology. 1990. Publisher Full Text\n\nKasana RC, Salwan R, Dhar H, et al.: A rapid and easy method for the detection of microbial cellulases on agar plates using Gram’s iodine. Curr Microbiol. 2008; 57(5): 503–507. PubMed Abstract \| Publisher Full Text\n\nAmoozegar MA, Malekzadeh F, Malik KA: Production of amylase by newly isolated moderate halophile, Halobacillus sp. Strain MA-2. J Microbiol Methods. 2003; 52(3): 353–359. PubMed Abstract \| Publisher Full Text\n\nBailey MJ, Biely P, Poutanen K: Interlaboratory testing of methods for assay of xylanase activity. J Biotechnol. 1992; 23(3): 257–270. Publisher Full Text\n\nSinghania RR, Sukumaran RK, Patel AK, et al.: Advancement and comparative profiles in the production technologies using solid-state and submerged fermentation for microbial cellulases. Enzyme Microb Technol. 2010; 46(7): 541–549. Publisher Full Text\n\nJa’afaru MI: Screening of fungi isolated from environmental samples for xylanase and cellulase production. ISRN Microbiol. 2013; 2013: 283423. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSingh Y, Gundampati RK, Jagannadham MV, et al.: Extracellular L-asparaginase from a protease-deficient Bacillus aryabhattai ITBHU02: purification, biochemical characterization, and evaluation of antineoplastic activity in vitro. Appl Biochem Biotechnol. 2013; 171(7): 1759–1774. PubMed Abstract \| Publisher Full Text\n\nMa L, Yang W, Meng F, et al.: Characterization of an acidic cellulase produced by Bacillus subtilis BY-4 isolated from gastrointestinal tract of Tibetan pig. J Taiwan Inst Chem Eng. 2015; 56: 67–72. Publisher Full Text\n\nVeras B: Screening of protease, cellulase, amylase and xylanase from the salt-tolerant and thermostable marine Bacillus subtilis strain SR60. Harvard Dataverse, V1. 2018. http://www.doi.org/10.7910/DVN/J5JCC0" }	[ { "id": "43316", "date": "16 May 2019", "name": "Jorge Olmos-Soto", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe isolated B. subtilis strain only has the capacity to degrade cellulose at high saline concentrations (1.25-1.5 M) and its xylanase activity is only developed at medium salt concentration (1 M). However, protease and amylase activity cannot be considered as halotolerant because these enzymes only have good activity at 0.25-0.5 M. In this sense, I believe the conclusion must be reoriented to point out these results.\n\nAll the figures contain the same legend, please correct this issue.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [] }, { "id": "72914", "date": "21 Apr 2021", "name": "Shohreh Ariaeenejad", "expertise": [ "Reviewer Expertise Enzyme biochemical characterization" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe main idea of this work is to observe the production of multiple thermoenzymes and haloenzymes, including protease, cellulase, amylase and xylanase, from bacterial strains isolated from coral reefs Cabo Branco, Paraiba State, Brazil. Strain SR60 was identified by the phylogenetic analysis to be Bacillus subtilis through a 16S ribosomal RNA assay. To screen multiples enzymes, B. subtilis SR60 was inoculated in differential media to elicit the production of extracellular enzymes with the addition of a range of salt concentrations (0, 0.25, 0.50, 1.0, 1.25 and 1.5 M NaCl).\nComments to the author(s):\n\nComment #1:\nThe introduction is very poorly written. No history of similar work has been mentioned in this regard.\nComment #2: At the end of the introduction: “This study aimed to produce multiple thermoenzymes and haloenzymes (protease, cellulase, amylase and xylanase) expressed by Bacillus subtilis strain SR60, a bacterial symbiont isolated from Siderastrea stellate (Verrill, 1868) in a Brazilian coral reefs ecosystem 7°08’50” S; 34° 47’51” W.” - What does this mean? You expressed some enzyme genes to Bacillus subtilis strain SR60? Or You found a novel bacterial symbiont isolated from Siderastrea stellate and this bacteria had some enzyme genes?\nThe Authors wrote in materials \"A total of 12 bacterial isolates were obtained, which were analysed for protease, cellulase, amylase and xylanase production capacity, and only the one with the simultaneous production capacity of these enzymes was selected.” - This is very different from the last sentence of the introduction.\n\nComment #3: At the end of the isolation of the thermophilic bacterial strain: “For further screening of enzymatic activity described below, two bacterial colonies, isolated using the above culturing conditions, were inoculated onto each plate.” - Why did you choose the two bacterial colonies?\nComment #4: In the part of Bacterial identification: Why are there no comparisons of the phylogenetic tree of these 12 identified strains? Why is their access number not in the article?\nComment #5: The results are very vague. Which strains showed what enzymatic activity? The whole article is written as a short story.\nComment #6: In the Table 2. Screening of enzyme production culture in different molarities NaCl. The results should be expressed as a comparative percentage (Relative activity). Please see my related article about effect of salt on enzymes1.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/7-1704
https://f1000research.com/articles/7-1703/v1	26 Oct 18	{ "type": "Research Article", "title": "Apoptosis inducing effects of chlorhexidine and essential oil mouthwashes on BHK-21 fibroblast cell line: An in vitro study", "authors": [ "Shaimaa Ali Hamouda Ali El Basuony", "Naglaa El Hossary", "Nermine Raouf Amin", "Naglaa El Hossary", "Nermine Raouf Amin" ], "abstract": "Background: The maintenance of oral health can be achieved mainly by mechanical and chemical means. Among chemical agents, mouthwashes are widely used for personal oral hygiene because of their ability to inhibit dental plaque. The antibacterial effects of essential oils (EOs) and chlorhexidine (CHX) are well documented; however, the reaction of host tissue to these substances has a poor documentation. Until now studies have not examined the effect of EOs with sodium fluoride (EOF) on fibroblast cell lines. The aim of this study was to examine the effect of mouth rinse EOs, EOF and CHX on the apoptosis of fibroblast cell line. Methods: BHK-21 fibroblast cell line was cultured and incubated in Eagle's Minimum Essential Medium containing EOs, EOF and CHX mouthwashes with different doses (15% or 25%) and various exposure times. Cell apoptosis was assayed using RT-PCR. Results: EOs, EOF and CHX induce apoptotic effects on fibroblasts in a dose and time dependent manner. Conclusion: CHX is the most cytotoxic mouthwash to fibroblasts as compared to mouthwashes containing EOs and EOF.", "keywords": [ "fibroblast", "chlorhexidine", "essential oils", "essential oils with sodium fluoride", "mouthwash." ], "content": "Introduction\n\nFibroblasts are the common cell type in connective tissue and plays a major role in normal function and in pathologic changes of oral tissues1,2. Various antiseptic agents are available in the form of mouthwashes, among them are essential oils (EOs), EOs with sodium fluoride (EOF) and chlorhexidine (CHX). The antibacterial action of these agents had been widely examined1,3,4.\n\nIn fibroblasts, EOs can induce depolarization of the mitochondrial membranes by decreasing the membrane potential, affecting ionic Ca++ cycling5,6. They change the permeability of membranes resulting in liberation of cytochrome, leading to apoptosis7. Fluoride inhibits protein synthesis and thus it influences certain signalling pathways included in apoptosis such as the P53 pathway8,9.\n\nCHX inhibits protein and DNA synthesis, resulting in cell apoptosis3,10. Moreover, CHX can damage mitochondrial membranes resulting in leakage of apoptosis-inducing proteins11,12.\n\nMany previous studies have compared the effect of both CHX and EOs mouthwashes on cultured fibroblasts3,5. However, to the best of our knowledge, the combined effect of EOs and EOF hasn’t been studied yet. In this study, we investigated the effect of different doses and different exposure periods of mouthwashes containing EOs, EOF and CHX on a fibroblast cell line.\n\n\nMethods\n\nThis study was performed on BHK-21 cell line in (Vacsera Co., Egypt). It was cultured according to a previous protocol3. Briefly, BHK-21 cells were cultured in Eagle's Minimum Essential Medium (ATCC, Manassas, VA, USA). They were sub-cultured into 96-well plates (ATCC, USA).\n\nCommercial mouthwashes (3 total) were diluted to 15% and 25% using distilled water. Mouthwashes used were composed as followed:\n\n- EO mouthwash: containing thymol (0.0060%), eucalyptol (0.09%), mentol (0.042%) and methyl-salicylate (0.064%) in a 26.9% hydroalcoholic vehicle;\n\n- EOF mouthwash: containing EOs as before and 0.2% sodium fluoride;\n\n- CHX mouthwash: containing 0.12% chlorhexidine hydrochloric acid.\n\nCells were incubated with mouthwashes for 0, 1, 2, 3, 5, 10 minutes in the 96-well plates at 37oC.\n\nAfter cell culturing, RNA extraction and quantitative real time polymerase chain reaction (PCR) was performed as follows1:\n\n1. Total RNA was isolated using the RNeasy Micro Kit (Qiagen, Germany) and RNA concentration was measured spectrophotometrically using Nanodrop ND-1000.\n\n2. Reverse transcription was performed on 600 ng of total RNA using oligo dT primers and MMLV Reverse Transcriptase in a final volume of 20 mL (Invitrogen, CA, USA) for 5 minutes at 65oC, followed by one hour at 37oC.\n\n3. Samples were subsequently heated for 15 minutes at 70oC to terminate the reverse transcription reaction.\n\n4. Real-time quantitative PCR was performed on the cDNA samples using an Applied Biosystem Real-Time Detection System. Annexin V was the target gene for apoptosis. Primer sequences for Annexin V: sense primer, 5' CAGTCTAGGTGCAGCTGCCG 3' and antisense primer, 5' GGTGAAGCAGGACCAGTGT3'. The following primer sequences were used for GAPDH (housekeeping gene): sense primer, 5' ATG GCC TTC CGT GTT CCT AC 3' and antisense primer, 5' GCC TGC TTC ACC ACC TTC TT 3'\n\n5. Real-time PCR was conducted by amplifying the cDNA with iQ SYBR Green Universal Master Mix (Applied Biosystems, CA, USA).\n\n6. Melting curve analysis of amplification products was performed at the end of the PCR reaction to confirm that a single PCR product was detected. For every PCR reaction, GAPDH was used as the internal control.\n\n7. A relative quantification method 2−ΔΔCT method was used13. The relative quantitation value of target was normalized to the internal control (GAPDH).\n\nData was analysed using IBM Statistical Package for Social Sciences (SPSS), version 21 (SPSS Inc., IL, USA). Numerical data was described as mean and standard deviation and comparisons between these were performed using ANOVA and a post-hoc Tukey test.\n\n\nResults\n\nMicroscopic examination of fibroblast cell line after application of different mouthwashes at different concentrations and for different time durations is shown in Figure 1.\n\nFrom L-R (A) Untreated fibroblast cell line demonstrated confluent growth of elongated cells. Some cells appeared bipolar and some were multipolar. (B–E) Essential oil mouthwash: (B) 15% for 1 minute showed many viable spindle shaped fibroblasts; (C) 15% for 10 minutes showed decreased spindle shaped fibroblasts; (D) 25% for 1 minute showed few viable spindle shaped fibroblasts; (E) 25% for 10 minutes showed obvious cell free areas. (F–I) Sodium fluoride mouthwash: (F) 15% for 1 minute showed many viable spindle shaped fibroblasts and few apoptotic cells; (G) 15% for 10 minutes showed destroyed fibroblasts; (H) 25% for 1 minute showed large cell free areas; (I) 25% for 10 minutes showed more obvious large cell free areas. (J-M) Chlorhexidine mouthwash: (J) 15% for 1 and (K) 10 minutes many viable fibroblasts were detected; (L) 25% for 1 minute showed massive reduction in cell viability, many dead or destroyed fibroblasts surrounded by large cell free areas; (M) 25% for 10 minutes, only few remnants of dead fibroblasts were obvious.\n\nIn all mouthwashes 25% concentration showed a statistically significant increase in apoptosis compared with 15% and the untreated control (Table 1 and Figure 2).\n\nGroups with different letters are statistically significantly different. ANOVA. EO, essential oil; EOF, sodium fluoride; CHX, chlorhexidine.\n\nEO, essential oil; EOF, sodium fluoride; CHX, chlorhexidine (n=3).\n\nIn the EO mouthwash, values for apoptosis continued to significantly increase after 2, 3, 5 and 10 minutes for the 25% concentration (Table 2 and Figure 3).\n\nGroups with different letters are statistically significantly different. *ANOVA\n\nIn the EOF mouthwash, at 25% concentration, a significant increase in apoptosis values was detected after 5 and 10 minutes (Table 3 and Figure 4).\n\nGroups with different letters are statistically significantly different. *ANOVA\n\nIn the CHX mouthwash, at 25% concentration, a significant increase in apoptosis values started after 1 minute and continued to increase by time (Table 4 and Figure 5).\n\nGroups with different letters are statistically significantly different. **ANOVA\n\n\nDiscussion\n\nThe effectiveness of CHX and EOs mouthwashes in controlling the formation of plaque and gingivitis has been demonstrated14. However, there are concerns that these products are harmful to oral cells15–17.\n\nCHX is toxic, even in low concentrations, for different cell types including fibroblasts in culture18,19. Topical application of CHX can result in its penetration through the epithelial barrier leading to tissue damage20. In our study, and increase in concentration for EOs, EOF and CHX mouthwashes resulted in an increase in apoptosis. This was also observed by Faria et al.1 who reported that CHX induced apoptosis of cultured fibroblasts in a concentration dependent manner. Values for apoptosis at 15% and 25% concentrations of EOF were slightly higher than EOs alone in our study, demonstrating the ability of fluoride to enhance apoptosis induction, as previously described9.\n\nA significant increase in apoptosis induction was seen at 15% concentration CHX and at 25% concentration EOs and EOF. This suggests that CHX is more effective than EOs and EOF in apoptosis induction. This result was in agreement with Tsourounakis et al.5 who reported that there was a significant reduction in cell survival that occurred at concentrations of 15% CHX and 25% EOs21.\n\nThe increase in duration of application of EOs, EOF and CHX didn’t significantly increase apoptosis at the low concentration of 15% in the present study. However, at 25% CHX, increase in the duration of treatment significantly enhanced apoptosis, which was significant obvious after 1 minute. This means that CHX is more efficient than EOs and EOF in apoptosis induction at lower concentrations. This was in accordance with Flemingson et al.3 who stated that the apoptotic effect of CHX on fibroblasts occurs early, after 1 minute exposure, and added that CHX had the maximum cytotoxicity followed by EOs.\n\n\nConclusion\n\nCHX mouthwash is the most cytotoxic to fibroblasts compared to EOs and EOF containing mouthwashes. Adding fluoride to EOs in low concentrations didn’t worsen the adverse effects of Eos as shown by the combination mouthwash containing EOs and fluoride.\n\n\nData availability\n\nF1000Research: Dataset 1. File containing: Part A, Raw numerical data of RNA concentration at each concentration and duration of application of the mouthwashes; part B, raw phase contrast photomicrograph of fibroblast cell line after application of mouthwashes at different concentrations and durations., 10.5256/f1000research.16337.d22068422.", "appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nFaria G, Celes MR, De Rossi A, et al.: Evaluation of chlorhexidine toxicity injected in the paw of mice and added to cultured l929 fibroblasts. J Endod. 2007; 33(6): 715–722. PubMed Abstract \| Publisher Full Text\n\nBarnhart BD, Chuang A, Lucca JJ, et al.: An in vitro evaluation of the cytotoxicity of various endodontic irrigants on human gingival fibroblasts. J Endod. 2005; 31(8): 613–615. PubMed Abstract \| Publisher Full Text\n\nFlemingson, Emmadi P, Ambalavanan N, et al.: Effect of three commercial mouth rinses on cultured human gingival fibroblast: an in vitro study. Indian J Dent Res. 2008; 19(1): 29–35. PubMed Abstract \| Publisher Full Text\n\nRajabalian S, Mohammadi M, Mozaffari B: Cytotoxicity evaluation of Persica mouthwash on cultured human and mouse cell lines in the presence and absence of fetal calf serum. Indian J Dent Res. 2009; 20(2): 169–73. PubMed Abstract \| Publisher Full Text\n\nTsourounakis I, Palaiologou-Gallis AA, Stoute D, et al.: Effect of essential oil and chlorhexidine mouthwashes on gingival fibroblast survival and migration. J Periodontol. 2013; 84(8): 1211–1220. PubMed Abstract \| Publisher Full Text\n\nCarson CF, Mee BJ, Riley TV: Mechanism of action of Melaleuca alternifolia (tea tree) oil on Staphylococcus aureus determined by time-kill, lysis, leakage, and salt tolerance assays and electron microscopy. Antimicrob Agents Chemother. 2002; 46(6): 1914–1920. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nArmstrong JS: Mitochondrial membrane permeabilization: the sine qua non for cell death. BioEssays. 2006; 28(3): 253–260. PubMed Abstract \| Publisher Full Text\n\nJeng JH, Hsieh CC, Lan WH, et al.: Cytotoxicity of sodium fluoride on human oral mucosal fibroblasts and its mechanisms. Cell Biol Toxicol. 1998; 14(6): 383–389. PubMed Abstract \| Publisher Full Text\n\nKarube H, Nishitai G, Inageda K, et al.: NaF activates MAPKs and induces apoptosis in odontoblast-like cells. J Dent Res. 2009; 88(5): 461–465. PubMed Abstract \| Publisher Full Text\n\nChang YC, Huang FM, Tai KW, et al.: The effect of sodium hypochlorite and chlorhexidine on cultured human periodontal ligament cells. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2001; 92(4): 446–450. PubMed Abstract \| Publisher Full Text\n\nNegrelo Newton AP, Cadena SM, Merlin Rocha ME, et al.: New data on biological effects of chlorhexidine: Fe2+ induced lipid peroxidation and mitochondrial permeability transition. Toxicol Lett. 2004; 151(3): 407–416. PubMed Abstract \| Publisher Full Text\n\nCribb AE, Peyrou M, Muruganandan S, et al.: The endoplasmic reticulum in xenobiotic toxicity. Drug Metab Rev. 2005; 37(3): 405–42. PubMed Abstract \| Publisher Full Text\n\nBoulter N, Suarez FG, Schibeci S, et al.: A simple, accurate and universal method for quantification of PCR. BMC Biotechnol. 2016; 16: 27. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHaffajee AD, Roberts C, Murray L, et al.: Effect of herbal, essential oil, and chlorhexidine mouthrinses on the composition of the subgingival microbiota and clinical periodontal parameters. J Clin Dent. 2009; 20(7): 211–217. PubMed Abstract\n\nWyganowska-Swiatkowska M, Urbaniak P, Szkaradkiewicz A, et al.: Effects of chlorhexidine, essential oils and herbal medicines (Salvia, Chamomile, Calendula) on human fibroblast in vitro. Cent Eur J Immunol. 2016; 41(2): 125–131. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSelwitz RH, Ismail AI, Pitts NB: Dental caries. Lancet. 2007; 369(9555): 51–59. PubMed Abstract \| Publisher Full Text\n\nGiannelli M, Chellini F, Margheri M, et al.: Effect of chlorhexidine digluconate on different cell types: a molecular and ultrastructural investigation. Toxicol In Vitro. 2008; 22(2): 308–317. PubMed Abstract \| Publisher Full Text\n\nGhabanchi J, Moattari A, Darafshi R, et al.: Effects of three Commercial Mouth Rinses on the Cultured Fibroblasts: An in Vitro Study. J Dent (Shiraz). 2013; 14(2): 64–7. PubMed Abstract \| Free Full Text\n\nEren K, Ozmeriç N, Sardaş S: Monitoring of buccal epithelial cells by alkaline comet assay (single cell gel electrophoresis technique) in cytogenetic evaluation of chlorhexidine. Clin Oral Investig. 2002; 6(3): 150–154. PubMed Abstract \| Publisher Full Text\n\nMüller HD, Eick S, Moritz A, et al.: Cytotoxicity and Antimicrobial Activity of Oral Rinses In Vitro. Biomed Res Int. 2017; 2017: 4019723. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAfonina E, Stauber R, Pavlakis GN: The human poly(A)-binding protein 1 shuttles between the nucleus and the cytoplasm. J Biol Chem. 1998; 273(21): 13015–13021. PubMed Abstract \| Publisher Full Text\n\nEl Basuony SAH, El Hossary N, Raouf Amin N: Dataset 1 in: Apoptosis inducing effects of chlorhexidine and essential oil mouthwashes on BHK-21 fibroblast cell line: An in vitro study. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16337.d220684" }	[ { "id": "39910", "date": "09 Nov 2018", "name": "Marwa Mokbel ElShafei", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn my opinion normality tests should be performed before statistical tests and considering the small sample size, non-parametric tests could have been used. The article is approved, though a recommendation can be added after the conclusion in order to recommend further studies to evaluate the amount and percentage of mouth wash absorption to the connective tissue and hence reaching fibroblasts as the direct effect on the cell line doesn't mimic the real life effect.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "40531", "date": "18 Jan 2019", "name": "Thomas E. Lallier", "expertise": [ "Reviewer Expertise Cell biology", "including cell survival", "cell attachment", "cell migration assays", "as well as qPCR", "ELISA", "and immunocytochemistry" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors demonstrate that chlorohexidine (CHX)- and essential oil (EO)-containing mouth rinses induce cell death in fibroblasts in culture. They attempt to prove that this is through apoptosis, using qPCR to quantify the expression of Annexin V (Annexin-A5) in these cells. Unfortunately, this assay is flawed. Annexin V is a normally-expressed protein that happens to bind to phosphatidylserine, a phospholipid normally on the cytoplasmic leaflet of the plasma membrane. During apoptosis, this inner leaflet is exposed to the extracellular environment. Thus, Annexin V can be used to identify apoptotic cells fluorometrically. An increase in Annexin V transcript expression has not been linked with apoptosis. Furthermore, apoptosis is a post-translational event involving cytochromes, caspases and other proteins (e.g. BCL2 and BAX), and thus cannot be detected using a transcriptional assay like qPCR.\nThus, the only finding of this manuscript is that these mouth rinses lead to cell death, which is neither novel nor worthy of independent indexing.\nThe manuscript also contains numerous other errors (e.g. the photomicrographs in Figure 1 are of too little contrast to demonstrate anything, and are not even properly identified), and this is not yet ready for indexing.\n\nIs the work clearly and accurately presented and does it cite the current literature? No\n\nIs the study design appropriate and is the work technically sound? No\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNo\n\nAre all the source data underlying the results available to ensure full reproducibility? No\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [] } ]	1	https://f1000research.com/articles/7-1703
https://f1000research.com/articles/7-1285/v1	14 Aug 18	{ "type": "Research Article", "title": "Aloe barbadensis Miller leaf exudate is a potential treatment for bovine mastitis", "authors": [ "Samira de Aquino Leite Fiordalisi", "Luciana Aparecida Honorato", "Shirley Kuhnen", "Samira de Aquino Leite Fiordalisi", "Luciana Aparecida Honorato" ], "abstract": "Background: Aloe barbadensis Miller is a well-known phytotherapeutic, and parts of its leaves are used for a wide range of medicinal purposes. This study seeks to assess the in vitro antimicrobial and cytotoxic effects of leaf exudate (LE) from A. barbadensis leaves against Staphylococcus aureus and MAC-T bovine mammary epithelial cells. Methods: Seasonal LE samples were collected, and the effect on total phenolic and aloin contents was determined. Antimicrobial activity of LE was evaluated using the broth microdilution method, and toxicity to MAC-T cells was determined by MTT assay. Results: Samples collected during different seasons of the year showed a seasonal effect on the chemical profile of LE (P<0.05). However, despite these chemical variations, we found no differences in antimicrobial activity against S. aureus. For all studied samples, the minimum inhibitory concentration (MIC) was 1,000 µg/ml. Furthermore, we found an elevated cytotoxic effect of LE on MAC-T cells with a significant reduction in cellular viability at 7.8 µg/ml (P<0.05) and an IC50 of 91.89 µg/ml. Conclusions: Despite the antimicrobial effects of LE, the high toxicity for MAC-T cells suggests that it is unsuitable for intramammary use, but does have potential as a topical antimicrobial.", "keywords": [ "phytotherapy", "Staphylococcus aureus", "MAC-T cells" ], "content": "Introduction\n\nBovine mastitis, which is characterized by inflammation of the mammary gland, is the most frequent infection found in dairy herds worldwide1. The treatment recommended for mastitis is the administration of intramammary antimicrobials. However, control of infections caused by Staphylococcus aureus, the principal etiological agent of bovine mastitis2, is very difficult. In addition to inactivating several antimicrobials, this microorganism can also survive in the intracellular environment after phagocytosis. As a consequence, the cure rate of mastitis caused by S. aureus is low, with a high incidence of recurrence3. As such, interest in the search for methods of control and prevention has increased, including the identification of new antimicrobials4–7.\n\nIn vivo methods are still commonly used to study bovine mastitis, but in vitro testing has been recommended8. Based on in vitro models, studies have produced a wide range of results, from identifying the prevalence of etiological agents of mastitis to evaluating the direct effects of products on the susceptibility of studied microorganisms9,10. Among these, in vitro tests on antimicrobials are some of the most widely used8. In vitro studies with bovine mammary gland explants or mammary epithelial cells (MEC) are commonly used to assess the different functions of mammary glands, such as the response to initial infection11,12. Recently, primary cultures of mammary explants and MECs were also suggested as adequate models in the search for new therapeutic agents5,13. In the case of mastitis, such in vitro methods can help evaluate the toxicity of antimicrobials, enabling the determination of safe doses and minimizing the potential risks during in vivo validation.\n\nAloe vera (Aloe barbadensis Miller) is a plant widely used and recognized for its antimicrobial, anti-inflammatory, wound-healing, antitumor, and antioxidant pharmacological properties14. Yet, until now, few studies have reported on its potential as a treatment for bovine mastitis. Most research on the pharmaceutical potential of Aloe vera has studied the mucilaginous gel, commonly known as aloe vera gel, that is rich in complex carbohydrates, particularly acemannans15,16. However, along with the gel, a yellow exudate with a strong odor and bitter taste, known as leaf exudate (LE), can also be extracted from the leaves17. Its release occurs as soon as the leaves are cut and it can be found within the phloem vessels18,19. Despite being composed of large amounts of 1,8-dihydroxyanthraquinone derivatives and their glycosides, the industry that uses aloe vera gel as a raw material considers LE a residue. Among the anthraquinones found in LE, the major compounds are a mixture of two readily oxidizable diastereomers, aloin A and aloin B, which are sometimes undesirable because of their toxic and cathartic potential. However, these compounds may be of therapeutic interest in the control of antimicrobial and tumor cell proliferation20,21.\n\nThus, the current study seeks to investigate the potential of LE from Aloe vera leaves in the control of bovine mastitis through in vitro models that evaluate the antimicrobial effects against S. aureus and cytotoxicity to MAC-T cells.\n\n\nMethods\n\nA total of 30 plant samples were collected from 3-year-old Aloe vera (Aloe barbadensis) at random from a commercial grower (Naturama Sucos Integrais do Brasil Ltda®; Paulo Lopes, SC, Brazil) in March, June, September and December of 2015, and one leaf was taken from the mid-position of each plant. Thus, 30 leaves in total were collected for each month, representing each season. Leaves were cut at the base and maintained vertically for 3 h in a beaker to collect the LE at room temperature. Subsequently, the LE was lyophilized and stored at -20ºC. LEs of six plants were combined for a total of five repetitions for each season of the year.\n\nTotal phenolics. The total phenolic content was determined using the colorimetric method of Folin-Ciocalteau14 and an external standard curve of gallic acid (10–100 μg/ml) (y= 0.0197x / r2= 0.987). The results were expressed in µg of gallic acid equivalents (GAE)/mg of extract (µg of GAE/mg). All tests were performed in triplicate.\n\nAloin. The aloin content in LE was obtained on an UHPLC Thermo Scientific UltiMate 3000 RS Dual System (Thermo Fisher Scientific, San Jose, CA), using a Thermo Scientific C18 reverse-phase column (4.6 x 250 mm; 5 µm; 120Å (AcclaimTM120, Thermo Scientific©) at 40°C, operating at 240, 260, 280 and 320 nm. The mobile phase was eluted at 1 ml/min flow rate, using a methanol/water (70/30, v/v) mixture22. The identification of aloin was based on a comparison of the chromatographic profile and retention time with the commercial standard (Sigma-Aldrich, St. Louis, MO, USA/ B6906). After the addition of the standard, samples were co-chromatographed to confirm identification of the compound. Aloin content was determined through an external standard curve of barbaloin (y= 302.73x / r2= 0.9822) and the result expressed in µg of aloin per mg of the sample (µg/mg).\n\nAntimicrobial activity was evaluated using a broth microdilution method according to the Clinical and Laboratory Standards Institute Manual23. We tested six different concentrations of LE (4000 to 125 µg/ml) against the standard strain of S. aureus ATCC 25923 (Collection of Reference Microorganisms on Health Surveillance, Fundação Oswaldo Cruz, Fiocruz, Brazil) and seven strains of mastitic milk isolates. Milk samples from cows were submitted to the California Mastitis Test (CMT). CMT-positive milk samples were plated on blood agar supplemented with 5% sterile ovine blood and incubated for 24–48h at 37ºC. Gram-positive, catalase-positive, and rabbit plasma coagulase-positive samples were biochemically confirmed as Staphylococcus aureus24. Each strain was considered one repetition of the experiment with five replicates/repetition. As such, we conducted eight repetitions for each LE sample.\n\nThe minimum inhibitory concentration (MIC) was determined through visual analysis of turbidity after 24 hours of incubation on plates at 37°C in addition to spectrophotometric reading at 600 nm to determine the percentage of inhibition of bacterial growth, using a previously described method23.\n\nBecause LE is an exudate with a yellow color that easily oxidizes to a dark coloration14, we confirmed the MIC through a colorimetric method. We added 50 µl of resazurin dye (100 µg/ml, Sigma-Aldrich, St. Louis, MO, USA) to each well after reading the plates by spectrophotometer (600 nm). The plates were left to incubate at 37°C for an additional 30 minutes25.\n\nMammary epithelial cells of the MAC-T (Mammary Alveolar Cells-T) lineage were maintained in culture, as indicated by the supplier (Banco de Células do Rio de Janeiro, Brazil). Briefly, MAC-T cells were cultivated in Dulbecco’s Modified Eagle’s Medium (DMEM) and supplemented with 100 U/ml of penicillin, 100 µg/ml of streptomycin, 20% (V/V) heat-inactivated fetal bovine serum (FBS, Gibco, CA, USA), 4 mM L-glutamine (Synth), 4.5 g/L glucose (Sigma-Aldrich, St. Louis, MO, USA), 1 mM of sodium pyruvate (Sigma-Aldrich, St. Louis, MO, USA), 1.5 g/L sodium bicarbonate (VetecTM Sigma-Aldrich, St. Louis, MO, USA), 5 μg/ml insulin (Sigma-Aldrich, St. Louis, MO, USA), and 1 μg/ml hydrocortisone (Sigma –Aldrich, St. Louis, MO, USA) at 37°C and 5% CO2 in a humidified incubator. We changed the medium every 48 h. After reaching confluence, the cells were treated with 0.25% trypsin with 1 mM EDTA (Gibco, CA, USA) to prepare the cellular suspension (105 cells/ml). The suspension was transferred a 96-well microplate (100 µl/well), followed by incubation (24 h) in culture conditions for adherence. Subsequently, varying concentrations (2000, 1000, 500, 250, 125, 62.5, 31.3, 15.6, 7.8 and 3.9 µg/ml) of LE from the summer samples were added for 24 h, and cytotoxicity was determined based on the MTT method (Sigma-Aldrich, St. Louis, MO, USA)26. The formed formazan was dissolved with dimethyl sulfoxide (DMSO) to give a purple color with characteristic absorption at 540 nm. Intensity of purple color is directly proportional to the cell number, thus indicating cell viability. The experiments were performed in triplicate.\n\nData were expressed as the mean ± standard deviation (SD), of at least three independent experiments. We analyzed the data using analysis of variance with a Tukey adjustment (GraphPad Prism 5.0). We considered the effects statistically significant for P<0.05. The inhibitory concentrations capable of reducing cellular viability by 50% (IC50) were calculated using a nonlinear regression of data obtained from the cellular viability tests with GraphPad Prism 5.0 software. The average accumulated precipitation (mm) in the study region was calculated using available data27.\n\n\nResults and discussion\n\nThe total phenolic and aloin content of LE from the aloe vera leaves varied based on the season during which the samples were collected. The highest levels were found in the samples taken during the summer and the lowest levels from the spring samples (P<0.05). The accumulation of total phenolics in the summer LE seems to be associated with the climatic conditions during the collection period (Figure 1a, b). Precipitation indices were lower during the summer months (January to March)27 (Figure 1a), possibly causing hydric stress in the plants. In the spring, the lower total phenolic content of LE coincides with a period of greatest precipitation that year. Aloe is a plant comprised of 96% water; thus, its chemical composition is heavily influenced by precipitation levels28, as well as other factors, such as the period of flowering29.\n\n(A) Average accumulated precipitation (mm) in the study region (Paulo Lopes, SC, Brazil) during 2015; Source: INMET26. (B) Total phenolic content (μg GAE/mg) of Aloe vera leaf LE from different seasons of the year (average of five repetitions ±SD) (P<0.05). (C) Average content (µg/mg) of aloin (average of three independent injections ± SD) in samples of LE collected from Aloe vera leaves during different seasons of the year. Data points with the same letter above them are not significantly different from each other (P<0.05 indicates a significant difference).\n\nThe husk of Aloe leaves has greater levels of total phenolics compared to the leaf interior and internal parenchyma. In the literature, these values range from 12.06 to 20.86 µg GAE/mg leaf, depending on the species14. Among the various phenolic compounds in the leaves of Aloe, anthraquinones are noteworthy, particularly aloin. Anthraquinones are free in phloem vessels directly below the leaf epidermis, and aloin, in particular, is distributed throughout the plant as part of its defense mechanism30. In the present study, we found the highest levels of aloin in the summer samples and the lowest in the spring samples (Figure 1 b, c). These results are correlated with the total phenolic content found in the studied samples (Figure 1). Previous studies have also suggested the effect of seasonality on aloin content, and its synthesis is strongly influenced by precipitation levels. Dry periods have been correlated with greater content of aloin in the analysis of aloe vera leaves15,31,32. However, other factors can influence aloin content of aloe vera leaves, including cultivation conditions, age, and plant health33. For example, higher levels of barbaloin, isobarbaloin, and aloin in Aloe sp. plants were found during periods of the year with higher temperatures32.\n\nDespite significant differences in the levels of total phenolics and aloin in the LE samples (Figure 1b, c), these levels did not influence antimicrobial activity against S. aureus. For all LE samples, the MIC was 1,000 µg/ml, as confirmed by resazurin oxidation. The lowest tested concentration of 500 µg/ml was incapable of reducing bacterial growth to values greater than 80%. The effect of other concentrations between 500 and 1,000 µg/ml was not included in the study (Figure 2).\n\nThe effectiveness of LE from aloe vera leaves as an antimicrobial agent has been demonstrated for a wide variety of Gram-positive and Gram-negative bacteria, including S. aureus and others34,35. In the literature, the MIC of aloe vera extracts against S. aureus varies. Previous studies have shown lower (195 µg/ml), similar (1560 µg/ml), and higher (5,000 µg/ml) MIC values compared to those in the present study36–38. This variation may be related to diverse factors, such as the Aloe species studied, the part of the aloe leaf used in the tests, and the type of extraction and resuspension vehicle used. In the current study, the LE samples were collected directly from the cut leaf without any type of posterior extraction of the compounds of interest. Some solvents are capable of extracting certain compounds that may possess greater antimicrobial activity than others39; however, resuspension in water may be the easiest way to use aloe vera leaf subproducts, making it accessible, even to the producer.\n\nAn interesting aspect to consider in the present study is that the concentration of total phenolics and aloin in the LE samples does not seem to affect antimicrobial activity. By contrast36, the antimicrobial and anti-inflammatory activity of aloe vera LE has been associated with the concentration of phenolic and aloin compounds, suggesting that older leaves have higher levels of these compounds, and as such, have greater biological activity and defense against microorganisms and herbivores.\n\nFor Fabry et al.40, the potentially useful activity defined for crude plant extracts with organic solvents is considered good when MIC values are <8000 µg/ml, while Gibbons41 suggests that phytochemical isolates must have MIC values <1,000 µg/ml. As such, the antimicrobial action of the LE samples in the present study can be considered good, even though neither extraction nor isolation of the principal components took place.\n\nThe LE showed high toxicity to MAC-T lineage cells, causing significant reduction in cellular viability at concentrations greater than 7.8 μg/ml (Figure 3). At higher concentrations, such as 500 μg/ml, the reduction in the percentage of viable cells was greater than 80%. The IC50 was 91.89 μg/ml. It is worth noting that the MIC of S. aureus growth was 1,000 μg/ml (Figure 2), a concentration that had a strong effect on the viability of mammary epithelial cells (Figure 3). This result is significant because it suggests that caution must be exercised when considering the intramammary use of LE in order to avoid inflammation, owing to the death of epithelial cells.\n\nData shown are an average of three independent experiments. Data points with the same letter above them are not significantly different from each other (P<0.05 indicates a significant difference).\n\nIn an in vivo situation, the administration of a toxic product to bovine mammary glands can result in the development of inflammation42, which is more severe than that caused by the infection of pathogens42. In these cases, the attempt to combat inflammation leads to the formation of connective tissue at the affected site, which can diminish the alveolar area responsible for the synthesis of milk and, consequently, reduce milk production. In more severe cases, the loss of complete mammary glandular function, or even death of the animal, can occur43,44.\n\nThe MAC-T cellular lineage45 is an established model that has been frequently used in the investigation of mammary glandular functions and mediators of inflammatory processes46. Nonetheless, studies reporting on the effects of Aloe sp. extract, or fractions on this type of cell, are scarce. The toxic effects of aloe vera LE on other types of cells are discussed in the literature and have been associated with the presence of aloin and aloe emodin47. These anthraquinones induce the apoptosis of cells caused by a reduction in the proportion of cells in the mitotic phase48. Another hypothesis is that disruptions to the cell cycle and cellular differentiation, stimulation of the immune system, and antioxidant activity also have an anti-proliferative effect49.\n\nWhile the results found for MAC-T cells show that aloe vera LE has a high toxic potential for bovine mammary glands, the topical use of this product on the external area of the udder, for example pre- and post-dipping, or on instruments used during the management of milking, can be recommended. In this case, its potential as a disinfectant should be investigated.\n\nFurthermore, it is important to highlight that the compounds present in the Aloe vera LE liquid oxidize easily in the presence of light, oxygen, and at room temperature50. As such, very high concentrations are required in order to achieve antimicrobial efficacy against S. aureus, concentrations that would be toxic to mammary epithelial cells. Thus, we suggest the standardization of a methodology that can preserve and conserve these oxidative compounds, such as nanoencapsulation, which can maintain the desired antimicrobial activity and diminish the toxic effects.\n\n\nConclusion\n\nAlthough seasonality interferes with the chemical composition of aloe vera LE, the seasonal samples we evaluated did not differ in relation to antimicrobial activity with a MIC of 1,000 µg/ml found for all samples. At this concentration, aloe vera LE shows strong toxic effects on bovine mammary epithelial cells of the MAC-T lineage. Despite the demonstrated antimicrobial activity of aloe vera LE, we suggest caution in recommending its intramammary use to treat bovine mastitis; instead, the topical use of this product on an external area, such as the udder, may be both efficacious and safe.\n\n\nData availability\n\nDataset 1. Raw data concerning the phytochemical characteristics of leaf exudate and its antimicrobial/cytoxicity activity. DOI: https://doi.org/10.5256/f1000research.15671.d21390151.", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by the CNPq (National Council on Scientific and Technological Development; Brasilia, Brazil) (403415/2013-6, Edital 39/2013).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nBradley A: Bovine mastitis: an evolving disease. Vet J. 2002; 164(2): 116–128. PubMed Abstract \| Publisher Full Text\n\nWatts JL: Etiological agents of bovine mastitis. Vet Microbiol. 1988; 16(1): 41–66. PubMed Abstract \| Publisher Full Text\n\nAnderson KL, Azizoglu RO: Detection and Causes of Bovine Mastitis with Emphasis on Staphylococcus aureus. Encyclopedia of Agriculture and Food Systems. Elsevier Ltd.; 2014; 2: 435–440. Publisher Full Text\n\nMushtaq S, Shah AM, Shah A, et al.: Bovine mastitis: An appraisal of its alternative herbal cure. Microb Pathog. 2018; 114: 357–361. PubMed Abstract \| Publisher Full Text\n\nFiordalisi SAL, Honorato LA, Loiko MR, et al.: The effects of Brazilian propolis on etiological agents of mastitis and the viability of bovine mammary gland explants. J Dairy Sci. 2016; 99(3): 2308–2318. PubMed Abstract \| Publisher Full Text\n\nBarlow J: Mastitis therapy and antimicrobial susceptibility: a multispecies review with a focus on antibiotic treatment of mastitis in dairy cattle. J Mammary Gland Biol Neoplasia. 2011; 16(4): 383–407. PubMed Abstract \| Publisher Full Text\n\nCollado R, Prenafeta A, González-González L, et al.: Probing vaccine antigens against bovine mastitis caused by Streptococcus uberis. Vaccine. 2016; 34(33): 3848–3854. PubMed Abstract \| Publisher Full Text\n\nBasso KM, Bracarense APFRL: Explantes teciduais: Um modelo redescoberto na experimentação animal. Semin Agrar. 2013; 34(6Supl2): 3951–3958. Publisher Full Text\n\nDe Medeiros ES, Veiga M, Wilton J, et al.: Avaliação in vitro da eficácia de desinfetantes comerciais utilizados no pré e pós-dipping frente amostras de Staphylococcus spp. isoladas de mastite bovina. Pesq Vet Bras. 2009; 29(1): 71–75. Publisher Full Text\n\nDiaz MAN, Rossi CC, MendonçaI VR, et al.: Screening of medicinal plants for antibacterial activities on Staphylococcus aureus strains isolated from bovine mastitis. Rev Bras Farmacogn. 2010; 20(5): 724–728. Publisher Full Text\n\nRabot A, Wellnitz O, Meyer HH, et al.: Use and relevance of a bovine mammary gland explant model to study infection responses in bovine mammary tissue. J Dairy Res. 2007; 74(1): 93–99. PubMed Abstract \| Publisher Full Text\n\nBoudjellab N, Chan-Tang HS, Zhao X: Bovine interleukin-1 expression by cultured mammary epithelial cells (MAC-T) and its involvement in the release of MAC-T derived interleukin-8. Comp Biochem Physiol A Mol Integr Physiol. 2000; 127(2): 191–199. PubMed Abstract \| Publisher Full Text\n\nWang K, Jin XL, Liu J, et al.: Potential of dietary propolis in protecting bovine mammary epithelial cells against mastitis pathogens using in vitro models. J Nutr Intermed Metab. 2016; 4: 39. Publisher Full Text\n\nLucini L, Pellizzoni M, Pellegrino R, et al.: Phytochemical constituents and in vitro radical scavenging activity of different Aloe species. Food Chem. 2015; 170: 501–507. PubMed Abstract \| Publisher Full Text\n\nRodríguez-González VM, Femenia A, González-Laredo RF, et al.: Effects of pasteurization on bioactive polysaccharide acemannan and cell wall polymers from Aloe barbadensis Miller. Carbohydr Polym. 2011; 86(4): 1675–1683. Publisher Full Text\n\nRay A, Gupta SD, Ghosh S: Evaluation of anti-oxidative activity and UV absorption potential of the extracts of Aloe vera L. gel from different growth periods of plants. Ind Crops Prod. 2013; 49: 712–719. Publisher Full Text\n\nSánchez-Machado DI, López-Cervantes J, Sendón R, et al.: Aloe vera: Ancient knowledge with new frontiers. Trends Food Sci Technol. 2017; 61: 94–102. Publisher Full Text\n\nFanali S, Aturki Z, D’Orazio G, et al.: Analysis of Aloe-based phytotherapeutic products by using nano-LC-MS. J Sep Sci. 2010; 33(17–18): 2663–2670. PubMed Abstract \| Publisher Full Text\n\nChinchilla N, Carrera C, Durán AG, et al.: Aloe barbadensis: How a miraculous plant becomes reality. Phytochem Rev. 2013; 12(4): 581–602. Publisher Full Text\n\nHamman JH: Composition and applications of Aloe vera leaf gel. Molecules. 2008; 13(8): 1599–1616. PubMed Abstract \| Publisher Full Text\n\nDuval J, Pecher V, Poujol M, et al.: Research advances for the extraction, analysis and uses of anthraquinones: A review. Ind Crops Prod. 2016; 94: 812–833. Publisher Full Text\n\nMandrioli R, Mercolini L, Ferranti A, et al.: Determination of aloe emodin in Aloe vera extracts and commercial formulations by HPLC with tandem UV absorption and fluorescence detection. Food Chem. 2011; 126(1): 387–393. Publisher Full Text\n\nCLSI: Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria That Grow Aerobically. Approved Standard-Tenth Edition. CLSI document M07-A10. 2015; 35(2): 1–87. Reference Source\n\nCLSI: Abbreviated Identification of Bacteria and Yeast; Approved Guideline. CLSI document M35A. 2002; 22(18): 1–19. Reference Source\n\nBauer J, Siala W, Tulkens PM, et al.: A combined pharmacodynamic quantitative and qualitative model reveals the potent activity of daptomycin and delafloxacin against Staphylococcus aureus biofilms. Antimicrob Agents Chemother. 2013; 57(6): 2726–2737. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMosmann T: Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. J Immunol Methods. 1983; 65(1–2): 55–63. PubMed Abstract \| Publisher Full Text\n\nINMET-Instituto Nacional de Meteriologia. Disponível em: Acesso em: janeiro de 2017. Reference Source\n\nPatel K, Patel DK: Medicinal importance, pharmacological activities, and analytical aspects of hispidulin: A concise report. J Tradit Complement Med. 2016; 7(3): 360–366. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKoshioka (Née Nishimura) M, Koshioka M, Takino Y, et al.: Studies on the evaluation of Aloe arborescens Mill. Var. Natalensis berger and aloe extract (JP IX). Pharm Biol. 1982; 20(2): 53–59. Publisher Full Text\n\nWintola OA, Afolayan AJ: The foliar anatomy and micromorphology of Aloe ferox Mill. (Asphodelaceae). Afr J Tradit Complement Altern Med. 2014; 11(2): 350–357. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGutterman Y, Chauser-Volfson E: Peripheral defence strategy: Variation of barbaloin content in the succulent leaf parts of Aloe arborescens Miller (Liliaceae). Bot J Linn Soc. 2000; 132(4): 385–395. Publisher Full Text\n\nBeppu H, Kawai K, Shimpo K, et al.: Studies on the components of Aloe arborescens from Japan - Monthly variation and differences due to part and position of the leaf. Biochem Syst Ecol. 2004; 32(9): 783–795. Publisher Full Text\n\nAzaroual L, Liazid A, Barbero GF, et al.: Improved Chromatographic Methods for Determination of Bioactive Compounds from Aloe vera Leaves. ISRN Chromatogr. 2012; 609095, 7. Publisher Full Text\n\nHabeeb F, Shakir E, Bradbury F, et al.: Screening methods used to determine the anti-microbial properties of Aloe vera inner gel. Methods. 2007; 42(4): 315–320. PubMed Abstract \| Publisher Full Text\n\nPellizzoni M, Ruzickova G, Kalhotka L, et al.: Antimicrobial activity of different Aloe barbadensis Mill. and Aloe arborescens Mill. leaf fractions. J Med Plants Res. 2012; 6(10): 1975–1982. Reference Source\n\nNdhlala AR, Amoo SO, Stafford GI, et al.: Antimicrobial, anti-inflammatory and mutagenic investigation of the South African tree aloe (Aloe barberae). J Ethnopharmacol. 2009; 124(3): 404–408. PubMed Abstract \| Publisher Full Text\n\nCardoso FL, Murakami C, Mayworm MAS, et al.: Análise sazonal do potencial antimicrobiano e teores de flavonoides e quinonas de extratos foliares de Aloe arborescens Mill., Xanthorrhoeaceae. Brazilian J Pharmacogn. 2010; 20(1): 35–40. Publisher Full Text\n\nSoyelu OT, Masika PJ: Traditional remedies used for the treatment of cattle wounds and myiasis in Amatola Basin, Eastern Cape Province, South Africa. Onderstepoort J Vet Res. 2009; 76(4): 393–397. PubMed Abstract \| Publisher Full Text\n\nAmoo SO, Aremu AO, Van Staden J: Unraveling the medicinal potential of South African Aloe species. J Ethnopharmacol. 2014; 153(1): 19–41. PubMed Abstract \| Publisher Full Text\n\nFabry W, Okemo PO, Ansorg R: Antibacterial activity of East African medicinal plants. J Ethnopharmacol. 1998; 60(1): 79–84. PubMed Abstract \| Publisher Full Text\n\nGibbons S: Anti-staphylococcal plant natural products. Nat Prod Rep. 2004; 21(2): 263–277. PubMed Abstract \| Publisher Full Text\n\nTroncarelli MZ, Langoni H, Brandão HM, et al.: Safety of a nanopropolis formulation intended for intramammary treatment of bovine mastitis in organic dairy herds. Rev Bras Hig e Sanidade Anim. 2014; 8(5): 517–45. Reference Source\n\nCunningham BK: Fisiología Veterinaria. Elsevier Saunders. 2009; 718. Reference Source\n\nDyce KM, Sack WO, Wensing CJG: Tratado de Anatomia Veterinária. Tratado de Anatomia Veterinária. 2004; 519–522.\n\nHuynh HT, Robitaille G, Turner JD: Establishment of bovine mammary epithelial cells (MAC-T): An in vitro model for bovine lactation. Exp Cell Res. 1991; 197(2): 191–9. PubMed Abstract \| Publisher Full Text\n\nLahouassa H, Moussay E, Rainard P, et al.: Lipopolysaccharides, cytokines, and nitric oxide affect secretion of prostaglandins and leukotrienes by bovine mammary gland during experimentally induced mastitis in vivo and in vitro. Vet J. 2014; 77(1): 90–99.\n\nCoran SA, Bartolucci G, Bambagiotti-Alberti M: Selective determination of aloin in different matrices by HPTLC densitometry in fluorescence mode. J Pharm Biomed Anal. 2011; 54(2): 422–425. PubMed Abstract \| Publisher Full Text\n\nEsmat AY, Tomasetto C, Rio MC: Cytotoxicity of a natural anthraquinone (Aloin) against human breast cancer cell lines with and without ErbB-2: topoisomerase IIalpha coamplification. Cancer Biol Ther. 2006; 5(1): 97–103. PubMed Abstract\n\nTomasin R, Gomes-Marcondes MC: Oral administration of Aloe vera and honey reduces Walker tumour growth by decreasing cell proliferation and increasing apoptosis in tumour tissue. Phyther Res. 2011; 25(4): 619–623. PubMed Abstract \| Publisher Full Text\n\nFreitas VS, Rodrigues RAF, Gaspi FOG: Propriedades farmacológicas da Aloe vera (L.) Burm. f. Rev Bras Plantas Med. 2014; 16(2): 299–307. Publisher Full Text\n\nFiordalisi SdAL, Honorato LA, Kuhnen S: Dataset 1 in: Aloe barbadensis Miller leaf exudate is a potential treatment for bovine mastitis. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15671.d213901" }	[ { "id": "37692", "date": "14 Sep 2018", "name": "Gaspar Diaz-Muñoz", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI read with interest the manuscript entitled: “Aloe barbadensis Miller leaf exudate is a potential treatment for bovine mastitis”. The manuscript is very interesting. In general, I found clearly written and only a few typos and errors were found as listed in the attached file that already were corrected.\nI did not understand why in the discussion of antimicrobial activity, the authors cited only the results of concentrations between 500 and 1000 ug/mL. In the methodology cited, the use of six concentrations between 500 and 4000 μg/mL.\nI think the authors should consider the possibility of Aloe exudate being toxic to epithelial cells as it showed toxicity to the MAC T cells. I suggest making a cytotoxicity test for this cell before suggesting this exudate as a disinfectant.\nOverall, I believe that this otherwise very good manuscript could benefit of a minor revision before indexing in F1000Reasearch. I have attached a commented copy of the manuscript for the authors to consider in the revision.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4076", "date": "26 Oct 2018", "name": "Shirley Kuhnen", "role": "Reader Comment", "response": "1- Six concentrations of leaf exudate were tested. However, under 500 ug/mL, the reduction of microbial growth was lower than 70%, varying between seasons. The statement was modified. 2- Thank you very much. You are correct. This information was included in the text. The statements were modified. 3- Thank you very much. The suggestions were included in the new version." } ] }, { "id": "37690", "date": "14 Sep 2018", "name": "Liliana Cardemil", "expertise": [ "Reviewer Expertise Prebiotic molecules from Aloe vera" ], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe work is interesting since the infection of bovine mammary epithelial cells by Staphylococcus aureus it's a problem in cattle farming.\nHowever, the study they performed is incomplete. I do not understand, why the investigators did not try the Aloin that they purified to determine the inhibition of bacterial growth and the degree of toxicity on mammary cells with this compound as they did with the total leaf exudate? The leaf exudate has too many compounds. Not only phenols, the exudate may have amino acids, anthraquinones, chlorophyll, and other pigments, etc. At least testing the Aloin then you could know if this compound has an antibacterial effect by itself and/or, causes the toxicity of mammary epithelial cells. If the results are positive for Aloin as antimicrobial without causing toxicity they will know that Aloin can be used for the mastitis treatment. If not, they will know that the leaf exudate has to be free of Aloin to treat the mastitis.\n\nThe literature of the manuscript is missing important new and relevant literature on the uses of Aloe barbadensis Miller as a prebiotic and inhibitor of colon cancer development.\n\nIn my opinion the need to perform the analyses considering the aloin. In my opinion, the manuscript should be approved after the authors complete these analyses.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4075", "date": "26 Oct 2018", "name": "Shirley Kuhnen", "role": "Reader Comment", "response": "We chose not to include a control with Aloin in the study. We acknowledge that there are many works involving the purification of compounds from natural products in the field of Pharmacology. However, our objective was to test a complex matrix for its use (in natura) by organic or agroecological small producers of milk. There are several studies in the literature with complex matrices derived from medicinal plants that do not include a control as mentioned by the reviewer. The use of a complex matrix has some advantages, among them the lesser emergence of resistance among microorganisms. We also point out that there are no methodological errors in the study such as the lack of the controls required by the CSLI Guidelines and Standards. Therefore, the lack of a control with Aloin does not invalidate the results obtained in the present study and the work should not be rejected for this reason. The literature suggested doesn’t fit in our paper because it’s related to polysaccharides’ biological effects. The assay was performed according to CSLI Guidelines and Standards. This does not require control with isolated compounds. However, to cater to the suggestion, we added in the text that this may be a line of future research." } ] } ]	1	https://f1000research.com/articles/7-1285
https://f1000research.com/articles/7-1132/v1	25 Jul 18	{ "type": "Research Note", "title": "Use and effectiveness of the Individual Development Plan among postdoctoral researchers: findings from a cross-sectional study", "authors": [ "Nathan L. Vanderford", "Teresa M. Evans", "L. Todd Weiss", "Lindsay Bira", "Jazmin Beltran-Gastelum", "L. Todd Weiss", "Lindsay Bira", "Jazmin Beltran-Gastelum" ], "abstract": "The individual development plan (IDP) is a career planning tool that assists PhD trainees in self-assessing skills, exploring career paths, developing short- and long-term career goals, and creating action plans to achieve those goals. The National Institutes of Health and many academic institutions have created policies that mandate completion of the IDP by both graduate students and postdoctoral researchers. Despite these policies, little information exists regarding how well the tool is used and whether it is useful to the career development of PhD trainees. Herein, we present data from a multi-institutional, online survey on the use and effectiveness of the IDP among a group of 183 postdoctoral researchers. The overall IDP completion rate was 54% and 38% of IDP users reported that the tool was helpful to their career development. Positive relationships with one’s advisor, confidence regarding completing training, one’s confidence about their post-training career, and a positive experience with institutional career development resources are associated with respondents’ perception that the IDP is useful for their career development. We suggest that there is a need to further understand the nuanced use and effectiveness of the IDP in general, to determine how to execute the use of the tool to maximize trainees’ career development, and to generally enhance the career development support for PhD trainees.", "keywords": [ "biomedical research", "career development", "careers in research", "career planning", "individual development plan", "PhD training", "postdoctoral researchers", "science and technology workforce" ], "content": "Introduction\n\nThe Individual Development Plan was first introduced by the U.S. Federation of American Societies for Experimental Biology in 2002, and in 2014 the National Institutes of Health implemented a policy requiring the reporting of the tool’s use by graduate students and postdoctoral researchers in grant progress reports1–3. Also in 2014, 19% of postdoctoral researchers were shown to use the IDP with 71% of those users finding it valuable4. We recently conducted a multi-institutional study on the use and effectiveness of the IDP in 663 PhD students and found that 57% completed an IDP and 51% of users reported it being useful to their career development5. The effectiveness of the tool was associated with a positive relationship with mentors, confidence regarding career plans, and positive interactions with institutional career development resources5. More research is needed to further characterize the IDP among all PhD trainees and to determine if there are ways to maximize the tool’s impact on trainees’ career development5,6.\n\n\nMethods\n\nThese data were collected as part of a broader health and wellbeing online, survey-based study of graduate students and postdoctoral researchers in the spring and early summer of 2016 (March to June). With regard to the use and effectiveness of the IDP, the data collection procedures, survey methodology, data analysis and statistical methods, and study limitations have been described previously5. Briefly, the work was approved by the University of Kentucky (protocol 15-1080-P2H) and University of Texas Health San Antonio (protocol HSC20160025X) institutional review boards. Respondents read a cover page and anonymously consented to the study by engaging the online survey. The survey was distributed via social media and direct email. Eligibility criteria included being a current postdoctoral researcher in the life/biological/medical or physical/applied sciences at a U.S. institution. Subjects responded to the IDP questions using the five-point Likert scale strongly agree, agree, neutral, disagree and strongly disagree. For data analysis, strongly agree and agree responses were grouped together as an agree category and neutral, disagree, and strongly disagree were grouped together into a does not agree category. One-way frequencies were calculated (Supplementary File 2) and the Pearson chi-square test was used to assess the univariate associations between the survey variables and the outcome “I Find the IDP Process Helpful to my Career Development” only among the respondents who completed an IDP (Supplementary File 3). All summaries and statistical analysis were performed in SAS 9.4.\n\n\nResults\n\nAmong 183 total postdoctoral respondents, 45.4% reported being required to complete an IDP, 27.5% reported completing the tool with their PI/advisor, and 33.9% completed the IDP, at some point, without discussing it with their PI/advisor (Figure 1 and Supplementary File 2). In total, 54.1% of respondents actually completed the IDP with or without their advisor (Supplementary File 2 and Supplementary File 3). Further, 24.3% of all respondents reported being able to have an honest conversation with their PI/advisor in the context of the IDP process and 22.4% of all respondents found the IDP helpful to their career development (Figure 1 and Supplementary File 2). Of the respondents that completed an IDP, 38.4% found the tool helpful to their career development (Supplementary File 3).\n\nShown here are rates for variables measuring whether respondents are required to complete an IDP, complete an IDP annually with their PI/advisor, complete an IDP but do not discuss it with their PI/advisor, can have an honest conversation with the PI/advisor in context of the IDP, and whether the IDP process is helpful to their career development. One-way frequencies for all other survey variables can be found in Supplementary File 2.\n\nAs we have recently shown with PhD students5, the effectiveness of the IDP among its users is associated with positive mentorship relationships (Figure 2 and Supplementary File 3). For example, 62.2% of those who indicated that they could have an honest conversation with their PI/advisor found that the IDP process was helpful to their career versus 24.2% of those who did not agree (p < 0.001). Likewise, 56.7% of those who indicated that their PI/advisor positively impacts their emotional/mental wellbeing versus 30.4% of those who did not agree found the IDP process to be helpful to their career (p = 0.01). IDP effectiveness was also associated with confidence regarding the completion of training, being prepared for one’s post-training career, and positive interactions with career development resources (Figure 2 and Supplementary File 3).\n\nIDP effectiveness was assessed among the subset of respondents who completed an IDP by determining the univariate associations between the survey variables and the outcome “I Find the IDP Process Helpful to my Career Development.” The Pearson chi-square test was used to measure statistical significance. * p < 0.001; p ≤ 0.01; * p ≤ 0.05.\n\n\nDiscussion\n\nCompared to the 2014 study in which 19% of postdoctoral researchers used the IDP and 71% of users found it valuable4, the current data suggests that there has been a general increase in IDP usage among postdoctoral researchers with 54.1% of respondents in this study indicating that they are required to complete an IDP while its perceived value has decreased to less than 40%.\n\nIn general, the trends presented here for postdoctoral researchers are similar to our recent findings on use and effectiveness of the IDP in PhD students5, but there are some nuanced differences. For example, compared to the rates in PhD students, the rates of required completion of the IDP among this study’s postdoctoral researchers are lower; the rates of completing the IDP but not discussing it with a PI/advisor are higher; and the rates of reporting that the IDP process is helpful to one’s career development are lower. The correlation of IDP effectiveness and mentorship relationships and use of career development resources are similar between PhD students and postdoctoral researchers. It will be important to conduct additional studies to further delineate differences and similarities in the usage and effectiveness of the IDP between PhD students and postdoctoral researchers.\n\nWhile this work will add to our understanding of the IDP, there are some limitations to the study including the potential lack of generalizability across all institutions and/or fields of study and potential data/outcome bias. Additionally, this study may not capture all the issues related to the IDP, respondents may not be aware of their institution’s IDP policies, the IDP structure and processes may vary within and between institutions, and the measure of the effectiveness of the IDP herein is subjective. Subjects’ responses may also reflect multiple experiences with the IDP during their training.\n\nAs we have previously recommended5, additional research is needed to further understand the use and effectiveness of the IDP. Further, in order to enhance career development support for PhD trainees, faculty should receive mentorship training and career development infrastructure should be improved.\n\n\nData availability\n\nDataset 1. Individual Development Plan survey data. Columns Q1–Q26 correspond to the questions listed in Supplementary File 4. 10.5256/f1000research.15610.d2256987", "appendix": "Grant information\n\nN.L.V. is supported by the University of Kentucky’s Cancer Center Support Grant [NCI P30CA177558] and the Center for Cancer and Metabolism [NIGMS P20GM121327]. T.M.E is supported by the University of Texas Health Science Center San Antonio's Science Education Partnership Award [NIGMS R25GM129182].\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe thank the Markey Cancer Center Research Communications Office for formatting and graphic design assistance; Dr. Paula Chambers, Versatile PhD, for her input on and aid in distributing the study survey; and the Graduate School of Biomedical Sciences at the University of Texas Health San Antonio for providing partial funding for the study.\n\n\nSupplementary material\n\nSupplementary File 1. Self-reported institution of all respondents.\n\nClick here to access the data.\n\nSupplementary File 2. One-way frequencies of all respondents, separated by demographic characteristics.\n\nClick here to access the data.\n\nSupplementary File 3. Univariate analysis of the survey’s variables and the perception of Individual Development Plan helpfulness among respondents who completed an Individual Development Plan.\n\nClick here to access the data.\n\nSupplementary File 4. Example copy of the survey questions relevant to this study.\n\nClick here to access the data.\n\n\nReferences\n\nClifford PS: Quality Time with Your Mentor. The Scientist. 2002; 16: 59. Reference Source\n\nHobin JA, Fuhrmann CN, Lindstaedt B, et al.: You Need a Game Plan. Science. 2012. Publisher Full Text\n\nNational Institutes of Health: Revised Policy: Descriptions on the Use of Individual Development Plans (IDPs) for Graduate Students and Postdoctoral Researchers Required in Annual Progress Reports beginning October 1, 2014. 2014. Reference Source\n\nHobin JA, Clifford PS, Dunn BM, et al.: Putting PhDs to work: career planning for today's scientist. CBE Life Sci Educ. 2014; 13(1): 49–53. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVanderford NL, Evans TM, Weiss LT, et al.: A cross-sectional study of the use and effectiveness of the Individual Development Plan among doctoral students [version 2; referees: 2 approved, 1 approved with reservations]. F1000Res. 2018; 7: 722. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTsai JW, Vanderford NL, Muindi F: Optimizing the utility of the individual development plan for trainees in the biosciences. Nat Biotechnol. 2018; 36(6): 552–553. PubMed Abstract \| Publisher Full Text\n\nVanderford NL, Evans TM, Weiss LT, et al.: Dataset 1 in: Use and effectiveness of the Individual Development Plan among postdoctoral researchers: findings from a cross-sectional study. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15610.d225698" }	[ { "id": "36439", "date": "09 Aug 2018", "name": "Jonathan S. Wiest", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn this study, the authors seek to determine the effectiveness of utilizing an IDP during the training stage of a biomedical career. Using survey data from a larger online study of overall health and well-being, 183 trainees responded to questions regarding the IDP using the Likert scale. While 45.4% of the respondents were required to complete an IDP, in total, 54.1% completed an IDP (with or without their mentor’s support). Only of third of those who completed the IDP, however, found the tool useful for career development. Most notably, the authors found positive correlation between positive mentoring and the effectiveness of utilizing an IDP.\n\nOverall, this is a well written manuscript addressing an important topic in the training community. However, there are several points that this reviewer found to be confusing.\n\nFirst, the authors state that 54.1% of respondents completed an IDP, however, based on the data in Figure 1 and Supplementary File 2, 112 respondents (61%) of respondents completed an IDP. Further, in Figure 2, the authors are basing their conclusions on 99 respondents. While this is 54% of the 183 respondents, it is unclear why the total of 99 was used as opposed to the 112.\n\nSecond, the authors state that 38.4% of respondents who completed an IDP found the tool to be useful to their career development and reference Supplementary File 3. Yet, upon closer inspection, a question about IDP usefulness is missing from that document altogether. Thus, it is unclear where that percentage was derived. The data in Supplemental File 2 shows that 22.4% of the total population found the tool effective, however, that is reflective of the total population, and not those who completed the IDP. Taking those findings into consideration, the percentage of those who utilized the IDP and found it effective is 36.6%.\nLastly, it is unclear how Figure 2 is an analysis of IDP effectiveness. To the authors credit, it is noted that effectiveness is subjective and the IDP structure can be a confounding factor. However, the questions in Figure 2 are more indicative of mentor effectiveness, and not that of the IDP. It is important to note that the respondents who have positive relationships with their mentors seem to be better prepared, but only half of those find the IDP effective.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Partly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4077", "date": "25 Oct 2018", "name": "Nathan Vanderford", "role": "Author Response", "response": "Dear Drs. Wiest and Case Borden, Thank you for your review. Your critique has been very helpful as we have revised the article. Below we address the major issues you raised. We apologize for the confusion regarding the percent of respondents that completed an IDP. We analyzed the unique responses to questions 2 and 3 in our survey (please refer to the survey instrument within Supplemental File 4) to arrive at both the number and the percentage of respondents that had completed an IDP. Out of the 112 respondents that agreed to both survey questions 2 and 3, 13 agreed to both questions and these respondents were subtracted from the total to obtain the number and percentage of total respondents who completed an IDP (112 – 13 = 99; 99/183 = 54.1%). Supplemental File 2 reports on the data from all respondents while the univariate association analysis shown in Supplemental File 3 reports on the data only from the 99 unique respondents that reported completing an IDP. We have clarified this in the new version of the article. We likewise apologize for the oversight of not listing the percentage (38.4%) of respondents that had used the IDP and found it useful to their career development in Supplemental File 3. We now include this data in the top portion of the table found in Supplemental File 3. It is important to note that this percentage (38.4%) is based only on those respondents that had completed an IDP (again based on unique respondents to questions 2 and 3 in the survey). The frequency data presented in Supplemental File 2 is not additive because of the 13 respondents who agreed to both questions 2 and 3. We have clarified this in the new version of the article. Figure 2 does show a set of two category-level variables that are associated with IDP effectiveness. The asterisks in particular point to the significant differences in the proportions of the outcome (IDP effectiveness) among the levels of a given variable using the Chi-square test of proportions. We measured the outcome, IDP effectiveness, by asking respondents the question “I Find the IDP Process Helpful to my Career Development” Again, we have clarified this in the new version of the article. In closing, we have revised the article to address your comments as well as those of the other reviewers. We hope that you will favorably review the revised version of the article in light of our changes. Thank you again for your comments as we strongly feel that they have strengthened our work. We look forward to reading your next review. Sincerely, Nathan L. Vanderford" } ] }, { "id": "36897", "date": "30 Aug 2018", "name": "Tammy R. L. Collins", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors report on the usage of IDPs by postdoctoral scholars, which is both a timely and fundamental topic within the broader graduate and postdoctoral professional development community. This work extends beyond the authors’ recently published article on IDP usage among doctoral students1 to showcase how the instrument is currently being used by postdocs. The aforementioned manuscript on doctoral IDP usage extensively discusses policy and other issues surrounding IDPs, while this manuscript is lean on discussion. It would therefore benefit from including prior literature on correlates of success associated with postdocs who have written plans—which would have the added benefit of placing this work into a broader context (for example, see: Davis 20092).\n\nThere are a number of points that should be addressed, and they are outlined as follows:\nThe survey instrument asks questions on a 5-point Likert scale of strongly agree, agree, neutral (neither agree nor disagree), disagree, & strongly disagree. However, when analyzing the data, the authors report percentages either as ‘does not agree’ (also reported as ‘disagree’) or ‘agree’ - with ‘strongly disagree, disagree & neutral’ all grouped together as a ‘does not agree’ response. It seems that lumping “neutral (neither agree nor disagree)” into the ‘does not agree’ category would skew results (both in this manuscript and the manuscript on doctoral IDP usage) towards ‘does not agree.’ It is recommended that the authors reanalyze the data and report the ‘neutral’ responses as a third category in order to more accurately reflect intended answers/percentages.\n\nI agree with reviewer 1 that it is unclear how the authors calculated that 54% completed an IDP; like reviewer 1, I also calculated that 61% completed an IDP either with or without their PI. Similarly, I also agree with reviewer 1 that it is unclear where the 38.4% value came from (percent of respondents that completed an IDP who found the tool helpful). Furthermore, in figure 1, a value of 22.4% of all respondents (whether they complete an IDP or not) is listed as saying the IDP process is helpful to their career development. Since “22.4%” also includes those who never completed an IDP, reporting it in this manner seems to bias perceptions of the tool as unhelpful.\n\nAside from the unclear derivation of percentages discussed in point #2, the questions in the survey instrument do not seem to allow confident discernment of who actually completed an IDP. For example, respondents who disagree with the question “I complete an IDP annually with my PI/advisor” could have actually completed an IDP with their PI at the beginning of their postdoc and/or possibly in year 3 of their postdoc—how much weight were respondents giving to the word ‘annually’? Another question reads: “I complete an IDP but I do not discuss it with my PI/advisor” - could respondents complete two different IDPs - one with and one without their PI (and thus agree to both the former and latter questions)?\n\nThe authors assert that the usefulness of an IDP has decreased since 2014, specifically comparing their results to that in (Hobin et al. 20143). However, in the (Hobin et al. 2014) paper, an IDP’s overall value is reported either as ‘not helpful’ ‘neutral’ or ‘helpful’. It seems that the results in this manuscript could be compared more accurately with the (Hobin et al. 2014) data by showing the ‘neutral’ responses, rather than lumping them with ‘does not agree’ (see point 1 above). Additionally, it would be helpful to point out the limitations of comparing these two studies (ex: address key differences between the two survey instruments regarding how the IDP questions were asked (and how this might bias responses), address potential respondent audience differences, etc.).\n\nLike reviewer 1, I also feel that there are limited questions that address what one might consider IDP ‘effectiveness.’ The manuscript would thus be strengthened by discussing IDP indicators that have been previously reported in the literature (such as measures outlined by (Hobin et al. 2014) - ex: the value of an IDP in helping with self-assessment, helping identify career paths, helping identify skills to strengthen, etc.). Furthermore, (Davis 20092) reports an in-depth analysis of results from a Sigma Xi Postdoc Survey - identifying many positive correlates of success associated with postdocs who develop a written plan at the outset of their careers (ex: higher publication rate, grant submission rate, better supervisor relationships, etc.). Since the main point of this manuscript is to discuss the use and effectiveness of IDPs among postdocs, it would benefit from elaborating upon such postdoc-specific contextual literature - as well as other literature that documents the general benefits of goal-setting, which is a primary function of IDPs (Locke et al. 20024).\n\n(Minor) The authors indicate that “additional research is needed,” and it would thus be beneficial to clarify the research questions that should be addressed. For example, it seems that it would be useful to determine the effects of various parameters on IDP effectiveness such as: 1) when in training an IDP is completed; 2) inclusion/ exclusion of IDP components (such as self-assessment, career exploration, skill-building, goal-setting, etc.); 3) prior experience with/completion of an IDP as a PhD student; 4) completing an IDP of their own accord versus doing so because it is required; 5) receiving training on how to create/implement an IDP; 6) using a specific IDP instrument (ex: myIDP) versus an institutionally-developed IDP, etc..\n\n(Minor) As an additional point - since ‘IDP effectiveness’ is subjective as the authors point out, perhaps future studies should address better-defining these parameters so that common IDP evaluation methods can be adopted within the broader community. It would also be especially helpful to ascertain what IDP tools are being used, and how they are being implemented so that standard ‘correlates of effectiveness’ could be tied to specific IDP instruments (or components) and the manner in which they are employed.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Partly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4078", "date": "25 Oct 2018", "name": "Nathan Vanderford", "role": "Author Response", "response": "Dear Dr. Collins, Thank you for your review. Your comments and critique have been very helpful in guiding our revisions. We respond to your major points below. We appreciate your comments and concerns regarding our decision to combine the neutral Likert scale responses with the disagree and strongly disagree responses. As such, we have now split these responses out and present three categories (agree, neutral, and disagree) in our analysis. The applicable dataset, figures, supplemental files, and text have been revised accordingly. As mentioned in our response to Drs. Wiest and Case Borden, we sincerely apologize for the confusion over our reporting of the percent respondents to several questions. In particular, we analyzed the unique responses to questions 2 and 3 in our survey (please refer to the survey instrument within Supplemental File 4) to arrive at both the number and the percentage of respondents that had completed an IDP. Out of the 112 respondents that agreed to both survey questions 2 and 3, 13 agreed to both questions and these respondents were subtracted from the total to obtain the number and percentage of total respondents who completed an IDP (112 – 13 = 99; 99/183 = 54.1%). Supplemental File 2 reports on the data from all respondents while the univariate association analysis shown in Supplemental File 3 reports on the data only from the 99 unique respondents that reported completing an IDP. We have clarified this in the new version of the article. It is important to note that the frequency data in Supplemental File 2 is not additive because of the 13 respondents who agreed to both questions 2 and 3. In retrospect, we agree that our questions can make it difficult to discern which respondents uniquely completed an IDP. In hindsight, we could have asked a simple yes/no question about whether trainees had completed an IDP. That said, we are confident that our method of combining and de-duplicating the responses to questions 2 and 3 allow us to determine which of our respondents have completed an IDP. The new analysis of our Likert scale data allows us to more clearly and directly compare our results to the 2014 Hobin et al. data. Additionally, we have also specified in our limitations section that care should be taken in making such comparisons because of the analysis of different populations (e.g., although both populations were postdocs, there could be institutional differences, etc.) and different study designs. We have added a bit of additional text and references to better contextualize our work, to further clarify some limitations of the study, and to better define future research questions. Of note, however, this article was submitted as a short Research Note article type which has a 1,000 word limit and thus space is limited regarding adding a comprehensive literature review related to PhD trainee career development. We have thus focused on the IDP-related literature. Thank you again for your review. We look forward to your second review in light of our revisions. We feel that your comments and that of the other reviewers have strengthened the article. Sincerely, Nathan L. Vanderford" } ] }, { "id": "37952", "date": "14 Sep 2018", "name": "Adriana Bankston", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nGeneral comments: This publication addresses the very important topic of use and effectiveness of the individual development plan (IDP) for trainees. This is not a trivial topic to address and I commend the authors for this analysis in the context of current literature. I also appreciated the valuable insights on how the effectiveness of the tool is associated with positive mentoring, pointing out the need for strong relationships with advisors that can affect career trajectories for graduate students and postdocs.\nGeneral considerations: On the background side, not being terribly familiar with the already existing literature on this topic, I would recommend a bit more description of prior studies that could frame this work, and its novelty in the context of existing literature. However, I recognize that it’s also possible literature on the effectiveness of the IDP may be limited, in particular given the abstract nature of this concept. Without having read more background, but judging from the information presented in the introduction, my immediate impression is that prior studies looked at PhD students whereas this work examines postdoctoral researchers. Moreover, the authors mentioned that this was a multi-institutional analysis, which leaves me wondering whether the novelty of this work is looking at a different population (postdocs vs. graduate students) or whether it is related to the number or type of institutions that prior studies hadn’t examined in this context.\nOn the technical side, the abstract mentions that the authors looked at data from 183 postdocs, although from the methods section it appears that both graduate students and postdocs were examined for a period of March to June 2016. Given that a study of 663 PhD students was previously performed as referenced in the introduction, it would be helpful to see an articulation of the novelty that this current study brings over previous work. It would also be beneficial to clarify whether both populations were included in all the analyses in this publication (in particular figure 2, which appears to refer to PhD students, whereas figure 1 refers to postdocs), and to expand more on the similarities and differences between the effectiveness of the IDP on these two populations. While this is mentioned in the discussion section, if both graduate students and postdocs were analyzed more in-depth in these studies (also taking into account other aspects of the data in this publication), it would make for an interesting comparison as to whether or not the IDP has more of an effect on the career trajectories of one group or another. And while positive mentoring is required for both populations, postdocs may already be well on their way towards a more obvious career path than PhD students are. While the authors state similar trends in these topics, it would be helpful to take this analysis a step further and determine how IDP effectiveness affects career choices for these groups.\nIn terms of data analysis, the authors state that they grouped together the agree/strongly agree, and neutral/disagree/strongly disagree responses. I wondered why this is the case, perhaps it could be a low sample size that may not enable meaningful conclusions. However, in order to gain a fully comprehensive picture of the issue at hand, I would recommend displaying and analyzing data from each of these categories separately. I believe that for such a topic that is difficult to quantify, it will be important to dissect the prevalence of each of these responses. Examining the various categories (availability of programs, attendance and usefulness) in Supplementary Data 2 could be utilized for a more thorough analysis of this topic. While I wonder how usefulness can be assessed in a practical sense, it was also disheartening to see the percentage of respondents who indicated that they do not attend or did not find available programs helpful. Perhaps this is an area that should be further explored in terms of which programs would likely be helpful for trainees to have. It would also be interesting to dissect further the correlation between the usefulness of the IDP and subsequent career paths chosen. For example, did individuals who found the IDP helpful end up in the top career path predicted by the IDP, and are they currently satisfied in their position? If so, this might indicate that the IDP was useful in helping them achieve desired career goals. The IDP could also open them up to career paths they hadn’t considered before, which would demonstrate the added value of this tool for training and career development.\nI appreciated the data transparency in this publication (for example in Dataset 1, Supplementary Data 4). Given the wealth of information and number of questions asks, further analyses of these existing data looking at the effect of other variables on IDP effectiveness would provide a thorough analysis of how we might improve the IDP process based on barriers faced by particular groups. I would also suggest detailing the data analysis and quantification procedure used in this publication (as opposed to referencing prior publications with the information), in order to clarify how percentages in the results were calculated.\nI also wonder whether it’s possible to examine other variables together to make predictions that would enrich this publication in the future. For example, how do factors such as race, ethnicity, U.S. citizenship, and others, affect the responses to the IDP process (Supplementary Data 3). While these may not have been the primary objectives of the authors, this type of analysis would add another layer of complexity to whether the IDP is useful to various groups, whose career decisions may also be affected by additional factors. This publication does contain a large amount of raw data that I think could be utilized for a more thorough analysis of how various factors contribute to the effectiveness of the IDP. However, given the data is self-reported and there may be a limited number of responses in each of these categories, it may be difficult to assess the effect of such variables with the current dataset.\nFeedback on results: It was somewhat disheartening to see the percentage of postdocs who had completed the IDP without discussing it with their advisor (Figure 1), and the percentage of those that had honest conversations with their advisors was also not terribly high. These factors point to barriers towards positive mentoring relationships in academe, as well as obstacles to career development for trainees. They could affect the ability of trainees to follow desired career paths, or having to prepare for transitioning into these careers without their advisors knowing, especially if the advisor does not approve of their non-academic career choice. This fundamentally points to systemic flaws in academe and how the enterprise needs to change in order to better support trainees who are using the IDP as a guide to explore career options. Importantly, this also requires advisors to point trainees in the right direction, and be a sounding board during career transitions. I also wonder whether there is a connection between the lack of discussions with the advisor and their ability to have honest conversations (Figure 1), as it appears that this could be a layered response (i.e. they were either having or not having these conversations, and if they were, how honest did the trainees feel they could be with their advisors in terms of desired career options?). I think drawing a connection between these two variables could be valuable to investigate in terms of the barriers affecting the ability of trainees to pursue various career paths, and assess the usefulness of the IDP process for these particular careers.\nFigure 2 somewhat addresses this concern, and it was valuable to see that positive mentoring and having honest conversations with advisors can influence the responses of trainees on IDP effectiveness. There is a lot of really valuable information in this figure in terms of how we can improve faculty training to be more supportive of the career choices of their trainees, so that they feel valued and prepared for taking on other careers besides academia. Given the importance of these factors, it would also be interesting in the future to look at how positive relationships with advisors affect other aspects of training and career preparation for trainees. While factors such as the advisor being an asset to their career, providing ample support and positively impacting their emotional and mental well-being, among others variables, are likely very difficult to assess, I believe these are critical investigations that should be pursued further and more in-depth to better understand how to train the next generation of researchers. Along the lines of these ideas, putting these findings into a larger context would be really helpful in discussing how to better equip faculty to help trainees be successful in their desired careers.\nAdditional recommendations: It was interesting to learn about the comparison between IDP use and effectiveness for PhD students and postdocs, as detailed in the discussion section. I was surprised to see that postdocs weren’t required to complete the IDP to the same extent that PhD students were, did not discuss the results of completed IDPs with their advisor as much, and found the IDP less helpful for their career development. This observation that merits further investigation, as to whether the lack of usefulness of the IDP for postdocs was due to their inability to discuss it with their advisors, or whether other factors were also involved. I would also be curious to know more about why there is a lesser requirement for postdocs to complete the IDP, and whether reversing this trend would result in a greater percentage of postdocs actually pursuing desired career paths as opposed to traditional academic routes.\nIn terms of comparing data from PhD students and postdocs, I wonder whether these surveys and subsequent analyses were performed on both populations at the same time (during March to June 2016 as described in the methods) or whether the data discussed here on PhD students came from a previous publication. This analysis could also provide insights into whether we should target certain populations more in terms of IDP assessments, and which populations within academe the IDP is likely to be more useful for in terms of career exploration. For a more extensive analysis, it would also be interesting to compare all of the aspects in Figure 2 between PhDs and postdocs, in order to determine the effect of mentoring relationships on career trajectories of trainees at various stages in their careers.\nI appreciated that the authors pointed out limitations of the study in the discussion section, including as it relates to institutional variability. Indeed, Supplementary File 1 indicates that there are very few individuals at the institutions shown in the dataset, and many are at missing institutions. I imagine there is also quite a variability between these institutions in terms of size, number of postdocs, and the type of career development opportunities available that could supplement the IDP findings for trainees. These variables could influence how trainees rate the usefulness of the IDP, in terms of whether additional resources exist for them to further explore careers that were indicated as a good fit from the IDP. For example, it is possible that a limited knowledge on available career options, either due to the lack of resources or the inability of their advisor to help (in addition to not being able to find another suitable mentor to assist with career exploration), trainees may rate the usefulness of the IDP as lower than those with more external information available.\nIncreasing the sample size of respondents from each institution would also provide a clearer picture of how institutional environments affect career trajectories for trainees. In addition, incorporating other variables into the evaluation of institutions would enable various types of comparisons to be made about IDP effectiveness by trainees from diverse backgrounds, or those in institutions of a certain size or geographical area. These are also factors that could affect their career development - for example a larger city might offer opportunities to interact with other postdocs and take advantage of multi-institutional career development opportunities, which trainees in other geographical areas may not have access to.\nBroader picture comments: I agree with the authors that faculty should receive mentorship training and it would be helpful to see further elaboration by the authors on how this could be achieved. For example, mentorship training for faculty could include manuals with both internal and external resources and contacts from various career paths that trainees might want to pursue, thus enabling them to better train their postdocs for appropriate careers. There is also currently the barrier of trainees not being able to have honest conversations with advisors about their career options, therefore faculty attitudes need to change in order to allow postdocs to pursue non-academic careers.\nI also agree that a better career development infrastructure is needed, and that this would be a massive undertaking. Incorporating the findings from this publication, however, into current literature on these topics (and efforts made by others to reform career training in universities) would be a helpful beginning to understanding how such an infrastructure could be developed. Implementing the IDP as a mandatory training for postdocs at the bench as part of their annual assessment may already be happening at some universities, however we should also keep in mind that IDPs are really only the beginning of the career development process - while it is the responsibility of trainees to utilize their IDP results for further career exploration, an infrastructure that supports this process is imperative to their success. This infrastructure I envision could be internal to the university, or there could be an external entity developing resources for several universities to utilize for helping trainees explore career options. As part of this work, the authors could also consider developing a rubric to measure IDP effectiveness as it relates to their ability to achieve career goals outlined in the IDP. I would also be curious to see a rubric for assessing other factors that can influence this effectiveness (such as those in Figure 2) and trying to understand more about particular elements that go into each of these factors.\n\nOverall, this is a well written manuscript tackling an issue that is difficult to quantify but very important to study from the context of training the next generation of scientists. I believe that more of an in-depth literature overview, further analysis of the existing data and collection of additional data, and a more extensive discussion of the recommendations for change around faculty training to support postdocs, would greatly strengthen the manuscript in the future. I believe these findings are a valuable foundational start to these questions, conducting further investigations on this topic can provide more in-depth understanding of the potential that the IDP could have for training graduate students and postdocs for being successful in their chosen careers.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4079", "date": "25 Oct 2018", "name": "Nathan Vanderford", "role": "Author Response", "response": "Dear Dr. Bankston, Thank you for your very extensive report. We have responded to your major critiques/comments below. We have added a bit more text and references to better contextualize our work. We apologize for the confusion over the study population. In this work, we report only on data obtained from postdoctoral trainees. Our previous report was focused on IDP use and effectiveness in PhD students. These data were collected at the same time, but we choose to analyze the postdoctoral trainee and PhD student data separately. As mentioned in several of the responses to other reviewers’ comments, we have now broken out the neutral responses to our Likert scale questions and we now present these data separately. All the applicable text, data, figures and supplemental files have been updated accordingly. Much of your additional comments focus on additional data analysis and comparisons of the postdoctoral trainee and PhD student data. We agree that your suggestions are very important and you have posed very interesting and essential questions. We, however, feel that your suggestions are out of scope for the current study. This study was submitted as a short Research Note (which has a 1,000 word limit). These article types are meant to convey findings that can be described in a short report. One of our goals of this work was to obtain preliminary findings that can inform other work on the use and effectiveness of the IDP. Additional IDP use and effectiveness data that should be collected with a revised survey instrument that is informed by our work will allow for such additional analyses in the future. Thank you for your time and comprehensive report. We hope that you will favorably consider our revisions in light of our changes that address your major critiques and those of the other reviewers. We look forward to reading your next review. Sincerely, Nathan L. Vanderford" } ] }, { "id": "37951", "date": "18 Sep 2018", "name": "Kristen L. W. Walton", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article summarizes data from a subgroup of individuals surveyed about their use of an Individual Development Plan and other factors. Data on the effectiveness and usefulness of the IDP is important to justify policies that require postdoctoral scholars and PhD students to complete an IDP. The findings are interesting and concisely presented. The authors appropriately acknowledge several limitations to this survey. Overall, this manuscript adds important data to a field that is very difficult to quantify, given the variability in the IDP across institutions and training programs. I do have some questions and suggestions to strengthen this manuscript:\n\nAs noted by other reviewers, the separation of the Likert scale data into “agree” and “disagree” categories, with “neutral” included in the “disagree” category, has the potential to skew results towards the “disagree” category. It would be helpful to analyze the data with neutral responses listed as a separate category. The survey population demographics as reported in Supplementary File 2 show that the population of respondents was 80.7% white. How do the demographics of the survey population reflect the national postdoc population demographics? The numbers in some categories are likely too small to analyze in a statistically meaningful way, but it would be interesting to determine whether different demographic groups (race, gender, etc) had similar responses regarding the usefulness of the IDP and/or mentoring relationships.\n\nI agree with other reviewers that this paper has relatively minimal introduction and discussion to place it in the context of other work. The issues faced by postdocs are not identical to those faced by PhD students, and there are multiple recent publications discussing the problems facing postdocs (for example, The Postdoc Experience Revisited, National Academies Press 20141; Alberts et al, PNAS 20142). The data in this manuscript that show that only 22.4% of survey respondents felt that the IDP process was helpful to career development suggest that this process may not be an effective strategy for improving the postdoctoral experience and outcomes.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Partly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4080", "date": "25 Oct 2018", "name": "Nathan Vanderford", "role": "Author Response", "response": "Dear Dr. Walton, Thank you for your review of our article. We have responded to your major concerns one-by-one below. As mentioned in several of the other responses to reviewers’ comments, we have reanalyzed our data using the neutral responses as a separate category. All the text, data, figures, and files have been updated accordingly. We did collect data on the race and ethnicity of our respondents and we observed no significant differences between the groups regarding their response to whether they found the IDP helpful to their career development. We agree that our sample size may be too low to definitively draw any hard conclusions in this regard, however. It will be interesting to re-visit this question with a much larger sample size. One could envision differences given what is known about minority populations and the training and career outcome pipelines. Understanding these differences is critical in order to develop interventions that can fit the needs of specific populations. We have added a bit of additional text and literature to further contextualize our work. Of note, however, this article type has a 1,000 word limit and thus there is limited space to house a comprehensive literature review on all the related trainee career development topics. As such, we have focused on discussing the pertinent IDP literature. Thank you again for your time and review. We look forward to your second review. Sincerely, Nathan L. Vanderford" } ] }, { "id": "37949", "date": "18 Sep 2018", "name": "Richard McGee", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAs noted by Review 1, the manuscript is well written and easy to follow. However, I would agree with both of the other reviews that problems arise from both the survey questions and decisions on method of analysis. The first 3 questions about completing an IDP don’t really lead to a clean number of how many people do and do not complete one. A person could answer no to being required to complete an IDP but could be doing it voluntarily. And the issue of doing it at all vs. doing it annually makes the numbers even more ambiguous and difficult to sort out. An additional challenge comes from the 5 choices for questions 1-3 when they are really only yes/no situations. It is hard to imagine what could lead a person to be neutral on these 3 questions.\nI would also agree with the other reviewers that there is no rational for combing neutral responses with negative responses. This will skew the interpretation to a negative side without any evidence the respondent meant it to. This is one reason surveys often don’t provide the neutral point as the data are very difficult to interpret. At least, as reviewer 2 points out, including neutral as a distinct option would allow readers to reach their own conclusions.\nI also agree that % responses to questions 4 and 5 should be based only on those who actually complete the IDP as others really don’t have any basis for judgement. But as noted above and by other reviewers, this number of those who completed it is illusive from the question designs.\nRE: Figure 2, as pointed out by the other reviewers, it really does not reveal effectiveness of the IDP. At most it displays associations between some of the questions. I also would raise a concern with the 2 questions about mentors: “My PI/advisor provides real mentorship” and “My PI/advisor provides ample support”. Both of these are very ambiguous – e.g. is the intent to separate ‘real’ mentorship from some other form of mentorship? And what kind of ‘support’ – financial, psychosocial, professional? This level of ambiguity adds to concerns for including neutral responses with the disagree categories because neutral could easily reflect not knowing what the questions are asking for.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate? Partly\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4081", "date": "25 Oct 2018", "name": "Nathan Vanderford", "role": "Author Response", "response": "Dear Dr. McGee, We thank you for your review, which has guided our revisions. We respond to your major comments below. In retrospect, we agree with your comments regarding our survey questions related to discerning a number and percentage of respondents who completed an IDP. Looking back, we could have asked a simple yes/no question(s). That said, what we have done in our analysis is to use the unique responses to questions 2 and 3 within our survey (please refer to the survey instrument within Supplemental File 4) to arrive at the number and percentage of respondents that completed an IDP. Out of the 112 respondents that agreed to both survey questions 2 and 3, 13 agreed to both questions and these respondents were subtracted from the total to obtain the number and percentage of total respondents who completed an IDP (112 – 13 = 99; 99/183 = 54.1%). While not as clear-cut as a simple yes/no question, we are confident that this approach allows us to discern the number/percentage of our respondents who completed an IDP. We also agree with you and several other reviewers regarding the analysis of the Likert data. We have now split out the neutral responses as an independent category and we have updated the text, data, figures, and files accordingly. We apologize for the lack of clarity regarding the analysis of variables that associate with IDP effectiveness with respect to analyzing only those respondents that completed an IDP. In fact, this is how we designed the analysis and we have clarified this in the text. Figure 2 does indeed show variables that correlate with IDP effectiveness. We measured IDP effectiveness by asking respondents the question “I Find the IDP Process Helpful to my Career Development” and then we used this as an outcome variable to understand if variables such as mentorship associate with IDP effectiveness. Again, we have clarified this in the new version of the article. We agree that there are levels of ambiguity in some of our questions. This was, in some cases, by design. In hindsight, however, we could have clarified some of the questions. That said, we believe that the lessons learned from our study design (including the survey design) and our data/findings will be informative to future studies that look to better understand the use and effectiveness of the IDP. Thank you again for your critique. We believe that your comments and those of the other reviewers have improved our work. We look forward to your next review. Sincerely, Nathan L. Vanderford" } ] } ]	1	https://f1000research.com/articles/7-1132
https://f1000research.com/articles/7-247/v1	28 Feb 18	{ "type": "Research Article", "title": "Features of the waterpipe tobacco industry: A qualitative study of the third International Hookah Fair", "authors": [ "Neil Singh", "Mohammed Jawad", "Andrea Darzi", "Tamara Lotfi", "Rima Nakkash", "Benjamin Hawkins", "Elie A. Akl", "Neil Singh", "Andrea Darzi", "Tamara Lotfi", "Rima Nakkash", "Benjamin Hawkins", "Elie A. Akl" ], "abstract": "Background: Little research has been done to uncover the features of the waterpipe tobacco industry, which makes designing effective interventions and policies to counter this growing trend challenging. The objective of this study is to describe the features of the waterpipe industry. Methods: In 2015, we randomly sampled and conducted semi-structured interviews with representatives of waterpipe companies participating in a trade exhibition in Germany. We used an inductive approach to identify emerging themes. Results: We interviewed 20 representatives and four themes emerged: industry growth, cross-industry overlap, customer-product relationship, and attitude towards policy. The industry was described as transnational, generally decentralized, non-cartelized, with ad hoc relationships between suppliers, distributors and retailers. Ties with the cigarette industry were apparent. The waterpipe industry appeared to be in an early growth phase, encroaching on new markets, and comprising of mainly small family-run businesses. Customer loyalty appears stronger towards the waterpipe apparatus than tobacco. There was a notable absence of trade unionism and evidence of deliberate breaches of tobacco control laws. Conclusion: The waterpipe industry appears fragmented but is slowly growing into a mature, globalized, and customer-focused industry with ties to the cigarette industry. Now is an ideal window of opportunity to strengthen public health policy towards the waterpipe industry, which should include a specific legislative waterpipe framework.", "keywords": [ "waterpipe", "shisha", "hookah", "tobacco", "smoking", "industry" ], "content": "Introduction\n\nUsing the waterpipe (also commonly known as narghile, hookah and shisha) for tobacco smoking has been commonplace in Asia and North Africa for centuries. Its popularity has grown in Europe and North America in the last two decades, against a backdrop of a plateauing or decreasing cigarette prevalence1–3. The proportion of adolescents who have ever smoked waterpipe tobacco was between 4 and 33% in Africa and Asia, between 6 and 11% in the USA, and up to 38% in some UK communities4. Waterpipe tobacco use amongst young youth in Jordan increased from 14.0% to 22.6% between 2008 and 2011, with similar patterns seen in Lebanon5, Canada6,7, and the US8.\n\nToxicological studies have consistently shown that waterpipe tobacco use exposes smokers to significant quantities of tar, nicotine, carbon monoxide and carcinogens9. One meta-analysis of 17 studies measuring toxicant exposure from waterpipe tobacco showed that a session contains about 4.1mg nicotine, 619.0mg of tar, and 192mg of carbon monoxide10. Unsurprisingly, epidemiological studies have shown associations between waterpipe tobacco use and several cancers, respiratory diseases, cardiovascular diseases and low-birth weight11. These harms are compounded by the fact that the manufacture, marketing and consumption of waterpipe tobacco is not adequately regulated, particularly when compared to cigarette smoking12–14. Accordingly, there have been calls for more in-depth research to understand the most effective tobacco policy responses to counter this15.\n\nLittle research has been done to uncover the features of the waterpipe tobacco industry, such as that undertaken on the global alcohol industry16,17. This makes designing effective interventions and policies to counter this growing trend challenging. A growing understanding of the cigarette industry has been important in advancing tobacco control globally, and the same is needed for the waterpipe tobacco industry. Our group previously attended a waterpipe trade exhibition in 2014 and showed that marketing material most commonly described waterpipe as a healthier alternative to cigarettes, with emphasis on its flavours, safety, and quality18. Furthermore, we found that transnational tobacco companies were partnered or affiliated with a number of waterpipe tobacco exhibitors19.\n\nAt the 2014 trade exhibition visit we also demonstrated an overlap between the electronic cigarette (e-cigarette) and waterpipe tobacco industry. For example, the majority of exhibitors displayed e-cigarettes of various sizes, rebranded as ‘electronic waterpipes’ (e-waterpipes)18. Based on an analysis of products found on marketing material, we concluded that electronic waterpipe products were offshoots of the e-cigarette industry competing against the waterpipe tobacco industry19. Whether e-waterpipes are the next evolutionary step in the waterpipe tobacco story remains uncertain.\n\nThe objective of this study was to describe the features of the waterpipe tobacco industry. This included an understanding of the distribution of waterpipe tobacco manufacturers, distributers and retailers, identifying unique selling points to products, exploring the concept of brand loyalty and understanding the challenges faced by the waterpipe tobacco industry.\n\n\nMethods\n\nWe conducted semi-structured interviews20 with representatives of waterpipe tobacco companies participating in the third International Hookah Fair. The fair took place on 2nd and 3rd March 2015 in Frankfurt, Germany. The fair organizers described it as “the only trade fair primarily specializing in waterpipes, electronic shishas, hookah tobacco, charcoal and its requisites”21.\n\nWe used our previously developed waterpipe product categorisation scheme18 to decide on eligibility. We included representatives of companies that sell waterpipe consumption products (tobacco or tobacco substitutes) or waterpipe accessories (e.g., apparatuses, charcoal). We excluded exhibitors displaying only e-waterpipe products, as we aimed to focus exclusively on the waterpipe tobacco industry.\n\nWe created a sampling frame by visiting each exhibition stand and judging whether it met inclusion criteria. We marked eligible exhibition stands on an exhibition map and then measured their surface area as shown on the exhibition map. We then aimed to randomly sample a minimum of 30% of eligible exhibitors using probability proportional-to-size sampling, to ensure the representation of exhibitions with varying placement, footfall, and product type.\n\nThree authors (AD/MJ/TL) collected data by approaching potential participants at their exhibition stands, introducing themselves, and stating our objective as conducting a research project on the features of the waterpipe tobacco industry. We adapted the interview guide from a similar study conducted among smokeless tobacco industry players in India22. Supplementary File 1 shows the questions included in the interview guide.\n\nWe chose not to audio-record the interviews for practical reasons (e.g. the constant loud music would result in poor quality recordings), and did not have ethical approval to do so. Instead, the data collectors audio recorded their recollection of the discussions outside the main exhibition hall within a few minutes of completion of each interview. All three researchers contributed to this discussion, ensuring the recollection was well-triangulated.\n\nWe transcribed all audio-recordings from the semi-structured interviews. We used a qualitative approach to identify emerging themes using an inductive approach, drawing on grounded theory23. We applied the following three coding procedures24, but not necessarily in the following order: open coding (whereby we looked for themes and subthemes and categorized responses based upon differences and similarities between responses); axial coding (checking for gaps and overlaps between the subthemes to ensure that each one was fully elaborated); and selective coding (whereby all subthemes were compared and unified around core themes).\n\n\nResults\n\nThirty-three exhibitors met the inclusion criteria. Despite time constraints, we managed to randomly sample 20 exhibitors (61% of all eligible exhibitors). The duration of the interviews varied widely depending on the level of engagement by the exhibitor. Most interviews lasted less than ten minutes, and the longest lasted over an hour.\n\nThe following key industry features emerged from the thematic analysis, which we detail below:\n\n1. Industry growth\n\n2. Cross-industry overlap\n\n3. Customer-product relationship\n\n4. Attitude towards policy\n\nIndustry growth was demonstrated by the increased globalisation of the waterpipe industry. The analysis suggested that individual components of the waterpipe (e.g. apparatus, tobacco, and charcoal) may be sourced from different countries based on regional material and industrial strengths. The industry itself appears to comprise a host of regional networks and relies on transnational links. Tobacco producers appear to be based largely in low and middle-income countries (e.g., Egypt, India), whilst the apparatus is often made in the Far East (e.g., China) because of lower production costs. A minority of the higher-end waterpipes were made in Eastern Europe because of their expertise with Bohemian crystal and glassware. The most popular form of charcoals in the West are either bamboo-based (usually made in Russia) or coconut shell-based (manufactured in South East Asia). Finally, the distributers and retailers of all products were predominantly based in the Middle East (e.g., in Lebanon, Turkey).\n\nGrowth was also evident in terms of waterpipe companies reaching out to new markets, both in terms of new regions and new demographic groups. Several waterpipe tobacco distributors mentioned that the European market for waterpipe tobacco has already witnessed growth, but is now stabilizing and potentially reaching saturation. Some interviewees stated that the waterpipe tobacco industry is now encroaching into previously unexplored markets, for instance New Zealand, Russia and Mexico. New pricing strategies would target younger adults with less disposable income, and the manufacturing quality of waterpipes would be lower to accommodate this. That said, we understood that many of the company representatives were family members of the owners, suggesting that small, family-led businesses are still commonplace and the emergence of an oligopoly (i.e. the industry is dominated by a handful of companies) has yet to occur.\n\nThe growth of the industry was also demonstrated by presence at the convention of international counterfeit products, which was described by two exhibitors. One exhibitor, the owner of a large waterpipe apparatus-producing firm, explained that a main motivation for being at the exhibition was to identify counterfeits of his product. He angrily pointed out an exhibitor displaying waterpipe apparatuses with a similar name, which he alleged was an attempt to counterfeit his well-established brand. He went on further to say that when he approached representatives of this exhibitor, they claimed that the product was named after the owner’s daughter, and not after his brand.\n\nFurther support of industry growth was seen in the development of new and innovative products. We saw at least three types of charcoal products (briquettes, quick lighting discs, and bamboo/coconut shell-based) in addition to electronic heating elements replacing the charcoal altogether. We witnessed hundreds of tobacco flavours, including a move away from flavour descriptors (e.g. ‘ecstasy’ flavour, ‘twist’ flavour, and ‘green’ flavour) and a growing number of exhibitors displaying tobacco substitutes such as flavoured steam stones and herbal, non-tobacco varieties. We noticed that most of the innovation was by the apparatus manufacturers. In one example, a manufacturer was selling a waterpipe apparatus that had an aquarium with fish incorporated into its base, giving the illusion that the smoke was passing through the aquarium. Several exhibitors were selling ‘diffusers’ – small devices placed on the descending stem of the apparatus which creates smaller bubbles as the smoke enters the water.\n\nInterviews suggested a number of cross-industry overlaps, linked directly to the different product types of the waterpipe industry. For example, one of the largest and most well-positioned stands at the exhibition (immediately in front of the main doors), displayed the logos of Al-Nakhla (a leading tobacco manufacturer) and Japan Tobacco International (JTI) on their banner (Al-Nakhla was purchased by JTI in 2012)19. Another company representative revealed that it is now commonplace for his waterpipe company to exhibit at general cigarette and tobacco trade exhibitions; the Dortmund Intertabac exhibition in Germany was directly mentioned, and other exhibitors mentioned exhibitions in France, England, and Poland.\n\nWe found no evidence that ties to the cigarette industry were present for other waterpipe manufacturers, such as charcoal and apparatus manufacturers. Rather, these manufacturers were connected with non-tobacco industries; rather than waterpipe companies reaching out to other industries, it was generally felt that it was non-tobacco industries reaching out to the waterpipe industry. For example, one company, a successful barbeque charcoal manufacturer for nearly 100 years, have now become a main player in the waterpipe industry. The owner of a Germany-based waterpipe apparatus manufacturer described how his family have been involved in glass making for generations – in the last eight months he moved to making waterpipe apparatuses after his son started using it.\n\nSeveral Exhibitors displaying waterpipe apparatuses explained that the engineering and design of the apparatus were their unique selling points, particularly for more expensive, high-end apparatuses which may appeal to those who see it as a source of pride. One apparatus manufacturer described how his high-end, crystal-based apparatuses were bought by several high profile celebrities. Another were selling their bohemian glass apparatuses for between 149 and 249 Euros each, which was about ten times more expensive than their standard range.\n\nEnsuring high quality for customers was a consistent theme across exhibitors displaying waterpipe products of all types. At least two exhibitors displaying waterpipe apparatuses boasted about how their parts were made of rust-free stainless steel, brass or even more expensive and long-lasting materials. One representative claimed that, in the last ten years, only four of their pipes had rusted, and only because they were scratched or damaged. In another example, exhibitors displaying coconut-based charcoals proudly explained how their products remain hotter for longer compared with traditional briquettes, reducing the ‘inconvenience’ of continuously needing to get up and change the charcoal when it cools.\n\nCustomer loyalty also emerged in several interviews. A more in-depth interview with one exhibitor revealed that customers would routinely 'shop around' trialling many different products before deciding which combination they like. A few of the more established companies at the exhibition described brand loyalty resulting from the reputation. The more newly established companies described loyalty towards their particular flavours or charcoal types, rather than loyalty to their company brand per se. In general, we were given the impression that loyalty was stronger towards the apparatus rather than to the tobacco or charcoal, and when looking only at tobacco product loyalty, this was stronger towards the flavour rather than to the company producing the flavour. In one example, warmer apple tobacco flavours are popular in winter, while in the summer months cooler tobacco flavours (e.g., mint) are the bestsellers, according to one exhibitor.\n\nWe asked directly about trade associations and lobby groups, and found these to be absent or severely lacking in the waterpipe tobacco industry. For example, the owner of a major apparatus-producing camping described how his company was one of few he knows of that had its own lawyers on board ready for trade disputes and other legal battles, and that other companies preferred short-term financial gains rather than long term legal protection. He explained that this may be because the industry was originally based in the Arab world, where there is a shorter history of trade unionism, political lobbying and a more laissez-faire attitude towards respecting the law.\n\nWe found two deliberate violations of tobacco policy. In the first example, a representative of a wholesale retailer described how their companies’ products are priced so that they are middle of the range and affordable. When probed for their specific target audience, this was reported as youth as young as ten years old upwards to those in their mid-thirties. In the second example, we found that some waterpipe tobacco manufacturers exploit the self-assembled nature of waterpipe tobacco smoking to deliberately avoid tobacco ingredient laws. One example that came up separately in four interviews was regarding a German law from the 1970s than prohibited more than 5% glycerin in tobacco products. We saw several instances of companies selling glycerin in separate bottles that end-users could mix into their tobacco to improve the flavor. One tobacco manufacturer at the exhibition explained that 20–30% glycerine was needed to keep the quality of the flavour. Another participant admitted to using more than 5% glycerine in the manufacture of his waterpipe tobacco in order to keep the flavour from being too dry, and said he felt pretty lucky that bypassing the authorities was not creating a problem for him. This indicates that enforcement of these laws may also be lacking.\n\n\nDiscussion\n\nThis study reports the features of the waterpipe industry under four key themes: industry growth, cross-industry overlap, customer-product relationship and attitude towards policy. Our understanding is that the waterpipe industry is in an early growth phase, demonstrated by increasing globalisation, reaching out to new markets, the growing presence of counterfeits, and the development of new and innovative products. However, it is still relatively immature, comprising many small companies, often family businesses, who may be specialized in non-tobacco sectors such as glassware and less interested in long term legal protection. A complex web of interactions occurs with neither centralized planning nor cartelized regulation, relying instead on ad hoc personal and professional relationships between partner companies. Further, the diversification of products at this early stage may be considered a threat to product loyalty, which in itself is already quite weak. Perhaps the most pertinent finding is that the “waterpipe industry” is multidimensional and difficult to define. The fact that the waterpipe industry is not a single entity, but rather a conglomeration of actors from both tobacco and non-tobacco industries, will make the development of effective public health tobacco policies challenging.\n\nOur interpretation of cross-industry overlap is that waterpipe tobacco and cigarette industries are not in competition, but rather in collaboration. These two smoking behaviours are not seen to be mutually undermining, but mutually reinforcing. This is a clear pattern historically seen with transnational tobacco companies producing smokeless tobacco products in addition to cigarette. Further, our interviews suggested that the waterpipe industry derives much of its legitimacy and endurance from its links with non-tobacco industries. The relaxed view, and sometimes deliberate breaches, towards policy is not unexpected for a tobacco industry; however the fragmentation of the industry across many small companies may make enforcement of policy challenging and resource-intense.\n\nLargely thanks to the presence of internal industry documents, we know much about the features of the cigarette industry. The cigarette industry has used its economic power, lobbying and marketing machinery, and manipulation of the media to discredit scientific research and influence governments. It has also injected large contributions into social programs worldwide to create a positive public image under the guise of corporate social responsibility. The waterpipe industry appears to mirror and replicates some, but not all, of these tactics. Our findings, in concordance with the literature, show that the most two salient tactics used by the waterpipe industry are deceptive marketing messages, mainly targeted towards youth12,18,25 and blatant disregard for nearly all tobacco policies13,26–29. However, we are yet to see evidence of lobbying and involvement in social programs, perhaps due to the lack of economic power within the industry. One of the authors (MJ) noted a weak attendance from the waterpipe industry at the 2016 public meeting hosted for the US Food and Drug Administration, where only one company representative from the US spoke briefly against the move towards stronger waterpipe tobacco policy. Furthermore, the waterpipe industry differs from tobacco in several key respects: the large number of small family businesses, the use of reusable apparatuses with user identification and product loyalty centred at least as much on this apparatus as on the tobacco, and the lack of a legal self-protection; how these features impact the waterpipe industry’s ability to emulate the known tactics of the cigarette industry, remains to be seen.\n\nA key strength of this research is that it is the first, to the best of our knowledge, to use qualitative methodology to explore the features of the global waterpipe industry. In the absence of internal industry documents, deciding to sample company representatives at the world's largest waterpipe tobacco trade exhibition is an informative first step in developing a greater understanding of this industry. Our findings, taken together with what we know about waterpipe tobacco and the cigarette industry, offer important insights into the development of the industry and potential foci for further research.\n\nThis study has several limitations. Those attending the trade exhibition in Germany may not represent the global waterpipe tobacco industry; rather we suspect our sample was more over-representative of German waterpipe companies given its location. While waterpipe tobacco use is highest in Middle Eastern countries, that this fair was targeted to Europeans is an important finding. It is possible that this fair was more representative of waterpipe companies that place importance on such marketing events rather than those who do not; this may also make our results non-generalizable. Not knowing which companies will be attending the event a priori limits our capacity to get background information and assess the importance of these companies before meeting their representatives. Our interviews were not recorded on tape, and we instead relied on researchers’ recall and interpretations of discussions to retain the key points made. While this could introduce recall bias, we made all efforts to record the information within minutes of the interview to maximize recall, and every interview was attended by three researchers, minimizing the possibility of bias.\n\nUnderstanding the modus operandi of the waterpipe industry can help design effective interventions and policies to counteract the increasingly widespread use of these products and its potential implications for public health. Given the vast number of small businesses in the sector, now is an ideal window of opportunity to strengthen public health policy towards the waterpipe tobacco industry. However, given the waterpipe industry derives much of its legitimacy and endurance from its links with non-tobacco industries, interventions aimed solely at tobacco are at risk of failing. Cigarette regulations will likely not be effective at controlling waterpipe tobacco use, since they are aimed at targeting large, established companies that mostly use traditional means of advertising to promote the purchase of their products from supermarkets and other regulated vendors. Further, considering waterpipe-specific charcoal manufacturers commonly market their products as ‘healthy’ or ‘healthier’ than cigarettes18 despite their highly toxic emissions30,31, we re-iterate previous calls to treat charcoal products designed for waterpipe tobacco as a proxy tobacco product. We therefore echo calls for a specific legislative waterpipe framework to be developed that accounts for these unique aspects of the industry32, and a call for licensing of commercial waterpipe-serving premises in a similar fashion to the alcohol industry13.\n\nWe have identified several important research implications. A myriad of tobacco and tobacco-free products were marketed and sold side-by-side at this fair19, indicating the need to assess whether non-tobacco products are a gateway to future waterpipe or cigarette tobacco. We tried to look indirectly into the workings of the waterpipe tobacco industry, but more work needs to be done to confirm our findings. We call for additional qualitative research to gain ethnographic information on waterpipe tobacco users, sellers and manufacturers similar to the insights into the workings of the cigarette industry33. Pressing questions include the need to identify the main players in the waterpipe industry, their market shares, and their influence on the supply and demand chains, if any. Given the pivotal role of Global South countries in the production and distribution of waterpipe tobacco products, the specific impact of waterpipe tobacco consumption in the West on developing countries is also a question that warrants asking34,35. A close collaboration between social scientists and public health researchers is needed to fully understand the political economy of the waterpipe tobacco industry.\n\n\nConclusion\n\nTo conclude, this qualitative study has provided insights into the waterpipe tobacco industry structure and features. Policy-makers could benefit from these findings when designing waterpipe tobacco-specific interventions, to curb the rise of waterpipe tobacco-related disease.\n\n\nEthical statement\n\nThe Imperial College Research Ethics Committee approved the study (Reference: ICREC_14_3_6). Written informed consent was not obtained for participation in the study as it was designed as covert participant observation.\n\n\nData availability\n\nDataset 1: Transcripts that formed the basis for this study. DOI, 10.5256/f1000research.13796.d19293736", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis study received no specific funding. Dr. Hawkins' research was partially supported by the National Cancer Institute of the National Institutes of Health under award number R01CA091021. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary material\n\nSupplementary File 1: Questions that guided the semi-structured interviews.\n\nClick here to access the data.\n\n\nReferences\n\nJawad M, Lee JT, Millett C: Waterpipe tobacco smoking prevalence and correlates in 25 Eastern Mediterranean and Eastern European countries: cross-sectional analysis of the Global Youth Tobacco Survey. Nicotine Tob Res. 2016; 18(4): 395–402. PubMed Abstract \| Publisher Full Text\n\nMaziak W, Taleb ZB, Bahelah R, et al.: The global epidemiology of waterpipe smoking. Tob Control. 2015; 24 Suppl 1: i3–i12. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCenters for Disease Control and Prevention: Drop in cigarette consumption offset by increases in other forms of smoked tobacco. 2012. In: 2014. Reference Source\n\nAkl EA, Gunukula SK, Aleem S, et al.: The prevalence of waterpipe tobacco smoking among the general and specific populations: a systematic review. BMC public health. 2011; 11: 244. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSaade G, Warren CW, Jones NR, et al.: Linking Global Youth Tobacco Survey (GYTS) Data to the WHO Framework Convention on Tobacco Control (FCTC): The Case for Lebanon. Prev Med. 2008; 47 Suppl 1: S15–S19. PubMed Abstract \| Publisher Full Text\n\nChan WC, Leatherdale ST, Burkhalter R, et al.: Bidi and hookah use among Canadian youth: an examination of data from the 2006 Canadian Youth Smoking Survey. J Adolesc Health. 2011; 49(1): 102–104. PubMed Abstract \| Publisher Full Text\n\nMinaker LM, Shuh A, Burkhalter RJ, et al.: Hookah use prevalence, predictors, and perceptions among Canadian youth: findings from the 2012/2013 Youth Smoking Survey. Cancer Causes Control. 2015; 26(6): 831–838. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nArrazola RA, Singh T, Corey CG, et al.: Tobacco Use Among Middle and High School Students - United States, 2011–2014. MMWR Morb Mortal Wkly Rep. 2015; 64(14): 381–385. PubMed Abstract\n\nShihadeh A, Schubert J, Klaiany J, et al.: Toxicant content, physical properties and biological activity of waterpipe tobacco smoke and its tobacco-free alternatives. Tob control. 2015; 24 Suppl 1: i22–i30. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPrimack BA, Carroll MV, Weiss PM, et al.: Systematic Review and Meta-Analysis of Inhaled Toxicants from Waterpipe and Cigarette Smoking. Public Health Rep. 2016; 131(1): 76–85. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWaziry K, Jawad M, Ballout RA, et al.: The effects of waterpipe tobacco smoking on health outcomes: an updated systematic review and meta-analysis. Int J Epidemiol. 2017; 46(1): 32–43. PubMed Abstract \| Publisher Full Text\n\nGriffiths M, Harmon TR, Gilly MC: Hubble Bubble Trouble: The Need for Education About and Regulation of Hookah Smoking. J Public Policy Mark. 2011; 30(1): 119–132. Publisher Full Text\n\nJawad M: Legislation enforcement of the waterpipe tobacco industry: a qualitative analysis of the London experience. Nicotine Tob Res. 2014; 16(7): 1000–1008. PubMed Abstract \| Publisher Full Text\n\nPrimack BA, Hopkins M, Hallett C, et al.: US health policy related to hookah tobacco smoking. Am J Public Health. 2012; 102(9): e47–51. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWorld Health Organization: Advisory note: waterpipe tobacco smoking: health effects, research needs and recommended actions for regulators - 2nd edition [online]. [Date of access 20 December 2016]. Reference Source\n\nHolden C, Hawkins B, McCambridge J: Cleavages and co-operation in the UK alcohol industry: A qualitative study. BMC Public Health. 2012; 12(1): 483. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJerginen DH: The global alcohol industry: an overview. Addiction. 2009; 104 Suppl 1: 6–12. PubMed Abstract \| Publisher Full Text\n\nJawad M, Nakkash RT, Hawkins B, et al.: Waterpipe industry products and marketing strategies: analysis of an industry trade exhibition. Tob Control. 2015; 24(e4): e275–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJawad M, Nakkash RT, Hawkins B, et al.: Gaining Insights Into the Waterpipe Tobacco Industry: Participant Observation and a Cross-Sectional Survey of Products at a Trade Exhibition. Nicotine Tob Res. 2016; 18(5): 874–878. PubMed Abstract \| Publisher Full Text\n\nBrinkmann S: Qualitative interviewing. Oxford University Pressl; 2013. Reference Source\n\nVaporfair. [online]. 2015; [Date of access 09 January 2016]. Reference Source\n\nSiddiqi K, Scammell K, Huque R, et al.: Smokeless tobacco supply chain in South Asia: A comparative analysis using the WHO Framework Convention on Tobacco Control. Nicotine Tob Res. 2016; 18(4): 424–30. PubMed Abstract \| Publisher Full Text\n\nMartin PY, Turner BA: Grounded theory and organizational research. J Appl Behav Sci. 1986; 22(2): 141–157. Publisher Full Text\n\nStrauss A, Corbin J: Basics of Qualitative Research. Newbury Park: Sage. 1990.\n\nKhalil J, Heath RL, Nakkash RT, et al.: The tobacco health nexus? Health messages in narghile advertisements. Tob control. 2009; 18(5): 420–421. PubMed Abstract \| Publisher Full Text\n\nJawad M, El Kadi L, Mugharbil S, et al.: Waterpipe tobacco smoking legislation and policy enactment: a global analysis. Tob control. 2015; 24 Suppl 1: i60–i65. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJawad M, McIver C: Waterpipe tobacco smoking prevalence and illegal underage use in waterpipe-serving premises: a cross-sectional analysis among schoolchildren in Stoke-on-Trent. Public health. 2017; 146: 32–38. PubMed Abstract \| Publisher Full Text\n\nNakkash R, Khalil J: Health warning labelling practices on narghile (shisha, hookah) waterpipe tobacco products and related accessories. Tob Control. 2010; 19(3): 235–239. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPrimack BA, Rice KR, Shensa A, et al.: U.S. hookah tobacco smoking establishments advertised on the internet. Am J Prev Med. 2012; 42(2): 150–156. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMonzer B, Sepetdjian E, Saliba N, et al.: Charcoal emissions as a source of CO and carcinogenic PAH in mainstream narghile waterpipe smoke. Food Chem Toxicol. 2008; 46(9): 2991–2995. PubMed Abstract \| Publisher Full Text\n\nSepetdjian E, Saliba N, Shihadeh A: Carcinogenic PAH in waterpipe charcoal products. Food Chem Toxicol. 2010; 48(11): 3242–3245. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSalloum RG, Asfar T, Maziak W: Toward a Regulatory Framework for the Waterpipe. Am J Public Health. 2016; 106(10): 1773–1777. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBenson P: Tobacco capitalism: growers, migrant workers, and the changing face of a global industry. Princeton University Press; 2011. Reference Source\n\nKawane H: The influence of the U.S. tobacco industry in foreign markets. N Engl J Med. 1991; 325(111): 815–816. PubMed Abstract \| Publisher Full Text\n\nBarry M: The influence of the U.S. tobacco industry on the health, economy, and environment of developing countries. N Engl J Med. 1991; 324(13): 917–920. PubMed Abstract \| Publisher Full Text\n\nSingh N, Jawad M, Darzi A, et al.: Dataset 1 in: Features of the waterpipe tobacco industry: A qualitative study of the third International Hookah Fair. F1000Research. 2018. Data Source" }	[ { "id": "33009", "date": "24 Apr 2018", "name": "Raed Bahelah", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an interesting article and well-written. I have some comments that I hope the authors find helpful:\nIntroduction\nThe context is missing in this sentence in the first paragraph: \"The proportion of adolescents who have ever smoked waterpipe tobacco was between 4 and 33% in Africa and Asia, between 6 and 11% in the USA, and up to 38% in some UK communities\", what is the year(s) for these estimates?\n\nIn the next sentence: \"Waterpipe tobacco use amongst young youth in Jordan increased from 14.0% to 22.6% between 2008 and 2011, with similar patterns seen in Lebanon, Canada, and the US\", do you mean ever or current waterpipe use?\n\nIn the second paragraph, I think the authors should add that waterpipe is perceived as less harmful and less addictive compared to cigarettes.\n\nIn the second paragraph, the authors need to highlight the differences between waterpipe and cigarettes as the tobacco industry techniques are well known and can be the same for waterpipe. However, one point to elaborate on, as well, is the involvement of the cigarette industry if the authors are aware of any literature, in the waterpipe business. This will support their discussion under cross-industry overlap.\n\nThird paragraph, while the authors did a good job highlighting the tactics of the cigarette industry and that similar understanding is needed for the waterpipe, I did not enjoy the comparison to the alcohol industry. It is, to me, out of context here.\nMethods\nMy main concern is that the sampling method may not necessarily capture exhibitors with \"varying product type\". Can the authors address this issue and justify their methodology?\n\nUnder \"Data Collection\", please make it clear that each interview was attended by all 3 researchers.\n\nI am also concerned that by introducing themselves as researchers and what the study is all about, this may influence participants' responses and may intentionally lie about connection with a cigarette industry. How did the authors reduce this possible source of bias?\n\nThe fact that the authors audio-recorded their own discussions after each interview can by itself introduce bias. Other than recall bias, how did the authors attempted to avoid conclusion bias while discussing the interviews?\n\nFollowing my previous question, how did the authors resolve any disagreement while transcribing the audio-recorded interviews?\nResults\nSome of the arguments for \"Industry Growth\" seem to me not really supporting growth rather globalization. For example, the proliferation of distributors and retailers in lower income countries can be due to a cheaper labor and because the waterpipe epidemic originated from the Middle East, rather than a growth. It can be also due to lax regulatory policies that allow such a proliferation. Can the authors comment on that?\n\nUnder \"4. Attitude towards policy\", please correct a typo: \"For example, the owner of a major apparatus-producing camping..\", should be \"company\" not \"camping\".\n\nUnder \"4. Attitude towards policy\", please correct the typo: \"One example that came up.......German law from the 1970s than..\", should be \"that\" not \"than\".\nDiscussion\nThe second paragraph: \"Our interpretation of cross-industry overlap is that waterpipe tobacco and cigarette industries are not in competition, but rather in collaboration.\", I could not see any results based on this study to support this conclusion.\nTranscripts\nI assume the authors reported the transcripts for 18 out of the 20 interviews. While some interviews show detailed discussions among the 3 authors, others show only \"Researcher 1\"?\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Partly\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4068", "date": "23 Oct 2018", "name": "Mohammed Jawad", "role": "Author Response", "response": "We would like thank the reviewer for their constructive comments, which we have used to improve the quality of the manuscript. The following changes have been made:Comment 1: The context is missing in this sentence in the first paragraph: \"The proportion of adolescents who have ever smoked waterpipe tobacco was between 4 and 33% in Africa and Asia, between 6 and 11% in the USA, and up to 38% in some UK communities\", what is the year(s) for these estimates?Our reply: Thank you for this comment, which was also raised by the other reviewer. We have added some context to this sentence and at the same time updated it a more recent systematic review of prevalence (the one originally cited was conducted in 2008). This sentence now reads:\"A systematic review conducted in 2016 showed that the proportion of adolescents who tried waterpipe tobacco in the last 30 days averaged at 10.6% for Europe, 10.3% for the Eastern Mediterranean, and 6.8% for the Americas.\"Comment 2: In the next sentence: \"Waterpipe tobacco use amongst young youth in Jordan increased from 14.0% to 22.6% between 2008 and 2011, with similar patterns seen in Lebanon, Canada, and the US\", do you mean ever or current waterpipe use?Our reply: Thank you for this comment. This refers to current (past-30 day) use, as outlined in McKelvey et al. 2013 (we pooled girls and boys in this calculation). We noted the reference for this trends was missing for the manuscript so we have added this in.Comment 3: In the second paragraph, I think the authors should add that waterpipe is perceived as less harmful and less addictive compared to cigarettes.Our reply: Thank you for this suggestion, which have now included in the first line of the manuscript and have cited a recent narrative review by Akl et al. to support it.Comment 4: In the second paragraph, the authors need to highlight the differences between waterpipe and cigarettes as the tobacco industry techniques are well known and can be the same for waterpipe. However, one point to elaborate on, as well, is the involvement of the cigarette industry if the authors are aware of any literature, in the waterpipe business. This will support their discussion under cross-industry overlap.Our reply: Thank you for this comment. The only involvement of the cigarette industry in the waterpipe industry that we are aware of is that of JTI's purchase of Nakhla tobacco, which is the predominant waterpipe tobacco brand in Egypt. This was also confirmed in our study. We are uncertain what the reviewer means by the \"differences between waterpipe and cigarettes\" and would welcome further clarification.Comment 5: Third paragraph, while the authors did a good job highlighting the tactics of the cigarette industry and that similar understanding is needed for the waterpipe, I did not enjoy the comparison to the alcohol industry. It is, to me, out of context here.Our reply: Thank you for this comment. We have removed reference to the alcohol industry.Comment 6: My main concern is that the sampling method may not necessarily capture exhibitors with \"varying product type\". Can the authors address this issue and justify their methodology?Our reply: Thank you for this question. We used a random sampling approach that factored in the size of the exhibition stand. We believe that the random component to the sampling was likely to capture exhibitions with varying product types.Comment 7: Under \"Data Collection\", please make it clear that each interview was attended by all 3 researchers.Our reply: Thank you for this comment, which was also raised by the other reviewer. We have made this clear by adding the following sentence to our methods:\"Researchers took turns to lead each interview but all participated by asking questions and making comments in a conversational manner.\"Comment 8: I am also concerned that by introducing themselves as researchers and what the study is all about, this may influence participants' responses and may intentionally lie about connection with a cigarette industry. How did the authors reduce this possible source of bias?Our reply: Thank you for this question. While appreciating that there is no way to be certain that our responses were completely truthful, the nature of the event was such that networking and conversations were encouraged. We did not treat our discussions as \"interviews\" and ensured they were kept conversational, casual and informal, while at the same time working through our interview guide. We have added the following to our limitations section:\"Finally, introducing ourselves as researchers may have influenced participants' responses. However, the nature of the event was such that networking and conversations were encouraged, and we did not expect more \"truthful\" answers had we not introduced ourselves as researchers, given we were unacquainted with our participants anyway.\" Comment 9: The fact that the authors audio-recorded their own discussions after each interview can by itself introduce bias. Other than recall bias, how did the authors attempted to avoid conclusion bias while discussing the interviews?Our reply: Thank you for this question. Given how little is known about the waterpipe tobacco industry, none of the researchers had any major pre-existing beliefs about the conduct of the industry prior to these discussions. There was therefore a small possibility of conclusion (or confirmation) bias.Comment 10: Following my previous question, how did the authors resolve any disagreement while transcribing the audio-recorded interviews?Our reply: Thank you for this question. The transcribing and analysis were conducted by one of the team members who did not attend the exhibition. This allowed for an independent analysis of the audio content. The analyst annotated the transcripts and all three researchers responded to the annotations to clarify ambiguous statements and resolve disagreements. There were very few disagreements between the reviewers, if not none.Comment 11: Some of the arguments for \"Industry Growth\" seem to me not really supporting growth rather globalization. For example, the proliferation of distributors and retailers in lower income countries can be due to a cheaper labor and because the waterpipe epidemic originated from the Middle East, rather than a growth. It can be also due to lax regulatory policies that allow such a proliferation. Can the authors comment on that?Our reply: Thank you for this interesting point. On re-reading the transcripts we probably meant globalisation rather than growth. In fact, we could not assess actual industry growth beyond suggestive indicators, such as the fact there was an international trade fair in Germany, that there were signs of globalisation, or that products on display were quite diverse. But these could happen without industry growth. Based on this comment we have rephrased this section to refer more to globalisation than to growth.Comment 12: Under \"4. Attitude towards policy\", please correct a typo: \"For example, the owner of a major apparatus-producing camping..\", should be \"company\" not \"camping\".Our reply: Thank you for spotting this typo, it has been corrected.Comment 13: Under \"4. Attitude towards policy\", please correct the typo: \"One example that came up.......German law from the 1970s than..\", should be \"that\" not \"than\".Our reply: Thank you for spotting this typo, it has been corrected.Comment 14: The second paragraph: \"Our interpretation of cross-industry overlap is that waterpipe tobacco and cigarette industries are not in competition, but rather in collaboration.\", I could not see any results based on this study to support this conclusion.Our reply: Thank you for this comment, which was also raised by the other reviewer. Having re-read the manuscript, we agree with it, and have rephrased this to:\"Our interpretation of cross-industry overlap, with particular reference to JTI's purchase of Al-Nakhla, is that waterpipe tobacco and cigarette industries are interacting with one another.\"Comment 15: I assume the authors reported the transcripts for 18 out of the 20 interviews. While some interviews show detailed discussions among the 3 authors, others show only \"Researcher 1\"?Our reply: All researchers contributed to the discussions, but when discussions were either relatively short or straightforward, one researcher led the summary and other two only interjected in the case of disagreement or omission of content." } ] }, { "id": "33008", "date": "11 May 2018", "name": "Randah Ribhi Hamadeh", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a very well written original and interesting article. However, I have some comments and queries: Abstract: Results: “20 representatives”, I suggest moving “20” to methods before “representatives” Introduction:\nPara 1: Specify time for this statement“ The proportion of adolescents who have ever smoked waterpipe tobacco was between 4 and 33% in Africa and Asia, between 6 and 11% in the USA, and up to 38% in some UK communities” or, if it is the latest available data, replace “was” by “is”. Para 2: “about 4.1mg nicotine, 619.0mg of tar, and 192mg of carbon monoxide” remove “of” Para 3:“This makes designing effective interventions and policies to counter this growing trend challenging. A growing”, I suggest replacing the first “growing” by “expanding” Para 3: I suggest moving the 2014 exhibition information to the next para as it is discussed as well.\nMethods:\nWritten consent was not taken from the participants and this should be stated. Since the audio recording was done following the interview, the authors should address recall bias and whether consensus between the three data collectors was necessary. Three researchers were present at each interview but did they all participate? How did you ensure privacy if others were present at the exhibition stand ?\nResults:\nPage 4, last para” “Rather, these manufacturers were connected with non-tobacco industries; rather than waterpipe companies reaching out to other industries..” rephrase by removing one of the two “rather” Page 5, Attitude towards policy, line 4: “camping”? Correct to “ company”\nDiscussion\nPage 5, last para: \"Our interpretation of cross-industry overlap is that waterpipe tobacco and cigarette industries are not in competition, but rather in collaboration.\", The conclusion is not supported with the results. Page 6, para 1: “Furthermore, the waterpipe industry differs from tobacco in several key respects”, shouldn’t tobacco be cigarette industry?\nTranscripts Authors need to explain why the transcripts vary in detail and the number of researchers reporting.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nNot applicable\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4067", "date": "23 Oct 2018", "name": "Mohammed Jawad", "role": "Author Response", "response": "We would like thank the reviewer for their constructive comments, which we have used to improve the quality of the manuscript. The following changes have been made:Comment 1: Abstract: Results: “20 representatives”, I suggest moving “20” to methods before “representatives”Our reply: Thank you for this suggestion, we have moved the number of participants to the methods section of the abstract.Comment 2: Para 1: Specify time for this statement“ The proportion of adolescents who have ever smoked waterpipe tobacco was between 4 and 33% in Africa and Asia, between 6 and 11% in the USA, and up to 38% in some UK Our reply: Thank you for this comment. These were taken from a systematic review which was conducted in 2008, so the prevalence estimates range from 2008 and earlier. We have since updated this review, and instead cite more recent findings with respect to prevalence. We have included the time frame in order to address the reviewer's comment. This sentence now reads:\"A systematic review conducted in 2016 showed that the proportion of adolescents who tried waterpipe tobacco in the last 30 days averaged at 10.6% for Europe, 10.3% for the Eastern Mediterranean, and 6.8% for the Americas.\"Comment 3: Para 2: “about 4.1mg nicotine, 619.0mg of tar, and 192mg of carbon monoxide” remove “of”Our reply: We have removed \"of\" from this sentence.Comment 4: Para 3:“This makes designing effective interventions and policies to counter this growing trend challenging. A growing”, I suggest replacing the first “growing” by “expanding”Our reply: We have used the word \"expanding\" as suggestedComment 5: Para 3: I suggest moving the 2014 exhibition information to the next para as it is discussed as well.Our reply: Thank you for this suggestion, we have moved information about the 2014 exhibition to the next paragraph.Comment 6: Written consent was not taken from the participants and this should be stated.Our reply: Thank you for this comment. That written consent was not taken from the participants is stated in the Ethical Statement of the manuscript.Comment 7: Since the audio recording was done following the interview, the authors should address recall bias and whether consensus between the three data collectors was necessary.Our reply: Thank you for this comment. We had addressed recall bias in the limitations section of the discussion, where is it stated: \"While this could introduce recall bias, we made all efforts to record the information within minutes of the interview to maximize recall, and every interview was attended by three researchers, minimizing the possibility of bias.\"Comment 8: Three researchers were present at each interview but did they all participate?Our reply: Yes, all researchers participated by asking questions and actively listening to the interviews. There were no predetermined allocation of roles as to whether only one person would speak, for example, in order to keep the discussions informal and natural. We have added the following sentence to the manuscript to clarify this:\"Researchers took turns to lead each interview but all participated by asking questions and making comments in a conversational manner.\"Comment 9: How did you ensure privacy if others were present at the exhibition stand?Our reply: Thank you for this question. It was not possible to ensure privacy during this research, given it was conducted in the public domain. We added the following sentence to include this point in our methods:\"Privacy with participants was not possible given the public setting.\"Comment 10: Page 4, last para” “Rather, these manufacturers were connected with non-tobacco industries; rather than waterpipe companies reaching out to other industries..” rephrase by removing one of the two “rather”.Our reply: Thank you for this comment, we have replaced the first \"rather\" with \"instead\"Comment 11: Page 5, Attitude towards policy, line 4: “camping”? Correct to “ company”Our reply: Thank you for spotting this typo, it has been corrected to \"company\".Comment 12: Page 5, last para: \"Our interpretation of cross-industry overlap is that waterpipe tobacco and cigarette industries are not in competition, but rather in collaboration.\", The conclusion is not supported with the results.Our reply: Thank you for this comment. Having re-read the manuscript, we agree with it, and have rephrased this to:\"Our interpretation of cross-industry overlap, with particular reference to JTI's purchase of Al-Nakhla, is that waterpipe tobacco and cigarette industries are interacting with one another.\"Comment 13: Page 6, para 1: “Furthermore, the waterpipe industry differs from tobacco in several key respects”, shouldn’t tobacco be cigarette industry?Our reply: You are absolutely correct, we have replaced \"tobacco\" with \"the cigarette industry\"" } ] } ]	1	https://f1000research.com/articles/7-247
https://f1000research.com/articles/7-1499/v1	20 Sep 18	{ "type": "Software Tool Article", "title": "Increasing workflow development speed and reproducibility with Vectools", "authors": [ "Tyler Weirick", "Raphael Müller", "Shizuka Uchida", "Tyler Weirick", "Raphael Müller" ], "abstract": "Despite advances in bioinformatics, custom scripts remain a source of difficulty, slowing workflow development and hampering reproducibility. Here, we introduce Vectools, a command-line tool-suite to reduce reliance on custom scripts and improve reproducibility by offering a wide range of common easy-to-use functions for table and vector manipulation. Vectools also offers a number of vector related functions to speed up workflow development, such as simple machine learning and common statistics functions.", "keywords": [ "bioinformatics", "reproducibility", "workflow", "vector", "matrix", "spreadsheet" ], "content": "Introduction\n\nAlthough the importance of computational analyses in biological research is increasingly appreciated, many analyses are time consuming to implement and remain complicated, as well as being difficult to reproduce1. Workflow-managers [e.g., Snakemake2] have greatly simplified many aspects needed for reproducibility. However, custom scripts (i.e., software not intended for use by a wider audience) remain a problem3. Custom scripts are often needed to further process data generated by high-use programs (i.e., programs intended for a wide user base). At the most basic level, analysis pipelines requiring custom scripts simply take more time to implement as additional code needs to be written. However, writing custom scripts also increases the chance of software bugs, which is concerning as even small bugs have led to retractions, such as mislabeling metadata4 or a sign change5. Furthermore, analyses using custom scripts also hamper reproducibility as the scripts may be publically unavailable, lack documentation, or does not work on certain operation systems. To reduce the impact of these problems, we introduce Vectools6, a command-line tool for working with vectors, matrices, and tables. Vectools reduces the need for custom scripts by offering an easy-to-use command-line tool with a wide range functions for manipulating tables, one of the most commonly used formats in bioinformatics. Further, Vectools incorporates a number of other useful vector-related functions, such as statistics and machine learning. Altogether, Vectools helps to speed up workflow development and improves reproducibility by offering a wide range of useful functions.\n\n\nMethods\n\nVectools can be run via command-line by simply typing “vectools”, which will print the main help menu. Vectools contains over 45 operations organized by headings. These are analysis, descriptors, manipulation, math, normalization, supervised learning, and unsupervised learning. A list of all headings and functions is available in (Supplementary File 1). To run an operation, simply type “vectools” followed by the operation name. If the “—help” argument is added after an operation name, a help menu with usage instructions and examples will be printed.\n\nA standard laptop computer with a recent version of Python3 will handle most applications.\n\n\nUse cases\n\nWhen manipulating data in tables, Core Utilities (Coreutils) programs (e.g., awk, grep, sed, and join) can be used instead of custom scripts. Using Coreutils helps to solve problems with availability as they are common to Unix-based systems. Here, we compared the usage of Vectools to Coreutils. Methods and output can be found in the archived data7. One downside of Coreutils programs is that they can be complex and difficult to understand. For example, joining multiple tables requires a Bash script using Coreutils-join, whereas this can be done with a single line with Vectools (Figure 1A). Furthermore, while common in Unix systems, the behavior of Coreutils programs may differ depending on the operating system. These differences can potentially cause errors or unexpected behavior, such as aggregating Gene Ontology (GO) terms by gene accession numbers with sed (Figure 1B). Instead of aggregating values on MacOS or other Berkeley Software Distribution (BSD) Unix systems, the Coreutils function prints the original input data. These errors can be caused by multiple reasons, such as BSD-sed not interpreting ANSI-C escape sequences (e.g., \\n for newline, \\t for tab) and differences in how regular expressions are evaluated. These problems can be overcome with Vectools with only one line of command. Vectools offers many functions that are currently unavailable in Coreutils, such as basic machine learning. Here, we show a simple example of using a support-vector machine to find potential novel carbonic anhydrases independent of sequence homology (Figure 1C). Carbonic anhydrases were chosen as they have multiple distinct classes, which arose via convergent evolution8. Vectools significantly simplifies a number of steps needed for this task. For example, the “svmtrain” operation handles hyper-parameter tuning via grid search, k-fold testing, and independent set testing. This significantly simplifies implementing machine learning in analysis pipelines.\n\n(A) Joining more than two files requires a single command using Vectools. The same operation using Coreutils requires a custom script. (B) Aggregating Gene Ontology terms by gene accession numbers with Vectools can be done with a simple command. The same operation using Coreutils requires a complex regular expression. Further, the regular expression does not work properly on MacOS. (C) Vectools also includes many operations unavailable in Coreutils, such as machine learning. Here, in five commands, we use supervised-learning for homology-independent prediction of enzyme function. Using Vectools we generated a support-vector machine model capable of predicting carbonic anhydrases with an estimated 99% accuracy and predict 15,018 of 1,223,287 uncharacterized proteins as potential carbonic anhydrases. Methods and output can be found in the archived data and analysis pipelines7.\n\n\nDiscussion\n\nHere, we show that Vectools reduces the need for custom scripts and is simpler to use than Coreutils. While Coreutils is faster and uses less memory, this is a minor issue given the increasing power and decreasing cost of computational resources. Vectools also offers various other functionalities, such as allowing easy incorporation of machine learning into analysis pipelines. Furthermore, Vectools helps to increase reproducibility by making analysis pipelines easier to share and reducing bugs. Users may also be interested in comparison with R. While certainly suited to the same tasks: 1) integrating R into a pipeline requires custom scripts; and 2) the use-cases for R and Vectools are different. R offers a large variety of functions at the cost of package dependency issues. Conversely, Vectools emphasizes ease-of-use by hosting a curated list of common functions. Thus, one common use-case of Vectools when combined with a workflow-manager is to replace work done in spreadsheets. This use-case offers a number of benefits. For example, it is in line with a recent technology feature in Nature, which argues that the concept of reproducibility extends to creating easy-to-update analysis pipelines9. With Vectools, these easy-to-update pipelines will also be easy to share, making valuable tool for bioinformatics research.\n\n\nData availability\n\nAll data used in the paper are archived in Zenodo7.\n\n\nSoftware availability\n\nSource code available from: https://vectools.bitbucket.io/.\n\nData and analysis pipelines: http://doi.org/10.5281/zenodo.14136667.\n\nSource code at time of publication: http://doi.org/10.5281/zenodo.14136716.\n\nLicense: The software, and data and analysis pipelines are available under a Creative Commons Attribution 4.0 International (CC BY 4.0) license.", "appendix": "Grant information\n\nThis study was supported by the start-up funding from the Mansbach Family, the Gheens Foundation and other generous supporters at the University of Louisville; University of Louisville 21st Century University Initiative on Big Data in Medicine (Z1762); and the Deutsche Forschungsgemeinschaft (SFB834 Z4).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary material\n\nSupplementary File 1. A list of operations offered by Vectools with short descriptions of their functions.\n\nClick here to access the data\n\n\nReferences\n\nFehr J, Heiland J, Himpe C, et al.: Best practices for replicability, reproducibility and reusability of computer-based experiments exemplified by model reduction software. AIMS Mathematics. 2016; 1(3): 261–281. Publisher Full Text\n\nKöster J, Rahmann S: Snakemake--a scalable bioinformatics workflow engine. Bioinformatics. 2012; 28(19): 2520–2522. PubMed Abstract \| Publisher Full Text\n\nLeVeque RJ: Top ten reasons to not share your code (and why you should anyway). SIAM News. 2013; 1. Reference Source\n\nHenson KE, Jagsi R, Cutter D, et al.: Retraction. J Clin Oncol. 2016; 34(27): 3358–3359. PubMed Abstract \| Publisher Full Text\n\nMa C, Chang G: Structure of the multidrug resistance efflux transporter EmrE from Escherichia coli. Proc Natl Acad Sci U S A. 2007; 104(9): 3668. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWeirick T, Müller R, Uchida S: Vectools source code at time of publication. Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1413671\n\nWeirick T, Müller R, Uchida S: Data and analysis pipelines used in Increasing workflow development speed and reproducibility with Vectools [Data set]. Zenodo. 2018. http://www.doi.org/110.5281/zenodo.1413666\n\nHewett-Emmett D, Tashian RE: Functional diversity, conservation, and convergence in the evolution of the alpha-, beta-, and gamma-carbonic anhydrase gene families. Mol Phylogenet Evol. 1996; 5(1): 50–77. PubMed Abstract \| Publisher Full Text\n\nPerkel JM: A toolkit for data transparency takes shape. Nature. 2018; 560(7719): 513–515. PubMed Abstract \| Publisher Full Text" }	[ { "id": "38645", "date": "26 Sep 2018", "name": "Yutaka Saito", "expertise": [ "Reviewer Expertise bioinformatics" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article describes Vectools, a command-line tool that can do various kinds of matrix operations for tsv-like data with simple one-liner programs. Vectools is similar to sed and awk commands in Unix Coreutils but has more functionalities, thereby reducing the cost for implementing custom scripts for daily data analyses. The authors claim this will improve the reproducibility problem in recent bioinformatics studies.\nAs a general comment, I think Vectools is useful and will be of interest for bioinformaticians who work in practical data analyses. Although I do not feel the tool has a theoretical novelty, its practical usefulness is worth post-publication evaluation by future users.\nI have several comments as follows:\n1:\nVectools is also similar to \"groupby\" function in Bedtools.\n\nSome functionalities of Bedtools groupby are not included in Vectools, and vice versa. The authors should refer to Bedtools, and if any, other command-line tools similar to Vectools.\n\n2:\nFor each analysis in Figure 1, please provide the size of input data (#rows, #columns, #sequences, etc.). Especially, I get the impression that SVM consumes a large memory. Although I partly agree with the authors' statement that the computational cost is a minor issue, it is still important to provide the information of memory usage along with data size.\n\n3 (minor points):\n(Top left in page 2) However --> In addition (?) (Top right in page 3) valuable tool --> valuable tools\n\nIs the rationale for developing the new software tool clearly explained? Partly\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4071", "date": "19 Oct 2018", "name": "Shizuka Uchida", "role": "Author Response", "response": "We would like to thank the reviewer for valuable comments. The followings are our point-by-point responses: > Comment #1: I have several comments as follows: Vectools is also similar to \"groupby\" function in Bedtools. Some functionalities of Bedtools groupby are not included in Vectools, and vice versa. The authors should refer to Bedtools, and if any, other command-line tools similar to Vectools. > Our response: Thank you very much for raising this point. We now clearly cite Bedtools in the Discussion section. To address the functionality issue, we have implemented two additional operations in Vectools, which are: 1) “mode” for calculating mode/antimode in Vectools; and 2) “colmerge” for combining or splitting columns based on a delimiter. We have also added the “--group” option to relevant operations (e.g., mean, mode, stdev). For cases in which the operation names or functionality do not match exactly, we list the equivalences between Bedtools Groupby and Vectools below: Groupby - Vectools count– shape \| slice collapse– aggregate distinct– unique count_distinct– unique \| sum sstdev– vrep \| stdev freqasc/ freqdesc– unique \| slice \| colmerge \| aggregate first/ last– chop > Comment #2: For each analysis in Figure 1, please provide the size of input data (#rows, #columns, #sequences, etc.). Especially, I get the impression that SVM consumes a large memory. Although I partly agree with the authors' statement that the computational cost is a minor issue, it is still important to provide the information of memory usage along with data size. > Our response: We have updated the figure by adding the file sizes for the SVM example. The first two examples display the entire file. Thus, we did not add file sizes in those examples. We have updated the figure legend to make this clearer. Further, all data used is assessable in the archived data. We have also updated the figure legend to make this more apparent. Finally, we fixed two typos in the figure. > Comment #3 (minor points): (Top left in page 2) However --> In addition (?) (Top right in page 3) valuable tool --> valuable tools > Our response: Thank you very much for reading our manuscript carefully. We have corrected the above grammatical errors as well as others." } ] } ]	1	https://f1000research.com/articles/7-1499
https://f1000research.com/articles/7-1687/v1	23 Oct 18	{ "type": "Software Tool Article", "title": "DataViz: visualization of high-dimensional data in virtual reality", "authors": [ "Eric Feng", "Xijin Ge", "Eric Feng" ], "abstract": "Virtual reality (VR) simulations promote interactivity and immersion, and provide an opportunity that may help researchers gain insights from complex datasets. To explore the utility and potential of VR in graphically rendering large datasets, we have developed an application for immersive, 3-dimensional (3D) scatter plots. Developed using the Unity development environment, DataViz enables the visualization of high-dimensional data with the HTC Vive, a relatively inexpensive and modern virtual reality headset available to the general public. DataViz has the following features: (1) principal component analysis (PCA) of the dataset; (2) graphical rendering of said dataset’s 3D projection onto its first three principal components; and (3) intuitive controls and instructions for using the application. As a use case, we applied DataViz to visualize a single-cell RNA-Seq dataset. DataViz can help gain insights from complex datasets by enabling interaction with high-dimensional data.", "keywords": [ "Virtual Reality", "Principal Component Analysis", "Visualization", "High-dimensional", "Unity" ], "content": "Introduction\n\nHistorically, we have heavily relied on 2-dimensional (2D) graphical displays to communicate large amounts of data. These graphs have also been useful in finding patterns within datasets and building intuition for more accurate and meaningful analysis. However, for large and complex datasets containing numerous dimensions, traditional 2D charts and graphs are inadequate in demonstrating the multi-faceted nature of relevant information.\n\nThe 3-dimensional (3D) visualization of datasets are valuable because they offer a starting solution to the problem above; the addition of another dimension allows for more information to be presented and thus decreases the potential for misinterpretation while concurrently increasing the possibility of pattern-matching and building intuition.\n\nThis paper researches the potential of using virtual reality (VR) as a platform to graphically render datasets in 3D by creating a visualization application. VR is already being used in a variety of fields including flight simulations1, mental health therapy2, and even visualizations of molecules and their interactions3. In the specific field of data visualization, several applications exist, including a surround-screen, projection-based visualizer named CAVE4, one developed using OpenGL that visualizes economic data5, and iViz6, an efficient and intuitive visualizer using VR that is also the most similar to the application developed in this research. DataViz attempts to make further progress by creating a modern, intuitive, and readily available application.\n\nWe continue to explore the potential of VR in the graphical rendering of large datasets; to do so, we have developed a Unity3D VR application for HTC Vive (HTC, New Taipei City, Taiwan) that runs principal component analysis (PCA) on datasets before graphing the subsequent projection into three dimensions. The software was designed to run efficiently with an intuitive interface.\n\n\nMethods\n\nIn the design of this application, special consideration was given to the following elements: the method of data analysis, the format of the input data, the limitations in computing power of the selected platform, and the mitigation of motion sickness.\n\nThe primary method of data analysis is PCA. The rationale behind this decision is that because humans live in three dimensions, the most intuitive manner of visualization is one that plots in that space. In this sense, PCA is excellent at taking large dimensional data and reducing them to plottable 3D coordinates, making the resulting graph more intuitive, and helping users discover patterns and develop scientific intuition.\n\nDataViz only accepts data in the table format (CSV or TXT). Occasionally, the user would want to analyze the transpose of the provided data. Although the transpose of a table could easily be found using specialized functions in Numpy or R, we decided to add the transpose functionality into the application.\n\nIn addition to transposition, DataViz also allows the user to omit specific columns from the file. This may be due to a variety of reasons including an unwanted dimension of data or column names. This functionality allows researchers to analyze only the columns they are interested in.\n\nThe user may also have a column that labels the points. Users can designate a specific column that differentiates the data with various tags, and these groups will show up in a graph legend during runtime.\n\nThe engine used in developing this application is Unity®. Unity is one of the most popular platforms for VR development but is not specifically designed for statistical analysis. Therefore, PCA on large datasets may result in slow run times, especially when there is a lack of an appropriate graphics card or other computational power involved. To overcome this limitation, the application can also accept coordinate data derived from PCA or other dimensionality reduction methods such as t-SNE7. In this manner, users can circumvent the slower computations associated with Unity.\n\nWhen implementing the VR aspect of the application, we concentrated on two main considerations: immersion and motion sickness. For the former, the primary goal was to allow the user to focus on the graphical rendering of his/her data without being bothered by the complicated details on how to use the tool. In pursuit of this, we designed an intuitive interface and series of menus, with clear instructions on the associated GitHub page in ‘Software Availability’.\n\nAnother concern when designing for VR was motion sickness. Motion sickness is a consequence of conflicting input between visual and inner ear senses and is a major problem in current VR simulations8. It has been found that motion sickness is a consequence of the action of motion and not displacement itself, and as a result, we designed our movement to be in short bursts of teleportation.\n\nThe application is built using the Unity® engine with scripting done in C#. The PCA and transpose implementation is from the Accord.Net 3.8 framework (http://accord-framework.net). The mouse embryonic development data used in the case study is from Ref 9.\n\nDataViz was designed to be an intuitive application for graphically rendering large datasets. Upon opening the software, a user should follow the onscreen prompts and fill out the appropriate parameters to input their dataset as well as use the extra functionalities described above. DataViz automatically runs PCA on the input dataset according to user configurations. If needed, more detailed instructions can be found on the associated GitHub page.\n\nVR is a resource intensive activity. The following are guidelines for ensuring the quality and performance of DataViz.\n\nSystem Requirements (https://www.vive.com/us/ready/):\n\nProcessor: Intel i5-4590 / AMD FX 8350 equivalent or greater\n\nGraphics card: NVIDIA GeForce GTX 1060 or AMD Radeon RX 480, equivalent or better\n\nMemory: 4 GB RAM or more\n\nVideo output: HDMI 1.4 or DisplayPort 1.2 or newer\n\nUSB: 1x USB 2.0 or newer\n\nOperating system: Windows 7 SP1 or newer\n\n\nUse case\n\nThe primary goal in the development of this application was to determine the viability of using VR to graphical render and analyze complex data sets. After development, we tested the DataViz by analyzing a high-dimensional dataset regarding mouse embryonic development9. Using single-cell RNA sequencing (scRNA-Seq), Deng et al. generated hundreds of expression profiles of individual embryonic cells from zygote blastocyst stages.\n\nBy graphically rendering the 3D PCA projection of the data and subsequent analysis, we were able to verify an expected trend of embryo development; initial cell division (zygote stage to 16-cell stage) results in large-scale physical changes inside the embryo. This is in contrast to later cell division where the various stages of embryo development are more similar to one another. We can also see the developmental trajectory in the transcriptomic landscape (Figure 1).\n\nThe graph displays the similarities among the blastocyst stages in comparison to changes in earlier stages of development. We can identify categories and general trends of the data using this method.\n\nThis method of analysis has some limitations, the foremost being an inability to account for all the data present. While reducing high-dimensional data to three dimensions simplifies the resulting plot and may help formulate testable hypotheses through further research or build intuition and comprehension regarding the data provided, it is inevitable that we lose some of the variance present in higher dimensions. In this test case, Table 1 reveals the proportion of the data retained per principal component. One way of overcoming this would be to use non-linear dimensionality reduction methods like multidimensional scaling (MDS) or t-SNE.\n\nFor example, in the test case of mouse embryo development, the resulting three-dimensional graph could only reveal 33% of the original dataset.\n\nPC, principal component.\n\nDespite the shortcomings involved in the provided analysis and plotting approach, DataViz is still useful for categorizing the data into disjoint groups.\n\n\nDiscussion\n\nTwo of the primary motivations for using VR to visualize data were the introduction of a third dimension as well as increased interactivity with data. As shown by the Use Case, although the current functionality is limited to PCA, the application is useful in demonstrating the potential that VR has to offer in the analysis and communication of large, complex datasets.\n\nTo understand this potential further, future research should focus on human trials in determining the statistical difference between the traditional 3D plot on a computer screen and a VR simulation regarding data comprehension and analysis. Additionally, in order to account for more variance in the original dataset, future research should consider other dimensionality reduction methods.\n\n\nConclusion\n\nWe have developed an application for visualizing high-dimensional data in VR. It reduces high-dimensional data using PCA before generating an immersive 3D scatter plot. It also contains a variety of functionalities including the ability to transpose the given input and to accept raw coordinate data. A major limitation of DataViz is its inability to account for the full variance in the dataset. Also, the amount of benefit that visualization receives from being in VR as opposed to on a 2D monitor is unknown.\n\n\nData availability\n\nThe data of mouse embryo development can be found in Deng et al., 20149\n\n\nSoftware availability\n\nSource code and additional instructions available at: https://github.com/thunder2011/DataViz\n\nArchived source code at time of publication: https://doi.org/10.5281/zenodo.145578710.\n\nLicense: GNU Lesser General Public License v2.1", "appendix": "Grant information\n\nThis material is based upon work supported by the National Science Foundation/EPSCoR Grant Number IIA – 1355423 and by the State of South Dakota. Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the view of the National Science Foundation.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nChittaro L, Buttussi F: Assessing Knowledge Retention of an Immersive Serious Game vs. a Traditional Education Method in Aviation Safety. IEEE Trans Vis Comput Graph. 2015; 21(4): 529–38. PubMed Abstract \| Publisher Full Text\n\nPowers MB, Rothbaum BO: Recent advances in virtual reality therapy for anxiety and related disorders: Introduction to the special issue. J Anxiety Disord. 2018; pii: S0887-6185(18)30342-6. PubMed Abstract \| Publisher Full Text\n\nO'Connor M, Deeks HM, Dawn E, et al.: Sampling molecular conformations and dynamics in a multiuser virtual reality framework. Sci Adv. 2018; 4(6): eaat2731. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCruz-Neira C, Sandin D, DeFanti T: Surround-Screen Projection-Based Virtual Reality: The Design and Implementation of the CAVE. Proceedings of the 20th Annual Conference on Computer Graphics and Interactive Techniques - SIGGRAPH '93, Anaheim, CA USA. 1993; 135–142. Publisher Full Text\n\nSullivan P: Graph-Based Data Visualization In Virtual Reality: A Comparison Of User Experiences. California Polytechnic State University. 2016. Publisher Full Text\n\nDonalek C, Djorgovski SG, Cioc A, et al.: Immersive and Collaborative Data Visualization Using Virtual Reality Platforms. 2014 Ieee International Conference on Big Data (Big Data). 2014; 609–614. Publisher Full Text\n\nvan der Maaten L: Accelerating t-SNE using Tree-Based Algorithms. Journal of Machine Learning Research. 2014; 15: 3221–3245. Reference Source\n\nBecker J, Ngo T: Mitigating Visually-Induced Motion Sickness in Virtual Reality. Stanford University. 2016. Reference Source\n\nDeng Q, Ramsköld D, Reinius B, et al.: Single-cell RNA-seq reveals dynamic, random monoallelic gene expression in mammalian cells. Science. 2014; 343(6167): 193–6. PubMed Abstract \| Publisher Full Text\n\nFeng E: thunder2011/DataViz: Centralize Package and Executible onto Github (Version v1.0.2). Zenodo. 2018. http://dx.doi.org/10.5281/zenodo.1455787" }	[ { "id": "40657", "date": "18 Dec 2018", "name": "David R. Glowacki", "expertise": [], "suggestion": "Not Approved", "report": "Not Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn my opinion the work outlined in this paper represents an interesting software prototype, but at this stage my impression is that this work is still very much in the ‘prototype’ phase and not yet ready for full publication. Developing a VR framework for visualizing PCA in 3d is extremely quick work using a tool like Unity, and does not represent a significant technical achievement per se. This paper feels to us like an interesting starting point for future research. For example, it would be good to see some examples of dataset visualisations which demonstrate cases where the VR really helped the end-user, e.g., in the form of measurable HCI type user studies, or perhaps through case study examples. We have inspected the code linked to in the GIT repository, and it appears to rely on standard Unity sphere prefabs. It would be good to understand how this framework would actually scale for visualizing massive data sets.\n\nTechnically, it would be quite useful if the application could outsource the computation of the PCA data to another program – e.g., via a library or through a communication protocol such as protobufs. That would maximize the application’s interactivity, and it would mean that users need not precompute their PCA.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1687
https://f1000research.com/articles/6-1688/v1	14 Sep 17	{ "type": "Research Article", "title": "Interfacial microscopic examination and chemical analysis of resin-dentin interface of self-adhering flowable resin composite", "authors": [ "Tamer M. Hamdy" ], "abstract": "Background: The newly introduced self-adhering flowable resin-composites decrease the required time for application by incorporation of an acidic adhesive monomer, thus reducing the number of the steps, but its bonding is still uncertain. The aim of this study was to evaluate the interfacial microscopic examination and chemical analysis at the resin-dentin interface of a self-adhering flowable resin composite (Vertise-Flow) versus a total-etch (Te-Econom Plus) resin composite, using an etching agent (Eco-Etch gel) and bonding agent (Single Bond Universal). Methods: Sixteen freshly extracted sound human posterior teeth were used. The teeth were randomly divided into two groups: 8 specimens per type of composite. Standard-shaped class V cavities were prepared on the buccal surface. One group was restored by Te-Econom Plus resin composite by total-etch technique using Eco-Etch gel, which was applied to dentine for 15 seconds, followed by rinsing, drying and bonding agent application (Single Bond Universal). The other group restored directly with self-adhering resin composite (Vertise-Flow) without application of etch or bond. Curing was done for 20 seconds using a light emitting diode light curing unit. Evaluation of the resin-dentin interface was done microscopically by examination of marginal gap distance in μm using scanning electron microscope (SEM), and chemical analysis of silver particles was observed using SEM with energy-dispersive X-ray spectrometry after 24 hours of specimen storage in ammoniacal silver nitrate. Results: Regarding marginal gap distance (µm) and silver atomic % mean values, teeth restored with self-adhering resin composite (Vertise-Flow) showed significantly higher mean values than the multi-step etch and rinse resin composite group (5.2 vs 0; 12.2 vs 8.2, respectively). Conclusions: Resin-dentin bonding using total-etch resin composite technique was more effective than self-adhering flowable resin composite (Vertise-Flow) regarding marginal gap formation and penetration of silver particles. Further studies for bond strength could be performed.", "keywords": [ "Self-adhering", "total-etch", "bonding system", "resin composite", "gap distance", "resin-dentin interface" ], "content": "Introduction\n\nAdhesive dentistry has seen a paradigm shift from the invasive to be minimally invasive, due to a revolution in bonding systems. There are great demands for simplified restorative materials. A new self-adhering flowable resin composite (Vertise™ Flow Self-Adhering Flowable Composite, Kerr Dental, USA), was recently introduced onto the market. Bonding is achieved by incorporation of an acidic adhesive monomer into the flowable composites1. It is still a big challenge to seal the resin-dentin interface, due to the heterogeneous nature of dentin, the wet tubular structure, composition and surface morphology and or improperly designed adhesives2,3.\n\nThe total-etch (etch and rinse) technique is a widely accepted technique to improve bonding of dental resins2. The dentin bonding mechanism is based on the micro-mechanical interlocking of the infiltrated resin monomers into porosities created by demineralized inorganic part of dentin4. Restoration debonding may arise from gap formation at the resin-dentin interface and hence recurrent caries, discoloration and tooth pain5. Thus sufficient marginal seal should be obtained. Recently, an innovative self-adhesive and flowable resin composite was developed. These materials are claimed to eliminate the need for a separate bonding application step, thus simplifying the restorative procedure. Thus, the aim of this study was to evaluate the sealing performance of this new material.\n\n\nMethods\n\nAfter attaining written informed consent from each patient to use their extracted teeth in research, sixteen sound human posterior molar teeth were extracted in a private dental clinic (Dr. Tamer Hamdy Dental Clinic), which were randomly divided into two groups (eight specimens per group). Standard-shaped class V cavities (3x3 mm, 2 mm of depth) were prepared in the teeth using a #169L carbide bur (KG Sorensen, Brazil) on the buccal surface. One group’s (Group A) cavities were filled with Te-Econom Plus® (Ivoclar Vivadent, Africa) resin composite after etching and bonding application. The etching agent, Eco-Etch gel (Ivoclar Vivadent), was applied to dentine for 15 seconds, followed by rinsing and drying. A bonding agent (Single Bond Universal, 3M ESPE, USA) was applied to the teeth for 20 seconds, then air-dried for 5 seconds, then light-cured for 10 seconds. Finally, the Te-Econom Plus resin composite was applied. The other group’s (Group B) cavities were filled with self-adhering resin composite (Vertise-Flow), which was applied without etch or bond. Curing was done for 20 seconds using a light emitting diode (LED) light curing unit (Satelec, Acteon, France).\n\nAll teeth were stored in distilled water for 24 hours at 37°C. Subsequently, the specimens were vertically sectioned with a diamond saw (Isomet, Buehler Ltd., USA) under water lubrication into approximately 1mm thick slabs. These were examined for marginal gap distance in μm using scanning electron microscope (SEM; Model Quanta 250 FEG; FEI, Thermo Fisher Scientific, USA): accelerating voltage 30 K.V., magnification 14x up to 1000000 and resolution for Gun.1n, to ensure high brightness and resolution at low accelerating voltage.\n\nSpecimens slab were then placed in freshly prepared 50 weight/volume % ammoniacal silver nitrate solution for an additional 24 hours at 37°C in the dark. Ammoniacal silver nitrate solution (pH=9.5) was prepared according to the protocol of Tay et al. (2002)6. After 24 hours of storage in ammoniacal silver nitrate, the silver impregnated specimens were then rinsed thoroughly in distilled water and placed in photo-developing solution for 8 hours under a fluorescent light (200 Watt)3.\n\nThe specimens were then observed under environmental SEM Model Quanta 250 FEG attached with energy-dispersive X-ray (EDX; Inspect S 50, FEI, Netherlands): accelerating voltage 30 K.V., magnification 4000x and resolution for Gun.1n. Backscattered electron mode was used for elemental analysis of the atomic silver %.\n\nNumerical data were explored for normality using Kolmogorov-Smirnov and Shapiro-Wilk tests, followed by Student’s t-test to compare between the two groups. The significance level was set at P ≤ 0.05. Statistical analysis was performed with IBM®SPSS® Version 20 for Windows (SPSS Inc., IBM Corporation, USA).\n\n\nResults\n\nRegarding marginal gap formation, Group A showed a significantly lower mean gap distance values than Group B (p<0.001), as shown in Table 1 and Figure 1 and Figure 2.\n\nGroup A, treated with total-etch technique; Group B, treated with self-adhering resin composite. Mean and standard deviation (SD) values and results of Student’s t-test for the comparison between gap distances are shown (n=8/group).\n\n: Significant at P ≤ 0.05\n\nImage representative of 8 teeth.\n\nImage representative of 8 teeth.\n\nThe SEM with EDX analysis results revealed that Group A showed significantly lower mean silver atomic % values than Group B (p<0.001), as shown in Table 2 A. Selected SEM/EDX analysis is shown in Figure 3 and Figure 4.\n\nGroup A, treated with total-etch technique; Group B, treated with self-adhering resin composite. Mean and standard deviation (SD) values and results of Student’s t-test for the comparison between silver atomic % values are shown (n=8/group).\n\n: Significant at P ≤ 0.05\n\nImage representative of 8 teeth.\n\nImage representative of 8 teeth.\n\n\nDiscussion\n\nA proper marginal sealing is essential to improve the durability of resin composite/bonding systems7. Most of the clinical studies assessing the performance of an adhesive system use class V cavities8. EDX analysis permits identification of silver particles, thus giving an indication about the chemical analysis of the interface9.\n\nFrom the present results, it was obvious that the multi-step etch and rinse technique provides better sealing regarding marginal gap formation and penetration of silver particles than that of self-etch “Vertise-Flow”. This may be attributed to the fact that phosphoric acid included in acid-etch step demineralized the smear layer, exposing the collagen fibers of superficially demineralized dentin. These could increase micromechanical interlocking of the bonding agent within the dentin surface10. The poorer sealing of Vertise-Flow may be due to the included adhesive monomer, which is called glycerol phosphate dimethacrylate “GPDM”, that etches instead of bonds to hydroxyapatite11.\n\n\nConclusions\n\nTotal-etch resin composite technique was more effective regarding marginal gap formation and penetration of silver particles as compared to a flowable resin composite (Vertise-Flow). Further studies for bond strength could be performed. It is important to emphasize that this study ignored the effect of oral condition, thus further clinical studies are suggested.\n\n\nData availability\n\nDataset 1: Raw values for silver atomic % in teeth treated with total-etch technique (Group A) and self-adhering resin composite (Group B) (n=8/group/method). doi, 10.5256/f1000research.12306.d17706112\n\nDataset 2: Scanning electron microscopy (SEM) showing gap formation (raw values included on the images) and SEM/energy-dispersive X-ray analysis (EDX) at the resin-superficial dentin interface in teeth treated with total-etch technique (Group A) and self-adhering resin composite (Group B) (n=8/group/method). doi, 10.5256/f1000research.12306.d17706213", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nYuan H, Li M, Guo B, et al.: Evaluation of Microtensile Bond Strength and Microleakage of a Self-adhering Flowable Composite. J Adhes Dent. 2015; 17(6): 535–543. PubMed Abstract \| Publisher Full Text\n\nMortazavi V, Fathi M, Ataei E, et al.: Shear bond strengths and morphological evaluation of filled and unfilled adhesive interfaces to enamel and dentine. Int J Dent. 2012; 2012: 858459. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYuan Y, Shimada Y, Ichinose S, et al.: Qualitative analysis of adhesive interface nanoleakage using FE-SEM/EDS. Dent Mater. 2007; 23(5): 561–9. PubMed Abstract \| Publisher Full Text\n\nTurp V, Sen D, Tuncelli B, et al.: Adhesion of 10-MDP containing resin cements to dentin with and without the etch-and-rinse technique. J Adv Prosthodont. 2013; 5(3): 226–33. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKopperud SE, Tveit AB, Gaarden T, et al.: Longevity of posterior dental restorations and reasons for failure. Eur J Oral Sci. 2012; 120(6): 539–548. PubMed Abstract \| Publisher Full Text\n\nTay FR, Pashley DH, Yoshiyama M: Two modes of nanoleakage expression in single-step adhesives. J Dent Res. 2002; 81(7): 472–6, Accessed August 2, 2015. PubMed Abstract \| Publisher Full Text\n\nKanca J 3rd, Greitzer G: Class II restorations with margins below the CEJ. J Esthet Restor Dent. 2009; 21(3): 193–201. PubMed Abstract \| Publisher Full Text\n\nCasselli DS, Faria-e-Silva AL, Casselli H, et al.: Marginal adaptation of class V composite restorations submitted to thermal and mechanical cycling. J Appl Oral Sci. 2013; 21(1): 68–73. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHashimoto M, De Munck J, Ito S, et al.: In vitro effect of nanoleakage expression on resin-dentin bond strengths analyzed by microtensile bond test, SEM/EDX and TEM. Biomaterials. 2004; 25(25): 5565–74. PubMed Abstract \| Publisher Full Text\n\nNeelima L, Sathish ES, Kandaswamy D: Evaluation of microtensile bond strength of total-etch, self-etch, and glass ionomer adhesive to human dentin: an in vitro study. Indian J Dent Res. 2008; 19(2): 129–33. PubMed Abstract \| Publisher Full Text\n\nYoshida Y, Nagakane K, Fukuda R, et al.: Comparative study on adhesive performance of functional monomers. J Dent Res. 2004; 83(6): 454–8. PubMed Abstract \| Publisher Full Text\n\nHamdy TM: Dataset 1 in: Interfacial microscopic examination and chemical analysis of resin-dentin interface of self-adhering flowable resin composite. F1000Research. 2017. Data Source\n\nHamdy TM: Dataset 2 in: Interfacial microscopic examination and chemical analysis of resin-dentin interface of self-adhering flowable resin composite. F1000Research. 2017. Data Source" }	[ { "id": "26488", "date": "06 Oct 2017", "name": "Rania El-Saady Badawy", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nAfter reviewing the entire manuscript, several modifications were added to different parts of the manuscript.\nFor the Introduction part, it shortly covered the idea of investigated resin composites describing the aim of the study.\nMaterials and Methods described steps and method of slab preparation (missing a few details).\nResults were objectively presented.\nThe results were vaguely discussed.\nThis manuscript is approved for publication after elaborating the changes required.\n\nSpecific Comments:\n\nBonding is achieved by incorporation of an acidic adhesive monomer into the flowable composites1.\nComment: Bonding of flowable composites to tooth structure is achieved by incorporation of an acidic adhesive monomer into the material.\n\nIt is still a big challenge to seal the resin-dentin interface, due to the heterogeneous nature of dentin, the wet tubular structure, composition and surface morphology and or improperly designed adhesives2,3\n\nComment: Too long sentence\n\nThe total-etch (etch and rinse) technique is a widely accepted technique to improve bonding of dental resins2.\nComment: improve bonding of dental resins to tooth structure\n\n...is based on the micro-mechanical interlocking of the infiltrated resin monomers into porosities created by demineralized inorganic part of dentin4.\nComment: created in demineralized organic part....\n\nRestoration debonding may arise from gap formation at the resin-dentin interface and hence recurrent caries, discoloration and tooth pain5\nComment: Debonding of restorations may arise from gap-formation ..........., discoloration and tooth pain may follow\n\nThese materials are claimed to eliminate the need for a separate bonding application step, thus simplifying the restorative procedure\nComment: to eliminate the need for a separte step of bond-application, finally simplifying the restorative procedure\n\nThus, the aim of this study was to evaluate the sealing performance of this new material\nComment: Therefore, the aim of.....\n\nPreparation of specimens. After attaining written informed consent from each patient to use their extracted teeth in research\nComment: Why should we get a patients approval to work on extracted teeth?? You should better start with: Sixteen sound human molar teeth.......\n\n...sixteen sound human posterior molar teeth were extracted in a private\n\nComment: Only say human molar teeth- it is known that molars are posterior\n\n(eight specimens per group). Standard-shaped class V cavities (3x3 mm, 2 mm of depth) were prepared\nComment: Dimensions of the cavity are unclear\n\nOne group’s (Group A) cavities were filled with Te-Econom Plus® (Ivoclar Vivadent, Africa) resin composite after etching and bonding application.\nComment: and bond application\n\nA bonding agent (Single Bond Universal, 3M ESPE, USA) was applied to the teeth for 20 Seconds\nComment: After rinsing, a bonding agent ......... was applied to the teeth for 20 sec\n\n...teeth for 20 seconds, then air-dried for 5 seconds, then light-cured for 10 seconds.\n\nComment: was applied to teeth for 20 seconds, afterwards the teeth were air-dried for 5 sec and light-cured for 10 sec\n\n...under water lubrication into approximately 1mm thick slabs.\nComment: Slab of what??? Please explain specific composition the slab. Please mention that it is a slab composed of tooth structure bonded to resin composite of any type!!!\n\nChemical analysis of the interface. Specimens slab were then placed in freshly prepared 50 weight/volume % ammoniacal silver nitrate solution for an additional 24 hours at 37°C in the dark.\nComment: Specimen slabs were placed.......weight or volume percent????\n\nAmmoniacal silver nitrate solution (pH=9.5) was prepared according to the protocol of Tay et al. (2002)6.\n\nComment: prepared according to Tay et al\n\nAfter 24 hours of storage in ammoniacal silver nitrate, the silver impregnated specimens were then rinsed thoroughly in distilled water and placed in photo-developing\n\nComment: storage in silver nitrate solution, the silver impregnated specimens were rinsed. Thoroughly\n\nThe specimens were then observed under environmental SEM Model Quanta 250 FEG attached with energy-dispersive X-ray (EDX; Inspect S 50, FEI, Netherlands): accelerating voltage 30 K.V., magnification 4000x and resolution for Gun.1n.\nComment: he treated specimens were observed.\nBackscattered electron mode was used for elemental analysis of the atomic silver %.\nComment: The Blackscattered ......\n...followed by Student’s t-test to compare between the two groups.\nComment: to compare between both groups\n\nlower mean gap distance values than Group B (p<0.001), as shown in Table 1 and Figure 1 and Figure 2.\nComment: and Figures 1 and 2\n\nThe SEM with EDX analysis results revealed that Group A showed significantly lower mean silver atomic % values than Group B (p<0.001), as shown in Table 2 A. Selected SEM/EDX analysis is shown in Figure 3 and Figure 4.\n\nComment: revealed significantly lower mean silver atomic % values for Group A compared to Group B (p< 0.001) A selected..... in Figures 3 and 4.\n\nTable 1. Marginal gap formation (μm) between groups of teeth treated with different composites\nComment: groups of teeth restored with different...\n\nFigure 1. Selected scanning electron microscopy shows absence of gap formation in dentin in teeth treated with total etch technique.\nComment: in dentin for teeth treated with....\n\nFigure 2. Selected scanning electron microscopy shows presence of gap formation in dentin in teeth treated with self-adhering resin composite.\n\nComment: in dentin for teeth........ with a self-adhering ....\n\nTable 2. Chemical analysis of the interface between groups of teeth treated with different composites. Group A, treated with total-etch technique; Group B, treated with self-adhering resin composite.\nComment: restored with self-adhering resin composite\n\nDataset 1. Raw values for silver atomic % in teeth treated with total-etch technique (Group A) and self-adhering resin composite (Group B) (n=8/group/method)\nComment: Group (A), …\n\nDataset 2. Scanning electron microscopy (SEM) showing gap formation (raw values included on the images) and SEM/energy dispersive X-ray analysis (EDX) at the resin-superficial dentin interface in teeth treated with total-etch technique (Group A) and self-adhering resin composite (Group B) (n=8/group/method)\nComment: Group (A), …\n\nA proper marginal sealing is essential to improve the durability of resin composite/bonding systems7\n\nComment: sealing of what???\n\nFrom the present results, it was obvious that the multi-step etch and rinse technique provides better sealing regarding\n\nComment: Our results revealed better sealing ability of composites treated with multi-step etch and rinse technique, presenting lower marginal gap formation and lower penetration of silver particles compared to Vertise-Flow.\n\nThis may be attributed to the fact that phosphoric acid included in acid-etch step demineralized the smear layer, exposing the collagen fibers of superficially demineralized dentin. These could increase micromechanical interlocking of the bonding agent within the dentin surface10\nComment: Presence of exposed collagen fibers could increase....\n\nThe poorer sealing of Vertise-Flow may be due to the included adhesive monomer, which is called glycerol phosphate dimethacrylate “GPDM”, that etches instead of bonds to hydroxyapatite11.\nComment: included adhesive monomer, the glycerol.........\n\nFurther studies for bond strength could be performed. It is important to emphasize that this study ignored the effect of oral condition,\nComment: Further studies on bond strength should be undertaken.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "3086", "date": "17 Oct 2017", "name": "Tamer Hamdy", "role": "Author Response", "response": "Thanks for your valuable revision. Corrections doneBelong to your commentComment: Why should we get a patients approval to work on extracted teeth?? You should better start with: Sixteen sound human molar teeth......It was done according to journal editor revision due to the rule of the journal." } ] }, { "id": "26333", "date": "11 Oct 2017", "name": "Nadia A. Badr", "expertise": [ "Reviewer Expertise Dental biomaterials" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe scientific idea of this article to evaluate the adaptability of resin/dentine interface of the most commonly used direct aesthetic restorative material is superb.\nThe article lacks the name of the company & country of the main investigated material (flowable composite, Vertise-Flow).\nDespite the used method is well known for detection of nanoleakage of restorative materials, it is smartly employed for assessment of adaptability of composite at resin/dentine interface.\nThe suggestions is logical but it is preferable to be separate item not involved in context of conclusion section.\n\nSummary:\nThe article is presented clearly & accurately The study design is appropriate Sufficient details were provided except the name of the company & country of the investigated material (Vertise-Flow) The used method for detection of marginal gap is smartly employed The data are reproducible And the unique conclusion is obvious The suggestion is promising for further researches\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "3123", "date": "23 Oct 2017", "name": "Tamer Hamdy", "role": "Author Response", "response": "Many thanks for your valuable comments, I do the required corrections, belong to \"the suggestions is logical but it is preferable to be separate item not involved in context of conclusion section\" I do that according to editorial request according to the journal role" } ] } ]	1	https://f1000research.com/articles/6-1688
https://f1000research.com/articles/7-608/v1	17 May 18	{ "type": "Research Article", "title": "Could differences in implicit attitudes to sexual concurrency play a role in generalized HIV epidemics?", "authors": [ "Chris R. Kenyon", "Kenny Wolfs", "Kara Osbak", "Maleeto Malataliana", "Guido Van Hal", "Sizwe Zondo", "Jacques van Lankveld", "Kenny Wolfs", "Kara Osbak", "Maleeto Malataliana", "Guido Van Hal", "Sizwe Zondo", "Jacques van Lankveld" ], "abstract": "Background: Sexual partner concurrency has been implicated in the genesis of generalized HIV epidemic in South Africa. Most South Africans, however, disapprove of concurrency in surveys. These surveys test individuals’ explicit attitudes which are susceptible to a number of important biases such as the social desirability bias. Assessment of implicit cognitions have been found to be better predictors of behaviour in socially sensitive domains. We hypothesized that South Africans may have implicit attitudes more tolerant of concurrency than lower concurrency prevalence populations. Methods: To test this hypothesis, we developed a concurrency-implicit association test (C-IAT) and compared the C-IATs of samples of South African and Belgian university students. Results: We found a large and statistically significant difference in the C-IAT between the South Africans (D600-score = -0.009, indicating absence of preference for concurrency or monogamy) and Belgians (D600-score = 0.783, indicating a strong preference for monogamy; t-test = 13.3; P < 0.0001). The effect size measure, Cohen’s d, was found to be 0.88, which is considered a large effect size in this field. Conclusions: Our results are compatible with the thesis that differences in implicit attitudes to concurrency play a role in the genesis of generalised HIV epidemics.", "keywords": [ "concurrency", "HIV", "sexual networks", "implicit association" ], "content": "Introduction\n\nA higher prevalence of sexual partner concurrency, were an individual has a series of overlapping sexual partners at once, is one of the factors implicated in the genesis of generalized HIV epidemics in Southern and Eastern Africa1–3. Qualitative research from the region has argued that a tolerance of concurrency plays an important role in generating high concurrency rates4–9. A quantitative analysis of South African survey data, however, found that most men and women disapproved of concurrency9. This discrepancy may be partly explained by the way that the social desirability bias may affect the accuracy of self-reported data pertaining to socially sensitive topics such as sexual norms10–14. Measures of implicit cognition assess cognitive processes less available to introspection and are thus less affected by these problems. Several studies have found implicit measures to be better predictors of behavior than explicit measures in these sensitive domains10,13–15. In a previous study, we developed a concurrency implicit association test (C-IAT) and tested it on a sample of 869 Belgian students16. The students revealed a strong implicit preference for monogamy as opposed to concurrency. No differences in C-IAT were found between men and women, but men who have sex with men and women who have sex with women were found to have a somewhat weaker implicit preference for monogamy than heterosexual men and women16.\n\nIn this study, we compare the results from this Belgian study with those obtained from a similar sample of South African students. We assess: (i) if implicit and explicit norms towards concurrency differ between Belgian and South African university students, (ii) if the variation between these two populations involves a difference in behavior of ‘core-groups’ or general population shifts (iii) the correlation between implicit and explicit attitudes to concurrency and reporting that one has engaged in concurrency at both individual and population levels.\n\n\nMethods\n\nImplicit Association Tests (IATs) are reaction-time measures that tap implicit associations without requiring conscious introspection17. We developed a Concurrency-IAT (C-IAT) that measures the implicit associations that individuals hold towards concurrency in relation to monogamy. Our C-IAT was constructed using the attribute categories “positive/negative” and the target categories “monogamy/multiple partners.” Participants had to categorize words as either positive or negative and pictures as either depicting two people in a monogamous relationship or two people of which one had another partner.\n\nOur C-IAT consisted of five different blocks. The C-IAT was programmed in OpenSesame, an open source program for reaction time experiments, for the offline version that was used in English in South Africa18. The online Dutch-language IAT used in in Belgium was hosted on the Project Implicit® Web site. The full C-IATs as well as all the words and images used in their construction can be obtained from Kenyon et al.16\n\nAfter completing the IAT the students were asked to complete a questionnaire pertaining to their sexual behavior and explicit attitudes to concurrency. These questions (variables they are intended to define) included: How many sex partners do you have? (Point prevalence concurrency); Where there any other times in your life when you had more than one sex partner at a time? (Life time concurrency). Three questions investigating explicit attitudes towards concurrency were assessed using a scale from 1 (strongly disagree) to 5 (strongly agree): It’s okay to have sex with others as long as your main partner does not find out? (Concealed concurrency); If you are in a sexual relationship with someone, it’s okay to have sex with others as long as you are honest with your main partner about this? (Liberalist concurrency); If my main partner has other sex partners, it is okay for me to have other partners as well? (Reactive concurrency)19. The questions used in this questionnaire are available from Kenyon et al.16\n\nAll procedures were approved by the Institutional Review Board of the Institute of Tropical Medicine (Antwerp) and the Ethics Committees of the University of Antwerp and Rhodes University.\n\nIn both countries all students at the two participating Universities were eligible for study inclusion. In Belgium they were recruited via an email sent to the entire student body. This was not possible in South Africa and thus students were recruited via posters and word of mouth.\n\nAll participants were tested independently either in the Department of Psychology or in a secure and quiet room at the Rhodes University library. After they had signed the informed consent form, students were first asked to perform the C-IAT behind a computer in the above mentioned locations. After students completed the C-IAT, they proceeded to answer the explicit, paper-and-pencil questionnaire measures.\n\nThe entire protocol was conducted online. Students received a link to the study website via the recruitment e-mail. The first step on the study website was signing the informed consent form. They then completed the C-IAT, and after this the explicit questionnaire.\n\nFor both student populations, the IAT and explicit measures took between 15 and 20 minutes to complete.\n\nD600-scores of the IAT were calculated according to the standard protocol suggested by Greenwald et al.20,21 Scores usually vary between -2 and +2, indicating strong implicit preferences for concurrency and monogamy, respectively, with zero indicating absence of preference. The minimum response time was 400 ms, the maximum response time was 2500 ms. Any responses below this interval were omitted while any responses above this interval were recoded to 2500 ms. Incorrect answers got a penalty of 600 ms.\n\nWe compare the distributions of C-IAT (D600-scores) scores between Belgium and South Africa visually using histograms and statistically using t-tests for independent samples. In keeping with standard practice in this field, we used Cohen’s d as a measure of effect size. Cohen’s d was calculated by dividing the South African minus the Belgian mean difference D600 by the pooled standard deviation. Pearson’s correlation was used to test the correlations between implicit and explicit attitudes as well as between these two and self-reported point-prevalence of concurrency. Chi-squared and t-tests were used to test differences between categorical and continuous variables. All analyses were repeated stratified by gender. There were differences by gender in self-reported concurrency and explicit (but not implicit) attitudes towards concurrency. These differences were, however, congruent between Belgium and South Africa and did not affect the results. As a result, only unstratified results are reported.\n\nPopulation level analyses: Sexual norms and behaviors such as concurrency have been shown to vary between different sexual orientations22–25. This provided the rationale for using Spearman’s correlation to assess the population level correlations between the point-prevalence of concurrency and intrinsic (mean D600) and extrinsic (mean values for each of the 3 variables considered separately) attitudes. The populations were defined according to self-reported sexuality by country and gender. Only populations with n > 10 were utilized for the analyses.\n\nAll analyses were performed in Stata 13 (StataCorp LP, College Station, TX, USA).\n\n\nResults\n\nA total of 869 students in Belgium and 70 in South Africa participated. The demographic characteristics of the populations are detailed in Table 1. The South African students reported more sexual partners in the past year than the Belgians (mean 3.5 and 1.4, respectively, P < 0.001), a higher point-prevalence of concurrency (38.7% and 3.0%, P < 0.001), ever having engaged in concurrency (61.5% and 22.2%, P < 0.001) and partner concurrency (50.8% and 16.9%, P < 0.001; Table 1).\n\n* P < 0.05, P < 0.001, * P < 0.0001 (P-values are for comparisons between South Africa and Belgium).\n\n# WSW – Woman who has sex with women, MSM – Man who has sex with men\n\nThe IAT results for the South African and Belgian populations both approximated normal distributions with similar standard deviations (SD) = 0.40 and 0.42 respectively; Figure 1). There was a large and statistically significant difference in the C-IAT between the South Africans (D600-score = -0.009, indicating absence of preference for concurrency or monogamy) and Belgians (D600-score = 0.783, indicating a strong preference for monogamy; t-test = 13.3; P < 0.0001). The effect size measure, Cohen’s d, was found to be 0.88 which is considered a large effect size in this field10. There was no difference between C-IAT results by gender in either country (data not shown).\n\nDistributions of D600 scores for South African and Belgian students (Positive and negative scores indicate preferences for monogamy and concurrency respectively).\n\nThe differences in implicit associations between countries were larger than those for explicit associations: South Africans were more pro-concealed-concurrency (d = 0.47), Belgians more pro-liberalist-concurrency (d = 0.27) and there was no difference in pro-reactive-concurrency (d = 0.03; Table 1).\n\nIndividual level: Self-reported concurrency behavior was slightly more strongly associated with explicit (r = 0.13 to 0.58) than implicit (r = 0.05 to 0.11) attitudes to concurrency by country (Table 2).\n\n* P < 0.05, P < 0.001, * P < 0.0001\n\n# IAT - Implicit Association Tests\n\nPopulation level: The prevalence of concurrency by sexual orientation was associated with the mean implicit attitude to concurrency (rho = 0.95, P = 0.0004, n = 8). The same relationship was present when the analysis was restricted to the Belgian students (rho = 0.87, P = 0.024, n = 6). The association between extrinsic attitudes and concurrency was not statistically significant (concealed concurrency: rho = 0.65, P = 0.06; liberal concurrency: rho = -0.50, P = 0.171; reactive concurrency: rho = -0.03, P = 0.932).\n\n\nDiscussion\n\nThe IAT results for the study populations in South Africa and Belgium were both normally distributed with a similar variance. Belgium’s population curve was however relatively right-shifted. There was no evidence of a ‘core high risk group’ with a distribution outside of the Gaussian distribution in either country. This variation of distributions between different populations via right or left shifting the mean value (but retaining the same variance) mimics the findings of Rose and others for a wide variety of physical and mental health attributes and behaviors, including number of sex partners25–27. Rose’s interpretation of this relationship was that populations do not tolerate ‘excessive’ variations in norms and behaviors and thus distributions of these characteristics move up and down as a whole27.\n\nSimilarly, large differences in implicit cognition via shifting the mean value whilst retaining the same variance have been found between different groups in other studies12,20,28. One 34 country study, for example, found large differences in mean implicit gender-science stereotype scores between countries28. Of interest these mean IAT scores were found to be better predictors of national sex differences in math and science achievement than corresponding explicit attitudes. Culturally determined differences in implicit attitudes are thought to emerge during childhood or early adolescence as individuals participate in the custom complexes of their cultures11,28,29. Historical and anthropological analyses suggest that the shift from polygamy to monogamy in Southern Africa over the last 150 years did not reduce the number of partners men had. Non-marital and non-main partnerships were however driven underground4,6,8,30. This created the norm which - although heavily contested - maintains that it is acceptable for men to have ‘kwapeni’s’ (secret lovers) as long as their main partner does not find out5–7. In our study, we found that 20.0% of South Africans versus 1.4% of Belgians agreed with this statement (P < 0.001). Because this is a sensitive topic it is possible that the IAT is providing an alternative measure of the acceptance of concurrency.\n\nIf this is the case and this acceptance is causally linked to higher concurrency rates then the distribution of implicit responses to concurrency amongst South African students suggests that population level interventions would be required to address this issue. Current efforts targeting concurrency are largely focused on higher risk individuals which are unlikely to result in a shift in the population distribution in implicit attitudes to concurrency5,27,31.\n\nA limitation with the line of reasoning outlined above is the low correlation found between self-reported concurrency and implicit attitudes to concurrency. This may be because the implicit attitudes play little or no role in driving high concurrency rates. Alternatively, the C-IAT may constitute an important marker of a population-wide greater tolerance of concurrency which broadly enables higher concurrency rates but that other specific risk factors then determine which individuals will engage in concurrency32. We found some support for this latter interpretation in the form of a population level correlation between implicit attitudes towards concurrency and the practice thereof. Further study limitations include: a small sample size in South Africa; only samples from two countries were included in the study; and there were slight differences in how participants were recruited and tested. In South Africa, explicit questionnaires were completed on paper and pencil and the IAT was run offline. In Belgium, both the explicit questionnaire and IAT were offered online and could be completed from home. Respondents were self-selected and thus the samples cannot be regarded as representative of the entire university student or national populations. Finally, in the Belgian sample the nature of the web-based IAT meant that we could not exclude the possibility of multiple participations by respondents.\n\n\nData availability\n\nDataset 1: STable 1: Concurrency implicit association tests 10.5256/f1000research.14951.d20342633", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nKenyon C, Buyze J, Colebunders R: HIV prevalence by race co-varies closely with concurrency and number of sex partners in South Africa. PLoS One. 2013; 8(5): e64080. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGoodreau SM, Cassels S, Kasprzyk D, et al.: Concurrent partnerships, acute infection and HIV epidemic dynamics among young adults in Zimbabwe. AIDS Behav. 2012; 16(2): 312–322. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKirby D: Changes in sexual behaviour leading to the decline in the prevalence of HIV in Uganda: confirmation from multiple sources of evidence. Sex Transm Dis. 2008; 84 Suppl 2: ii35–ii41. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDelius P, Glaser C: The myths of polygamy: a history of extra-marital and multi-partnership sex in South Africa. South African Historical Journal. 2004; 50(1): 84–114. Publisher Full Text\n\nJana M, Nkambule M, Tumbo D: Multiple and concurrent sexual partnerships in Southern Africa: a ten country research report. In. Johannesburg: Soul City; 2008. Reference Source\n\nLeclerc-Madlala S: Cultural scripts for multiple and concurrent partnerships in southern Africa: why HIV prevention needs anthropology. Sex Health. 2009; 6(2): 103–110. PubMed Abstract \| Publisher Full Text\n\nParker W, Makhubele B, Ntlabati P, et al.: Concurrent sexual partnerships amongst young adults in South Africa. Challenges for HIV prevention communication. In: CADRE; 2007. Reference Source\n\nHunter M: Cultural politics and masculinities: multiple-partners in historical perspective in KwaZulu-Natal. Cult Health Sex. 2005; 7(3): 209–223. PubMed Abstract \| Publisher Full Text\n\nKenyon C, Osbak K, Buyze J, et al.: Variations of Sexual Scripts Relating to Concurrency by Race, Class, and Gender in South Africa. J Sex Res. 2015; 52(8): 878–86. PubMed Abstract \| Publisher Full Text\n\nGreenwald AG, Poehlman TA, Uhlmann EL, et al.: Understanding and using the Implicit Association Test: III. Meta-analysis of predictive validity. J Pers Soc Psychol. 2009; 97(1): 17–41. PubMed Abstract \| Publisher Full Text\n\nSteffens MC, Buchner A: Implicit Association Test: separating transsituationally stable and variable components of attitudes toward gay men. Exp Psychol. 2003; 50(1): 33–48. PubMed Abstract \| Publisher Full Text\n\nRudman LA, Greenwald AG, Mellott DS, et al.: Measuring the automatic components of prejudice: Flexibility and generality of the Implicit Association Test. Social Cognition. 1999; 17(4): 437–465. Publisher Full Text\n\nCzopp AM, Monteith MJ, Zimmerman RS, et al.: Implicit attitudes as potential protection from risky sex: Predicting condom use with the IAT. Basic Appl Soc Psych. 2004; 26(2–3): 227–236. Publisher Full Text\n\nMarsh KL, Johnson BT, Scott-Sheldon LA: Heart versus reason in condom use: implicit versus explicit attitudinal predictors of sexual behavior. Z Exp Psychol. 2001; 48(2): 161–175. PubMed Abstract \| Publisher Full Text\n\nNock MK, Banaji MR: Prediction of suicide ideation and attempts among adolescents using a brief performance-based test. J Consult Clin Psychol. 2007; 75(5): 707–715. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKenyon CR, Wolfs K, Osbak K, et al.: Implicit attitudes to sexual partner concurrency vary by sexual orientation but not by gender - a cross sectional study of Belgian students. PLoS One. (In Press). 2018; 13(5): e0196821. PubMed Abstract \| Publisher Full Text\n\nNosek BA: Moderators of the relationship between implicit and explicit evaluation. J Exp Psychol Gen. 2005; 134(4): 565–584. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMathôt S, Schreij D, Theeuwes J: OpenSesame: an open-source, graphical experiment builder for the social sciences. Behav Res Methods. 2012; 44(2): 314–324. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJohnson S, Kincaid L, Figueroa M, et al.: The Third National HIV Communication Survey, 2012. South Africa: Pretoria. 2013. Reference Source\n\nGreenwald AG, McGhee DE, Schwartz JL: Measuring individual differences in implicit cognition: The implicit association test. J Pers Soc Psychol. 1998; 74(6): 1464–1480. PubMed Abstract \| Publisher Full Text\n\nGlashouwer KA, Smulders FT, de Jong PJ, et al.: Measuring automatic associations: validation of algorithms for the Implicit Association Test (IAT) in a laboratory setting. J Behav Ther Exp Psychiatry. 2013; 44(1): 105–113. PubMed Abstract \| Publisher Full Text\n\nGlick SN, Morris M, Foxman B, et al.: A comparison of sexual behavior patterns among men who have sex with men and heterosexual men and women. J Acquir Immune Defic Syndr. 2012; 60(1): 83–90. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHigher Education and Training HIV/AIDS Programme: National Student Sexual Health HIV Knowledge, Attitude and Behaviour Survey: Focusing on Student Men who have Sex with Men at 14 Higher Education Institutions in South Africa. In. Johannesburg: HEAIDS; 2014. Reference Source\n\nGrulich AE, de Visser RO, Smith AM, et al.: Sex in Australia: homosexual experience and recent homosexual encounters. Aust N Z J Public Health. 2003; 27(2): 155–163. PubMed Abstract \| Publisher Full Text\n\nKenyon CR, Tsoumanis A, Schwartz IS: A population's higher-risk sexual behaviour is associated with its average sexual behaviour-An ecological analysis of subpopulations in Ethiopia, Kenya, South Africa, Uganda and the United States. Epidemics. 2016; 15: 56–65. PubMed Abstract \| Publisher Full Text\n\nRose G: Mental Disorder and the Strategies of Prevention. Psychol Med. 1993; 23(3): 553–555. PubMed Abstract \| Publisher Full Text\n\nRose G, Day S: The population mean predicts the number of deviant individuals. BMJ. 1990; 301(6759): 1031–1034. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNosek BA, Smyth FL, Sriram N, et al.: National differences in gender-science stereotypes predict national sex differences in science and math achievement. Proc Natl Acad Sci U S A. 2009; 106(26): 10593–10597. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHaidt J: The emotional dog and its rational tail: A social intuitionist approach to moral judgment. Psychol Rev. 2001; 108(4): 814–834. PubMed Abstract \| Publisher Full Text\n\nKenyon C, Zondo S: Why do some South African ethnic groups have very high HIV rates and others not? Afr J AIDS Res. 2011; 10(1): 51–62. PubMed Abstract \| Publisher Full Text\n\nShisana O, Rehle T, Simbayi LC, et al.: South African National HIV Prevalence, Incidence and Behaviour Survey, 2012. Cape Town: HSRC Press; 2014. Reference Source\n\nMah T: Prevalence and correlates of concurrent sexual partnerships among young people in South Africa. Sex Transm Dis. 2010; 37(2): 105–108. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKenyon C, Wolfs K, Osbak K: Dataset 1 in: Could differences in implicit attitudes to sexual concurrency play a role in generalized HIV epidemics? F1000Research. 2018. Data Source" }	[ { "id": "34132", "date": "06 Jun 2018", "name": "Hsun-Ta Hsu", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIntroduction 1. I would recommend the authors do a final review of the paper to ensure it is free of grammatical errors and typos (e.g., in the first sentence in the introduction, \"were\" should be \"where\"). 2. The authors stated: \"This discrepancy may be partly explained by the way that the social desirability bias may affect the accuracy of self-reported data pertaining to socially sensitive topics such as sexual norms.\" I would like to see more elaboration as it pertains to concurrency. 3. The statement regarding the findings of applying C-IAT on Belgium students is great. Since one of the study purposes is to identify the discrepancies between implicit and explicit norms between Belgium and South African participants, I would like to see the authors add their findings on whether there were discrepancies regarding explicit and implicit concurrency norms in their previous Belgium research. Currently, the authors only provide information regarding implicit norms of concurrency among Belgium students. 4. It would be great if the authors would discuss their rationale for comparing South African and Belgium students. 5. Also, it would be beneficial if the authors could provide more justification for the study aims (e.g., why look at shifts and correlations of the norms). 6. Finally, the rationale of looking at college students should also be provided. Is it because they more likely to be sexually active? More likely to engage in concurrency? Or, more likely to at risks of HIV/STIs? Method 1. Information regarding validity and reliability of the measurements used may be needed. 2. The authors mentioned that they recruited participants who were \"similar\" to participants in the Belgium study for the comparison of implicit and explicit norms. How did the authors determine similarity when recruiting participants in South Africa? Since they are all college students, I am assuming their age might be similar? How about the concentrations/majors? 3. More info regarding sampling may be needed. For example, how many students were approached? What was the refusal rate? 4. The measurements were used to measure implicit and explicit norms. Why would there be incorrect answers? Could the authors provide some examples? Results 1. Some of the numbers stated in the text do not match the tables. For example, the author stated: \"Self-reported concurrency behavior was slightly more strongly associated with explicit (r = 0.13 to 0.58) than implicit (r = 0.05 to 0.11) attitudes to concurrency by country.\" But, in Table 2, it seems that the correlation of explicit and concurrency behavior is 0.08-0.58. The authors might want to review the numbers. Discussion 1. The authors provide great information in that based on the finding, current efforts of targetting high-risk individuals may not be enough. Could the authors provide more information on, specifically, what may be the potential approach to address concurrency, especially targetting implicit attitudes on a population level? 2. In the discussion section, the authors seemed to imply that the findings may apply to population level. However, given the sampling strategy, I am not sure if such argument may be the case. Could the authors provide more evidence or explanation?\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "3750", "date": "18 Oct 2018", "name": "Chris Kenyon", "role": "Author Response", "response": "Dear Dr HsuThank you for your useful suggestions and comments, which we respond to below:Introduction1. I would recommend the authors do a final review of the paper to ensure it is free of grammatical errors and typos (e.g., in the first sentence in the introduction, \"were\" should be \"where”).Reply:This will be done in the next version.2. The authors stated: \"This discrepancy may be partly explained by the way that the social desirability bias may affect the accuracy of self-reported data pertaining to socially sensitive topics such as sexual norms.\" I would like to see more elaboration as it pertains to concurrency. Reply:This will be done in the next version of the paper.3. The statement regarding the findings of applying C-IAT on Belgium students is great. Since one of the study purposes is to identify the discrepancies between implicit and explicit norms between Belgium and South African participants, I would like to see the authors add their findings on whether there were discrepancies regarding explicit and implicit concurrency norms in their previous Belgium research. Currently, the authors only provide information regarding implicit norms of concurrency among Belgium students.Reply: These results are provided in detail at:Kenyon CR, Wolfs K, Osbak K, et al.: Implicit attitudes to sexual partnerconcurrency vary by sexual orientation but not by gender - a cross sectionalstudy of Belgian students. PLoS One. (In Press). 2018; 13(5): e0196821.4. It would be great if the authors would discuss their rationale for comparing South African and Belgium students.Reply:This was based on opportunity sampling related to the institutions/countries where the authors work.5. Also, it would be beneficial if the authors could provide more justification for the study aims (e.g., why look at shifts and correlations of the norms).Reply:This will be done in the next version of the paper.6. Finally, the rationale of looking at college students should also be provided. Is it because they more likely to be sexually active? More likely to engage in concurrency? Or, more likely to at risks of HIV/STIs? Reply: Concurrency and STI rates are particularly high in this age group.Method1. Information regarding validity and reliability of the measurements used may be needed.Reply:As we note in this paper this was assessed in the previous paper we refer to:Kenyon CR, Wolfs K, Osbak K, et al.: Implicit attitudes to sexual partnerconcurrency vary by sexual orientation but not by gender - a cross sectionalstudy of Belgian students. PLoS One. (In Press). 2018; 13(5): e0196821.2. The authors mentioned that they recruited participants who were \"similar\" to participants in the Belgium study for the comparison of implicit and explicit norms. How did the authors determine similarity when recruiting participants in South Africa? Since they are all college students, I am assuming their age might be similar? How about the concentrations/majors? Repy:We did not collect information as to the subjects they were studying.3. More info regarding sampling may be needed. For example, how many students were approached? What was the refusal rate? Reply: We do not have this data. The recruitment method used is detailed in the section:In both countries all students at the two participating Universitieswere eligible for study inclusion. In Belgium they were recruitedvia an email sent to the entire student body. This was not possiblein South Africa and thus students were recruited via posters andword of mouth.4. The measurements were used to measure implicit and explicit norms. Why would there be incorrect answers? Could the authors provide some examples?Reply:The reference to “incorrect answers\" on the IAT test refers to misclassifying an image or word during the IAT test e.g. placing an image of monogamy in the concurrency section.Results1. Some of the numbers stated in the text do not match the tables. For example, the author stated: \"Self-reported concurrency behavior was slightly more strongly associated with explicit (r = 0.13 to 0.58) than implicit (r = 0.05 to 0.11) attitudes to concurrency by country.\" But, in Table 2, it seems that the correlation of explicit and concurrency behavior is 0.08-0.58. The authors might want to review the numbers.Reply:Thank you for pointing out this error which will be corrected in the next version of the paper.Discussion1. The authors provide great information in that based on the finding, current efforts of targetting high-risk individuals may not be enough. Could the authors provide more information on, specifically, what may be the potential approach to address concurrency, especially targetting implicit attitudes on a population level?Reply:This paragraph has now been expanded with the following text to address this question:Current efforts targeting concurrency are largely focused on higher risk individuals which are unlikely to result in a shift in the population distribution in implicit attitudes to concurrency 5, 27, 31 . One approach may be follow the Know Your Network concurrency reduction intervention which was shown to be feasible and acceptable in a rural Kenyan population ( Knopf A, et al. \"This is the medicine:\" A Kenyan community responds to a sexual concurrency reduction intervention. Social Science & Medicine 2014; 108: 175-184).2. In the discussion section, the authors seemed to imply that the findings may apply to population level. However, given the sampling strategy, I am not sure if such argument may be the case. Could the authors provide more evidence or explanation?Reply:This is true. In the new version we have changed this claim to the following weaker claim:We found weak support for this latter interpretation in the form of a population level correlation between implicit attitudes towards concurrency and the practice thereof. This is however based on a small number of samples and this finding should be regarded as tentative." }, { "c_id": "4063", "date": "18 Oct 2018", "name": "Chris Kenyon", "role": "Author Response", "response": "Thank you for your useful suggestions and comments, which we respond to below:Introduction1. I would recommend the authors do a final review of the paper to ensure it is free of grammatical errors and typos (e.g., in the first sentence in the introduction, \"were\" should be \"where”).Reply:This has been done in the new version.2. The authors stated: \"This discrepancy may be partly explained by the way that the social desirability bias may affect the accuracy of self-reported data pertaining to socially sensitive topics such as sexual norms.\" I would like to see more elaboration as it pertains to concurrency. Reply:The following text has been added to make this clearer:Respondents to surveys asking about attitudes to sexual partner concurrency may consider that the interviewer holds negative attitudes towards concurrency. They may therefore bias their reported attitudes towards concurrency towards that of the interviewer.3. The statement regarding the findings of applying C-IAT on Belgium students is great. Since one of the study purposes is to identify the discrepancies between implicit and explicit norms between Belgium and South African participants, I would like to see the authors add their findings on whether there were discrepancies regarding explicit and implicit concurrency norms in their previous Belgium research. Currently, the authors only provide information regarding implicit norms of concurrency among Belgium students.Reply: These results are provided in detail at:Kenyon CR, Wolfs K, Osbak K, et al.: Implicit attitudes to sexual partnerconcurrency vary by sexual orientation but not by gender - a cross sectionalstudy of Belgian students. PLoS One. 2018; 13(5): e0196821.4. It would be great if the authors would discuss their rationale for comparing South African and Belgium students.Reply:This was based on opportunity sampling related to the institutions/countries where the authors work. Various studies have noted concurrency prevalences to be relatively low in Western Europe and relatively high in South Africa 1-3.5. Also, it would be beneficial if the authors could provide more justification for the study aims (e.g., why look at shifts and correlations of the norms).Reply: The following text has been added to the end of the introduction to make this clearer:Our rationale for exploring if the variation between these two populations involves a difference in behavior of ‘core-groups’ or general population shifts is based on the work of Rose and others 25– 27. They argued that if one finds a bimodal distribution in behavior 'A' in population 'B' compared to a normal distribution in a comparison population 'C' then this finding would be compatible with the existence of a core-group with higher risk behaviour in population 'B' being responsible for some of the differences in behavior 'A' between the two groups. The concept of a 'core-group' is well established in the HIV field and typically refers to a subpopulation with a high level of sexual network connectivity (conferred by features such as partner concurrency and rate of partner change) that contributes disproportionately to the spread of HIV in that population . Kenyon CR, Tsoumanis A, Schwartz IS: A population's higher-risk sexual behaviour is associated with its average sexual behaviour-An ecological analysis of subpopulations in Ethiopia, Kenya, South Africa, Uganda and the United States. Epidemics. 2016;15:56–65. 27266849 10.1016/j.epidem.2016.02.002In addition, we have explained the rationale for looking at shifts in C-IAT between Belgium and South Africa in the following text in the discussion:The IAT results for the study populations in South Africa and Belgium were both normally distributed with a similar variance. Belgium’s population curve was however relatively right-shifted. There was no evidence of a ‘core high risk group’ with a distribution outside of the Gaussian distribution in either country. This variation of distributions between different populations via right or left shifting the mean value (but retaining the same variance) mimics the findings of Rose and others for a wide variety of physical and mental health attributes and behaviors, including number of sex partners 25– 27 . Rose’s interpretation of this relationship was that populations do not tolerate ‘excessive’ variations in norms and behaviors and thus distributions of these characteristics move up and down as a whole 27 . As far as assessing the correlation between C-IAT and explicit norms is concerned, this is a standard assessment in this type of IAT research. We have provided this rationale in the following text in the discussion: A limitation with the line of reasoning outlined above is the low correlation found between self-reported concurrency and implicit attitudes to concurrency. This may be because the implicit attitudes play little or no role in driving high concurrency rates. Alternatively, the C-IAT may constitute an important marker of a population-wide greater tolerance of concurrency which broadly enables higher concurrency rates but that other specific risk factors then determine which individuals will engage in concurrency 32 . We found some support for this latter interpretation in the form of a population level correlation between implicit attitudes towards concurrency and the practice thereof.6. Finally, the rationale of looking at college students should also be provided. Is it because they more likely to be sexually active? More likely to engage in concurrency? Or, more likely to at risks of HIV/STIs? Reply: Concurrency and STI rates are particularly high in this age group 3, 4.Method1. Information regarding validity and reliability of the measurements used may be needed.Reply:As we note in the current paper the assessment of validity and reliability were perfomed in the previous paper reporting the development of the C-IAT:Kenyon CR, Wolfs K, Osbak K, et al.: Implicit attitudes to sexual partnerconcurrency vary by sexual orientation but not by gender - a cross sectionalstudy of Belgian students. PLoS One. 2018; 13(5): e0196821.2. The authors mentioned that they recruited participants who were \"similar\" to participants in the Belgium study for the comparison of implicit and explicit norms. How did the authors determine similarity when recruiting participants in South Africa? Since they are all college students, I am assuming their age might be similar? How about the concentrations/majors? Repy:We did not collect information as to the subjects they were studying.3. More info regarding sampling may be needed. For example, how many students were approached? What was the refusal rate? Reply: We do not have this data. The recruitment method used is detailed in the following section:In both countries all students at the two participating Universitieswere eligible for study inclusion. In Belgium they were recruitedvia an email sent to the entire student body. This was not possiblein South Africa and thus students were recruited via posters andword of mouth.4. The measurements were used to measure implicit and explicit norms. Why would there be incorrect answers? Could the authors provide some examples?Reply:The reference to “incorrect answers\" on the IAT test refers to misclassifying an image or word during the IAT test e.g. placing an image of monogamy in the concurrency section. This is standard terminology in the IAT field.Results1. Some of the numbers stated in the text do not match the tables. For example, the author stated: \"Self-reported concurrency behavior was slightly more strongly associated with explicit (r = 0.13 to 0.58) than implicit (r = 0.05 to 0.11) attitudes to concurrency by country.\" But, in Table 2, it seems that the correlation of explicit and concurrency behavior is 0.08-0.58. The authors might want to review the numbers.Reply:Thank you for pointing out this error which has been corrected in the new version of the paper.Discussion1. The authors provide great information in that based on the finding, current efforts of targetting high-risk individuals may not be enough. Could the authors provide more information on, specifically, what may be the potential approach to address concurrency, especially targetting implicit attitudes on a population level?Reply:This paragraph has now been expanded with the following text to address this question:Current efforts targeting concurrency are largely focused on higher risk individuals which are unlikely to result in a shift in the population distribution in implicit attitudes to concurrency 5, 27, 31 . One approach may be follow the Know Your Network concurrency reduction intervention which was shown to be feasible and acceptable in a rural Kenyan population ( Knopf A, et al. \"This is the medicine:\" A Kenyan community responds to a sexual concurrency reduction intervention. Social Science & Medicine 2014; 108: 175-184).2. In the discussion section, the authors seemed to imply that the findings may apply to population level. However, given the sampling strategy, I am not sure if such argument may be the case. Could the authors provide more evidence or explanation?Reply:This is true. In the new version we have changed this claim to the following weaker claim:We found weak support for this latter interpretation in the form of a population level correlation between implicit attitudes towards concurrency and the practice thereof. This is however based on a small number of samples and this finding should be regarded as tentative. References 1. Leridon H, van Zessen G, Hubert M. The Europeans and their sexual partners. Sexual behaviour and HIV/AIDS in Europe: Comparisons of national surveys 1998: 165-96.2. Kenyon C, Buyze J, Colebunders R. HIV prevalence by race co-varies closely with concurrency and number of sex partners in South Africa. PLoS One 2013; 8: e64080.3. Mah T. Prevalence and correlates of concurrent sexual partnerships among young people in South Africa. Sex Transm Dis 2010; 37: 105-8.4. Maughan-Brown B, Kenyon C, Lurie MN. Partner Age Differences and Concurrency in South Africa: Implications for HIV-Infection Risk Among Young Women. AIDS and Behavior 2014: 1-8." } ] }, { "id": "37802", "date": "05 Oct 2018", "name": "Maddalena Marini", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe present study investigates the explicit and implicit attitudes to sexual concurrency in a sample of Belgium and South African students. Explicit attitudes were measured using self-reported items, while implicit attitudes were assessed by means of an Implicit Association Test (IAT). At the explicit level, both groups reported negative attitudes towards sexual concurrency, while at the implicit level, strong implicit preferences for monogamy were found only among Belgian students. No implicit preference between monogamy and multiple partners were indeed observed among South African students.\nGeneral Comments: The aim of this study is very interesting because it highlights an implicit attitude to sexual concurrency that may underlie the origin of generalized HIV epidemic in South Africa. However, I have some reservations about the methodology, the statistical analysis, and the conclusions that prevent me from supporting the publication of this manuscript in its present form.\nAbstract\nAuthors state that their results “are compatible with the thesis that differences in implicit attitudes to concurrency play a role in the genesis of generalized HIV epidemics”. Since the present study did not investigate directly the relationship between implicit attitudes and HIV epidemic, I would suggest to be more careful in drawing these conclusions.\nIntroduction\nOn page 3, the authors state: “(ii) if the variation between these two populations involves a difference in behavior of ‘core-groups’ or general population shifts (iii) the correlation between implicit and explicit attitudes to concurrency and reporting that one has engaged in concurrency at both individual and population levels”. What did the authors exactly mean with ‘core-group’? Why is it important to evaluate the correlation between implicit and explicit attitudes at the individual and population levels? I suggest the authors to discuss more in details these points, their rationale and relevance.\nMethods\nFor the IAT structure and its stimuli, authors refer to one of their previous publications1. However, I would recommend the authors to provide a description of the IAT procedure as well as examples of stimuli used in their study, so that readers can better understand this instrument.\nStatistical analysis\nCould the authors please provide additional information on how they computed the D-scores? In addition, it would great if the authors could also provide more details on the procedure used for individual and population level analyses.\nResults\nThe authors state that “there was no difference between C-IAT results by gender in either country (data shown)”. I suggest the authors to report also the relative statistics. Could the authors please provide additional information on how they computed the Cohen’s d scores reported in the results of implicit and explicit attitudes? The r range of explicit measures reported in the text and in Table 2 is different. In the text, it is from 0.13 to 0.58, while in Table 2 it is from 0.08 to 0.58. Please clarify.\nDiscussions\nCould the authors please clarify the following statement reported on page 6 “If this is the case and this acceptance is causally linked to higher concurrency rates then the distribution of implicit responses to concurrency amongst South African students suggests that population level interventions would be required to address this issue”? The authors state “the C-IAT may constitute an important marker of a population-wide greater tolerance of concurrency which broadly enables higher concurrency rates but that other specific risk factors then determine which individuals will engage in concurrency”. Could they please elaborate on that?\nMinor comment: I would suggest the authors to read carefully the text as it presents several typos and problems in the grammatical structure.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? No\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4062", "date": "18 Oct 2018", "name": "Chris Kenyon", "role": "Author Response", "response": "General Comments:The aim of this study is very interesting because it highlights an implicit attitude to sexual concurrency that may underlie the origin of generalized HIV epidemic in South Africa. However, I have some reservations about the methodology, the statistical analysis, and the conclusions that prevent me from supporting the publication of this manuscript in its present form.Abstract Authors state that their results “are compatible with the thesis that differences in implicit attitudes to concurrency play a role in the genesis of generalized HIV epidemics”. Since the present study did not investigate directly the relationship between implicit attitudes and HIV epidemic, I would suggest to be more careful in drawing these conclusions. Reply:The abstract conclusion has been changed to the following:South African students were found to have less of a preference for monogamy than Belgian students.Introduction On page 3, the authors state: “(ii) if the variation between these two populations involves a difference in behavior of ‘core-groups’ or general population shifts (iii) the correlation between implicit and explicit attitudes to concurrency and reporting that one has engaged in concurrency at both individual and population levels”. What did the authors exactly mean with ‘core-group’? Why is it important to evaluate the correlation between implicit and explicit attitudes at the individual and population levels? I suggest the authors to discuss more in details these points, their rationale and relevance. Reply:The following text has been added to the end of the introduction to make this clearer:Our rationale for exploring if the variation between these two populations involves a difference in behavior of ‘core-groups’ or general population shifts is based on the work of Rose and others 25– 27. They argued that if one finds a bimodal distribution in behavior 'A' in population 'B' compared to a normal distribution in a comparison population 'C' then this finding would be compatible with the existence of a core-group with higher risk behaviour in population 'B' being responsible for some of the differences in behavior 'A' between the two groups. The concept of a 'core-group' is well established in the HIV field and typically refers to a subpopulation with a high level of sexual network connectivity (conferred by features such as partner concurrency and rate of partner change) that contributes disproportionately to the spread of HIV in that population . Kenyon CR, Tsoumanis A, Schwartz IS: A population's higher-risk sexual behaviour is associated with its average sexual behaviour-An ecological analysis of subpopulations in Ethiopia, Kenya, South Africa, Uganda and the United States. Epidemics. 2016;15:56–65. 27266849 10.1016/j.epidem.2016.02.002 Methods For the IAT structure and its stimuli, authors refer to one of their previous publications1. However, I would recommend the authors to provide a description of the IAT procedure as well as examples of stimuli used in their study, so that readers can better understand this instrument. Reply:This information has been provided as a two new online files:S1 IAT Test. Concurrency implicit association test.S2 Figures and words used in concurrency implicit association test. Statistical analysis Could the authors please provide additional information on how they computed the D-scores? Reply:The D600 scores were calculated as follows:D600-scores of the IAT were calculated according to the standard protocol suggested by Greenwald et al. Reaction times of the second target-attribute combination were subtracted from the first combination, correcting for combination sequence, and divided by the pooled standard deviation of all practice and test phases. Scores usually vary between -2 and +2, with high scores indicating strong implicitpreferences for monogamy and concurrency, respectively, with zero indicating absence of preference, positive scores indicating a positive implicit association with concurrency (and a negative association with monogamy), and negative scores indicating a negative implicit association with concurrency (and a positive association with monogamy). Before calculating the D600 score, the minimum response time was set at 400 ms, the maximum response time at 2500ms. Any responses below this interval were omitted while any responses above this interval were recoded to 2500 ms. Reaction times of incorrect answers were raised using a penalty of600 ms. In addition, it would great if the authors could also provide more details on the procedure used for individual and population level analyses. Reply:For the population level analyses, Spearman’s correlation was used to assess the population level correlations between the point-prevalence of concurrency and intrinsic (mean D600) and extrinsic (mean values for each of the 3 variables considered separately) attitudes. The populations were defined according to self-reported sexuality by country and gender. Only populations with n > 10 were utilized for the analyses.For the individual level analyses, the statistical analytical strategy used is described in the methods:We compare the distributions of C-IAT (D600-scores) scores between Belgium and South Africa visually using histograms and statistically using t-tests for independent samples. In keeping with standard practice in this field, we used Cohen’s d as a measure of effect size. Cohen’s d was calculated by dividing the South African minus the Belgian mean difference D600 by the pooled standard deviation. Pearson’s correlation was used to test the correlations between implicit and explicit attitudes as well as between these two and self-reported point-prevalence of concurrency. Chi-squared and t-tests were used to test differences between categorical and continuous variables. All analyses were repeated stratified by gender. There were differences by gender in self-reported concurrency and explicit (but not implicit) attitudes towards concurrency. These differences were, however, congruent between Belgium and South Africa and did not affect the results. As a result, only unstratified results are reported. Results The authors state that “there was no difference between C-IAT results by gender in either country (data shown)”. I suggest the authors to report also the relative statistics. Reply:This information has been added to the second paragraph of the results as follows:There was no difference in mean C-IAT score between men and women in Belgium (-0.81, SD = 0.39 and -0.78, SD = 0.40, respectively) or South Africa (0.05, SD = 0.46 and -0.01, SD = 0.40, respectively) Could the authors please provide additional information on how they computed the Cohen’s d scores reported in the results of implicit and explicit attitudes? Reply:Cohen’s d was calculated by dividing the South African minus the Belgian mean difference D600 by the pooled standard deviation. The r range of explicit measures reported in the text and in Table 2 is different. In the text, it is from 0.13 to 0.58, while in Table 2 it is from 0.08 to 0.58. Please clarify. Reply:Thank you for pointing this out. It has been corrected to 0.08 to 0.58 in the text. Discussions Could the authors please clarify the following statement reported on page 6 “If this is the case and this acceptance is causally linked to higher concurrency rates then the distribution of implicit responses to concurrency amongst South African students suggests that population level interventions would be required to address this issue”? Reply:We have rewritten this section to make it clearer. It now reads:If our IAT results from South Africa are indeed reflective of a broader acceptance of concurrency than a population such as Belgium and this acceptance is causally linked to higher concurrency rates then the distribution of implicit responses to concurrency amongst South African students suggests that population level interventions would be required to address this issue. The authors state “the C-IAT may constitute an important marker of a population-wide greater tolerance of concurrency which broadly enables higher concurrency rates but that other specific risk factors then determine which individuals will engage in concurrency”. Could they please elaborate on that? Reply:The following text has been added to this section to make it clearer:What this suggests is that there is a broader acceptance of concurrency in South Africa at an implicit level. This might play a role in determining the higher prevalence of concurrency in South Africa. Other factors such as previous experience of partner concurrency may then determine which specific individuals engage in concurrency . Minor comment:I would suggest the authors to read carefully the text as it presents several typos and problems in the grammatical structure. Reply:The paper has been re read and a number of typos corrected." } ] } ]	1	https://f1000research.com/articles/7-608
https://f1000research.com/articles/7-1656/v1	17 Oct 18	{ "type": "Method Article", "title": "Orchestrating a community-developed computational workshop and accompanying training materials", "authors": [ "Sean Davis", "Marcel Ramos", "Lori Shepherd", "Nitesh Turaga", "Ludwig Geistlinger", "Martin T. Morgan", "Benjamin Haibe-Kains", "Levi Waldron", "Marcel Ramos", "Lori Shepherd", "Nitesh Turaga", "Ludwig Geistlinger", "Martin T. Morgan", "Benjamin Haibe-Kains", "Levi Waldron" ], "abstract": "The importance of bioinformatics, computational biology, and data science in biomedical research continues to grow, driving a need for effective instruction and education. A workshop setting, with lectures and guided hands-on tutorials, is a common approach to teaching practical computational and analytical methods. Here, we detail the process we used to produce high-quality, community-authored educational materials that are available for public consumption and reuse. The coordinated efforts of 17 authors over 10 weeks resulted in 15 workshops available as a website and as a 388-page electronic book. We describe how we utilized cloud infrastructure, GitHub, and a literate programming approach to robustly deliver hands-on tutorials to participants of the annual Bioconductor conference. The scripts, raw and published workshop materials, and cloud machine image are all openly available. Our approach uses free services and software and can be adapted by workshop organizers and authors in other contests with appropriate technical backgrounds.", "keywords": [ "education", "software", "informatics", "bioinformatics", "Bioconductor", "R", "literate programming", "markdown", "github", "open source" ], "content": "Introduction\n\nMethods of biomedical data analysis are rapidly evolving, creating a crucial need for constantly up-to-date learning materials. Workshops given by topical experts that combine didactic lectures with hands-on, guided tutorials are a common approach to teaching data analysis. The educational materials produced by such workshops, however, are often difficult to find or utilize by the rest of the community after the workshop is over. Each year, the Bioconductor project organizes and hosts a scientific conference that features scientific talks, poster presentations, networking sessions, and hands-on workshops. Nearly half of the conference time is devoted to hands-on workshops ranging from introductory to specialized and advanced topics. We sought an organized approach to developing workshop materials that meets the goals of allowing multi-disciplinary content by multiple contributors, providing “literate” code1 that is presented in context with its explanation and runs reproducibly, can be disseminated broadly in a self-contained format, and can be efficiently updated for the next iteration. This article describes the approach adopted by the Bioconductor 2018 conference to coordinate 15 workshops, contributed by 17 authors, deliver these to conference participants, and then freely disseminate a book of the materials. The book is available at https://bioconductor.github.io/BiocWorkshops/.\n\nWith a relatively large number of contributors, we set out to produce workshop materials that:\n\n1. Maintained a basic level of functionality and standardized style,\n\n2. Could be used for interactive sessions and as standalone educational materials;\n\n3. Could be improved by community contribution and input;\n\n4. Could be formally published online and as a “published” work;\n\n5. Would incorporate some educational best-practices;\n\n6. Would promote smooth workshop offerings by allowing students to work on cloud instances known to run all workshops without error; and\n\n7. Would allow easy re-use of the instructional materials and cloud instances by others after the conference.\n\nThe approach involved a call for workshop syllabi that were vetted by committee, and requiring that authors of accepted workshops contribute their materials in R Markdown format to a central GitHub repository. Workshop editors compiled the collated book with Bookdown2, using Packer (https://packer.io) to reproducibly create an Amazon Machine Image (AMI) capable of compiling the workshops. The AMI was also used to provide cloud-based virtual machines to participants during the workshops. By performing ongoing code testing on this same AMI, and using issue tracking to inform authors of issues with their code, workshop organizers ensured smooth delivery of workshops without software installation slowdowns or unexpected issues with incompatible computational environments. We describe the process in order to enable others to use the approach and the specific code we created, and to highlight the areas identified for further improvement. The process uses free services and software and can be replicated without monetary cost given appropriate technical background of authors and editors.\n\n\nRequisite technical skills\n\nThe process described requires at least one editor with the computational skills to:\n\n1. Create a Bookdown project;\n\n2. Test workshops and build the collated book;\n\n3. Provide debugging help to workshop developers;\n\n4. Use git branches and GitHub pages to publish the book; and\n\n5. Create the virtual image containing all workshop materials and required software, ideally using Packer and/or Docker (https://www.docker.com) for reproducibility.\n\n6. Use GitHub issue labels and assignments to organize and assign tasks.\n\nWe expect that the workshop syllabus template, bookdown configuration files, build scripts, Packer definitions, and model GitHub repository may be helpful for other groups considering applying a similar approach to workshop creation.\n\nWorkshops authors must to be able to:\n\n1. Author R Markdown document (or another literate programming style agreed on by the organizers) and supporting files; and\n\n2. Make commits to GitHub.\n\nThe editors exempted one author from the third requirement, but this was for an already well-tested workshop. This would not be practical for other workshops that required iterations of debugging and testing. All other authors made their own commits to GitHub.\n\n\nMethods\n\nThe process began with a public call for workshop proposals in the form of a syllabus including a summary, a schedule, prerequisites, and learning goals. Requiring such a syllabus allowed i) the committee to carefully vet the instructional qualities of proposals, ii) participants to better anticipate the contents and learning outcomes of each workshop, including those needing to justify conference attendance in terms of specific learning objectives, iii) prompted instructors to devise learning goals and objectives for their workshops, and iv) promoted standardization of workshop formats. The syllabus proposal was also included at the start of each workshop and “chapter” of the produced book.\n\nA template in Markdown format was provided with the call for syllabi (see http://bit.ly/2OfqlI9). The syllabus template included a minimal summary of relevant pedagogical theory to guide workshop authors in creating meaningful learning goals and objectives.\n\nFor authors of accepted syllabi, deadlines were given for i) submission of a draft workshop, and ii) a complete workshop that compiles without errors. The first deadline, approximately 4 weeks before the conference, proved important to engaging authors at a sufficiently early stage to provide them with “collaborator” access to the repository and begin the sometimes lengthy process of debugging both individual workshops and the collection as a whole. The final deadline, 1 week before the conference, gave editors time to debug individual workshops with errors, build the conference workshop book, and create the cloud image to be used by conference participants. In hindsight, these deadlines seemed adequate provided that editors were available for an intensive period of engagement starting from just before the first deadline. In the end, 2 days after the final deadline were allowed for resolving errors, at which point the deadline was considered passed and access privileges of everyone but editors was changed to read-only. This process left 4 days for the editors to fix remaining overall and workshop-specific problems without new ones being introduced, successfully build the book of materials, review the book for more subtle problems like missing images and incorrect formatting, and create an AMI for use by participants at the workshop.\n\nWe used the GitHub Issue-tracking system to track the progress of individual workshops and to resolve project-wide issues. Authors were asked to submit a new issue as a way of providing their GitHub usernames, after which they were given collaborator write permission to the repository. A single issue was created for each accepted workshop and given the custom label “Not Started”, using the `ghi` GitHub Issue command-line tool (https://github.com/stephencelis/ghi) to streamline creation and labeling of 15 new issues. This process could be streamlined by creating the per-workshop issues first, then asking authors to make a comment on the issue for their workshop. As workshop authoring progressed, the “Not Started” label was replaced with color-coded labels of “Incomplete,” “Problems,” “Looks OK,” and “Status Unknown.” Editors used these issues to document and share workshop-specific build errors, and authors used them to respond and ask workshop-specific debugging questions. We did not find the Travis Continuous Integration (CI) system3 usable due to the resource intensiveness of building the workshops, and the need to test each workshop individually as well as the book as a whole. Automated testing of such a project, including posting of workshop-specific errors to the correct issue, is an area for future improvement.\n\nWe applied the approach of the Bioconductor project to ensure smooth-running workshops in which i) instructors could present a document containing runnable code and results, while ii) students could follow along by running the provided code, without wasting limited workshop time due to students running different software versions, having to install software and dependencies, or trying to run broken code. To accomplish this, we implemented the workshops in a development cycle characterized by rapid changes and continuous integration, followed by a release cycle frozen to new functionality. The development cycle included the initial phase of author contribution, an initial freeze of 3 days where authors were prevented from further contributions but editors continued debugging and building, and a final release freeze 2 days before the start of the workshop. The final freeze was used to resolve dependencies and to build and test the Cloud image used by conference participants. This schedule provided maximum development time to authors but required an intensive development period for organizers immediately before the conference, so we emphasize the importance of these final deadlines being “hard” ones.\n\nAt the start of the initial development phase we provided instructions in the README.md file of the project GitHub repo (https://github.com/Bioconductor/BiocWorkshops)4 on how authors could test their workshops individually. This testing procedure was a departure from procedures most authors were already familiar with, and some authors submitted untested code. In an ideal Continuous Integration environment, each new commit would have been automatically and centrally tested, but the editors resorted to a manual process of regularly testing the individual workshops and providing feedback. This was in part due to limitations of the free tier of most continuous integration systems. (e.g., https://docs.travis-ci.com/user/customizing-the-build/#build-timeouts). Considering some features of the BiocWorkshops repository, such as a 20 gigabyte repository (including cache and data files) and about 130 R package dependencies, it may be more practical to set up a paid or in-house Continuous Integration environment for future work. Another possibility is to build and test each workshop in an individual GitHub repository, with these being incorporated as Git submodules in a master repository (https://git-scm.com/book/en/v2/Git-Tools-Submodules).\n\nWorkshops were authored using R Markdown, and compiled into a book (PDF and ePub) and website using Bookdown R package. Bookdown, in turn, uses the gitbook publishing system (https://www.gitbook.com/) to produce a variety of formats from the same source material. R Markdown files intended to be part of a Bookdown project do not contain the required front matter of a typical stand-alone R Markdown document. To help authors use and test the correct format, we seeded each workshop document with the syllabus that had been submitted by that author, and successfully built the book of the submitted syllabi. Each workshop represented a chapter of a book compiled using the Bookdown software. This approach provided several advantages:\n\nR markdown syntax is already familiar to any developer of a Bioconductor package, since it is the standard approach to creating the package “vignette” or prose documentation.\n\nR markdown implements “literate programming” by including formatted text, runnable code, and output of the code\n\nBookdown allows collating chapters as a clean, lightweight online book format, and pandoc additionally allows creation of PDF and ePub formats\n\nThese formats can then be self-published with options to order paper copies through companies such as https://leanpub.com\n\nThis approach allowed automatic installation of required packages by listing them in the DESCRIPTION file required by R packages.\n\nThe organization and coordination process benefited from a long-term working relationship among editorial team members, prior experience with organizing similar conferences, and good communication among team members. Division of responsibilities was natural based on team member interests and skills. Additional deadlines or organizational structure might be required in cases where editors and organizers are less familiar with each other’s skills and interest.\n\nThe following editorial jobs were required. Except where noted as “committee”, each task was assigned to a single individual, clarifying responsibilities.\n\nCreating the syllabus template\n\nReviewing and selecting submitted syllabi (committee)\n\nReminding of upcoming deadlines, and chasing after authors who have missed deadlines\n\nContent scoping including:\n\nLabeling workshops as “Learn”, “Apply”, or “Develop” (committee)\n\nAdvising some authors on overly lengthy or out-of-scope material\n\nCleaning up formatting such as lengthy automatically generated messages\n\nAssigning course numbers to group similar workshops together\n\nCreating the publication scripts (Bookdown shell)\n\nTesting individual workshop documents and posting issues to GitHub (multiple editors)\n\nCentral debugging of individual workshops when needed (multiple editors)\n\nBuilding the whole book and publishing it through GitHub-pages\n\nCreating a packer script to automate AMI creation\n\nFinal AMI creation and cloning for the conference\n\nWe chose to use commercial cloud service (Amazon Web Services (AWS)) to provide stable networking, common hardware and software, and minimize venue IT requirements and installation problems during the workshops. An AMI is a template that can be used to launch many identical virtual machines. The Bioconductor project maintains AMIs for recent versions of R and Bioconductor (http://bioconductor.org/help/bioconductor-cloud-ami/), which we extended iteratively throughout the workshop development and editing process. The final AMI, including RStudio server and R, required Bioconductor and R packages, and all workshop materials and dependencies, was used to launch an identical virtual machine for each workshop participant. Based on the ongoing testing, we knew that the image (AMI ID: ami-bac2c5c5) was adequate to run all workshop materials. The approximate cost of running all instances was less than $1000 for 5000 compute hours. The technology for distributing unique images to each workshop participant is an area for future improvement. For example, using Kubernetes (https://kubernetes.io/) on a Cloud cluster to instantiate Docker images for each participant may be a more cost-effective and portable solution.\n\nRather than manually installing new software or modifying virtual machine configuration by hand, we adopted the infrastructure-as-code paradigm using the Packer toolkit (https://packer.io). The requirements to build the virtual machine image were described in a JSON format (https://www.json.org/) file which Packer can use as input to reproducibly create an AMI.\n\nThe AMI was used by workshop organizers to build and test the workshop materials during the workshop editing process and was regenerated multiple times as individual workshop authors updated their respective materials and dependencies. In particular, with each addition of new R packages or dependencies, a new AMI was created.\n\nThe R package system already includes a well-defined approach for describing dependencies, the “DESCRIPTION” file (https://cran.r-project.org/doc/manuals/r-release/R-exts.html#The-DESCRIPTION-file). We asked workshop contributors to make additions to a single DESCRIPTION file in the workshop git repository when adding or changing dependencies. The presence of a correctly formatted DESCRIPTION file is sufficient to comprise an R package, so including the DESCRIPTION file in the top-level directory of workshop materials allowed standard R installation mechanisms to add dependencies to each AMI version.\n\nAfter the finalized AMI was available, a custom application was used to launch a virtual machine for each participant using the AMI as a template. For this particular set of workshops, the machines were configured to launch an m4.xlarge instance (4 virtual CPUs, 16 GB of RAM) with 100 GB of disk storage. As the virtual machines were accessed via an Rstudio server, all participants shared a common user experience through the Rstudio interface that nearly all were already familiar with.\n\nTo gain access to a personal virtual machine with the workshop materials, each user supplied a user ID (the user’s email) and a password that was the same for all participants. A new virtual machine was then launched and the attached IP address associated with the user’s email address. Further attempts by the same user to launch a virtual machine would simply supply the participant with the same instance. The virtual machines were created using a common Bioconductor account and were terminated automatically after the conference.\n\n\nResults\n\nThe process organized 15 workshops presented at the Bioc2018 conference into a single publicly available book. Each workshop comprised a single self-standing chapter of the book. Instructors directly presented their chapter materials at the conference, potentially complemented by slides and live-coding. Workshop participants were able to run the workshop code without software installation or errors, making efficient use of the 1- or 2-hour workshop time for learning through both observation and application.\n\nWorkshop materials are available via a website, a PDF book, an ePub book, and the commercial self-publication service, https://leanpub.com. The PDF version includes 388 pages, the result of 19 contributors making 313 separate changes (commits) to the materials over the course of 10 weeks. Workshops were qualitatively classified as 100-level “Learn”, 200-level “Apply”, or 500-level “Develop”, with 4, 9, and 2 chapters, respectively. Whereas conference attendance was ~120 participants, approximately 10 times as many have viewed the published materials in the 2 months since they were posted online.\n\n\nDiscussion\n\nWe have described the approach adopted by the Bioconductor 2018 conference organizers to coordinate 15 workshops, contributed by 17 authors, and deliver these to participants over 22 total hours of instruction and hands-on tutorials (see http://bioc2018.bioconductor.org/schedule). The process described above includes a combination of social contracts such a deadlines and clear responsibilities of organizers and contributors that, when combined with social coding practices and modern publishing tools, facilitated the creation of 388 pages of content (PDF version) over the course of 6 weeks in a format that can readily be updated annually. A workshop-specific AMI and virtual machines enhanced the participant experience by eliminating time-consuming installation problems and ensured that appropriate IT infrastructure (compute, storage, and performant networking) were available.\n\nBioconductor, as a project and as a community, values openness and sharing. Workshops were developed publicly with the goal of creating not only published materials, but having a set of modifiable and iterable raw material that others can repurpose for their own learning or instruction outside the context of the Bioconductor conference.\n\nWe were able to quickly develop a coherent and cohesive set of workshop materials that were appropriate for interactive sessions and as standalone educational materials. Providing a content template with specific required fields resulted in materials that could be quickly evaluated by potential participants with respect to i) included content, ii) learning objectives and goals, and iii) prerequisites. Using R markdown-based content ensured that included code would run without error. Cloud infrastructure minimized the need for local compute resources, limited network bandwidth needs to participant computers, and guaranteed an identical and tested compute environment. Clear deadlines and responsibilities for both contributors and editors resulted in 15 out of 16 proposed workshops to be included in the final materials, suggesting that contributors understood the requirements at the time of proposal and were able to follow through without our process introducing an undue burden.\n\nThe workshops proved easy to create due to the set pattern of guidelines present for submitting a workshop. The guidelines helped instructors to think about the structure and time limitations of the workshop, and to communicate precise learning goals, providing little chance to skew off topic. The R markdown format is particularly helpful for reproducibly compiling and collating the workshops ahead of time, because this format is widely familiar in the Bioconductor community. This format is completely open source, and authors did not have the need to create new “accounts” or “pay” for any services to contribute to this format. All work happened on Github in a centralized open source environment, allowing authors to take technical and pedagogical insights from other workshop authors. Testing that an individual contributed workshop was compatible with the whole bookdown document was straightforward given the use of common tools in our community. This ease of local testing made producing the workshop quite efficient. It was important that each workshop author was responsible only for ensuring that her own content could build successfully, without concern for formatting or integration into the entire document and website.\n\nAll the non-technical aspects addressed in this manuscript, such as prescribing deadlines and standard syllabi, are equally applicable in other environments. The Packer system for creating reproducible, reusable AMIs is also quite general and can be tailored to install essentially any software, language, or workshop materials. The bookdown package and R markdown software stack that we describe supports more than 40 additional languages besides R (https://bookdown.org/yihui/rmarkdown/language-engines.html), albeit at a less granular level than the R language support. Jupyter notebooks share some similarities with R markdown and are also quite popular for developing computational workshop materials. The gitbook publishing software (https://www.gitbook.com/), upon which the bookdown R package is based, is capable of taking the resulting markdown files and producing a book or website.\n\nAreas of future work include (i) enhancing the quality of individual workshops as well as adding new material, (ii) streamlining the iterative development and build process through continuous integration, and (iii) adopting containers such as Docker as the delivery mechanism rather than the proprietary AMI. We expect that using a social coding paradigm will result in continuous improvement and easy reuse of workshop materials. We plan to adopt a continuous integration pipeline that builds and delivers complete (and versioned) workshop materials. Adjustments to the current build process to leverage parallel processing can speed workshop builds, facilitating the continuous integration approach. Adopting a container technology such as Docker (https://docker.io) to deliver pre-built workshop materials will increase portability and remove the requirement to use a proprietary system such as Amazon Web Services. Finally, we plan to improve modularity of workshop materials to promote the creation of custom combinations tailored to workshop organizer needs.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\n\nResource availability\n\nThe workshop book is available from: https://bioconductor.github.io/BiocWorkshops/.\n\nAll source code and configurations are available at: https://doi.org/10.18129/B97H034.\n\nLicense: CC BY 4.0\n\nThe Amazon Machine Image with the pre-installed workshop materials and required software is available as ID: ami-bac2c5c5.", "appendix": "Grant information\n\nResearch reported in this publication was supported by the National Human Genome Research Institute of the National Institutes of Health under award number U41HG004059, the National Cancer Institute of the National Institutes of Health under award number U24CA180996, and the Center for Cancer Research, part of the Intramural Research Program at the National Cancer Institute at the National Institutes of Health. Part of this work was performed on behalf of the SOUND Consortium and funded under the EU H2020 Personalizing Health and Care Program, Action contract number 633974.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nThe authors wish to acknowledge contributors of the useful workshop materials to the Bioconductor workshops repository.\n\n\nReferences\n\nKnuth DE: Literate Programming. Comput J. 1984; 27(2): 97–111. Publisher Full Text\n\nXie Y: bookdown: Authoring Books and Technical Documents with R Markdown [Internet]. Boca Raton, Florida: Chapman and Hall/CRC; 2016. Publisher Full Text\n\nVasilescu B, Yu Y, Wang H, et al.: Quality and Productivity Outcomes Relating to Continuous Integration in GitHub. In: Proceedings of the 2015 10th Joint Meeting on Foundations of Software Engineering. New York, NY, USA: ACM; (ESEC/FSE 2015), 2015; 805–16. Publisher Full Text\n\nBioconductor: Bioconductor Workshop materials for 2018. Bioconductor. Software. 2018. Publisher Full Text" }	[ { "id": "39598", "date": "05 Nov 2018", "name": "Ewy Mathe", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe work presented here addresses key topics in the biomedical research community at large, including reproducibility, team coding, installation-independent and error-free code environments. The main goal of the authors was to deliver public resources through published materials through modifiable and raw materials and source codes that can readily be leveraged by participants during and after the Bioconductor training. With this in mind, authors present a generalizable model for producing an open-source workshop in an efficient manner that includes a published work of contributed sessions/education materials that can be used beyond the conference, a guide to best practices in education, an installation-free environment for workshop participants with error-free code. Overall, this approach to creating workshops and educational materials is exemplary for the biomedical research community as it provides public, transparent, reproducible, and free solutions.\nSome minor suggestions that could be included:\nOne practical aspect that could be reported is the specific number of editors that was used, particularly given the heavy work load of editors throughout this process. Do participants still have access to an AMI to avoid having to install everything themselves on their own after the workshops? There is a small typo in the second sentence of the Discussion “social contracts such a deadlines” should read “social contracts such as deadlines”. Under Future directions and challenges, the concept of using a Docker rather than the proprietary AMI is mentioned twice. Have authors collected feedback from workshop users, particularly regarding whether they are easily able to use material outside of the conference workshops?\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes", "responses": [] }, { "id": "40382", "date": "13 Nov 2018", "name": "Kevin Blighe", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis work addresses issues that are now faced in the biological and medical (biomedical) sciences, namely: reproduciblility (broadly speaking); ease of learning and accessibility to learning materials; maintenance of software, analysis pipelines, and methods; et cetera. The authors have a positive track record in this area and are in a good position to address the community in this regard. In particular, the authors focus on workshops as an ideal form of learning, which is certainly true, and make their material available as an e-book and other resources free to the global community.\nThroughout their work, the authors place emphasis on development and testing of code/programs as a team such that these can be later run in any end-user environment seamlessly and without error. These are indeed core fundamentals that need to be embraced in the wider community irrespective of programming language. Of course, other technical factors that are outside of the control of the authors may still hinder workshop progression. Finally, the emphasis on learning in a well-structured and standardised fashion will help future workshops to be successful.\nThe text is clear throughout and the authors provide sufficient details such that their efforts can be reproduced by others. Broadly speaking, the push toward standardisation is key to the integration between 'big data' and the biomedical sciences, and here the authors address one component of that.\nSome minor comments:\nOn page 3, left column: Could you elaborate on what you mean by 'Would incorporate some educational best-practices'? Do you mean a set of pre-existing best practices or just speaking generally? In the same listing, point 7 somewhat overlaps with the second part of point 2 - not a major issue and can be left as is. On Page 3, right column: In 'Required technical skills', it is stated 'The editors exempted one author from the third requirement, but this was for an already well-tested workshop.' - do you mean the 3rd requirement from the list of requirements for editors? There are only 2 requirements in total listed for authors. Page 4, left column: in 'Deadlines' section, a Gantt chart could help here, but not necessary. Page 5, left column: In 'Organization, coordination, and editorial responsibilities', is 'committee' the same as 'multiple editors'? Could probably clarify this at the start of the list with 'Except where noted as “committee” or \"multiple editors\", each task was assigned to a single individual, clarifying responsibilities.' Page 7: it may help to very briefly explain what needs to be done in order to load the AMI for those who have not done this previously\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Yes\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1656
https://f1000research.com/articles/7-1655/v1	17 Oct 18	{ "type": "Opinion Article", "title": "Publishing peer review materials", "authors": [ "Jeffrey Beck", "Kathryn Funk", "Melissa Harrison", "Jo McEntyre", "Josie Breen", "Andy Collings", "Paul Donohoe", "Michael Evans", "Louisa Flintoft", "Audrey Hamelers", "Phil Hurst", "Thomas Lemberger", "Jennifer Lin", "Niamh O'Connor", "Michael Parkin", "Sam Parker", "Peter Rodgers", "Magdalena Skipper", "Michael Stoner", "Josie Breen", "Andy Collings", "Paul Donohoe", "Michael Evans", "Louisa Flintoft", "Audrey Hamelers", "Phil Hurst", "Thomas Lemberger", "Jennifer Lin", "Niamh O'Connor", "Michael Parkin", "Sam Parker", "Peter Rodgers", "Magdalena Skipper", "Michael Stoner" ], "abstract": "Publishing peer review materials alongside research articles promises to make the peer review process more transparent as well as making it easier to recognise these contributions and give credit to peer reviewers. Traditionally, the peer review reports, editors letters and author responses are only shared between the small number of people in those roles prior to publication, but there is a growing interest in making some or all of these materials available. A small number of journals have been publishing peer review materials for some time, others have begun this practice more recently, and significantly more are now considering how they might begin. This article outlines the outcomes from a recent workshop among journals with experience in publishing peer review materials, in which the specific operation of these workflows, and the challenges, were discussed. Here, we provide a draft as to how to represent these materials in the JATS and Crossref data models to facilitate the coordination and discoverability of peer review materials, and seek feedback on these initial recommendations.", "keywords": [ "peer review", "scholarly publishing", "JATS", "JATS4R", "Crossref" ], "content": "Introduction\n\nPeer review is the practice of subjecting a scholarly article, such as a research paper submitted to a journal, to scrutiny or review by others ('peers') who are experts in the same field. Generally, if the author of the article addresses the concerns raised during peer review to the satisfaction of an editor, the article is accepted for publication. The peer review process produces a trail of documents, which can include: different versions of the article; the reviewer reports (with or without the name of the reviewer); responses by the author to the reports; and various letters (including cover letters from the author and decision letters from the editor). It is also possible for an article to go through two or more rounds of peer review, which increases the number of documents generated.\n\nTraditionally the documents generated during the peer review process were only ever seen by the author, the editor and the reviewers, but a small number of publishers now publish some peer review materials alongside articles. Moreover, support for this practice has been slowly gaining momentum, driven by a wish to increase transparency and provide credit for peer reviewers (Polka et al., 2018). There were 10 journals identified in the PMC corpus that archive some peer review materials. These journals take a variety of different approaches, which results in differing levels of discoverability for these materials. Additionally, some journals were identified as publishing peer review materials but not consistently archiving them in a repository.\n\nHere we report the findings of a workshop, held at the BMA in London, UK on July 6, 2018, at which representatives from publishers, PubMed Central (PMC)/Europe PMC, and Crossref discussed the practical challenges involved in publishing peer review materials. We focus on what has to happen after a publisher decides to start publishing peer review materials, and discuss how to do this in a way that is sustainable, improves discoverability, and supports machine readability and archiving. We do not discuss the relative merits of the different approaches to peer review that have emerged over the past decade, notably the many different flavours of 'open peer review' (Ross-Hellauer, 2017), but we feel that many of our suggestions and recommendations are relevant to most if not all of these approaches.\n\n\nWhat are peer review materials?\n\nAs mentioned above, peer review materials can include: the reviewer reports (with or without the reviewer name); responses by the author to the reports; and various letters (including cover letters from the author and decision letters from the editor). Some articles go through two or more rounds of peer review, which increases the number of documents generated. While each document is usually accompanied by a date, other metadata concerning the correspondence can be highly variable. For example, peer review reports and editors decision letters may or may not include names of reviewers or editors, or their ORCID IDs; the individual materials may or may not have DOIs. In subscription journals it is also possible for peer review materials to appear in front of or behind the paywall.\n\nPublishers are approaching the publication of peer review materials in a number of ways. The aims of this group were not to prescribe what should be done from an editorial point of view, but to enable what is published to be found by readers and machines alike. Prior to the workshop, data were collected from each publisher attending and we found the following materials are published:\n\n○ Peer review reports, anonymized or with report author names\n\n○ Author responses/Rebuttals\n\n○ Editor decision letters\n\nSome journals also provide appeal and resubmission information (including previous versions of the article, dates, and actors involved).\n\nIn some cases publishers make peer review materials available as a single PDF with versioned reports linked to specific revisions of the article. Other publishers create separate artifacts, each with unique DOIs, and still others edit and amalgamate various reports into one narrative. The variety can be found in the meeting notes and a table filled out before the meeting (See Supplementary File 1 and Supplementary File 2).\n\nAfter collecting these data, representatives from each of the identified publishers were contacted to attend a workshop in London on July 6, 2018; all but one publisher was able to attend. Further publisher representatives were invited following communication with ASAPbio – these publishers are embarking on this practice. Crossref, Europe PMC and PMC were also represented at this meeting, as downstream recipients of this content or the metadata related to it. Journal representatives expanded on the data previously collected and shared details on how they collect and publish peer review materials, how these artifacts are represented in the ANSI/NISO Z39.96-2015 Journal Article Tag Suite (JATS) document standard (these principles can be applicable to other DTDs), and whether they send relevant metadata to Crossref. Crossref shared details of their schema extension to represent this form of content (https://www.crossref.org/services/content-registration/peer-reviews/). As downstream recipients of peer review materials, PMC and Europe PMC presented the perspective of archives that collaborate with journals to ensure that content is being captured in a sustainable and consistent format that fosters long-term preservation and access to the scholarly record. Understanding the goals of, workflows, and limitations on each stakeholder allowed the group to refine the scope of the discussion and its outcomes.\n\n\nPeer review materials need to “stand alone”\n\nIn order to advance the transparency and recognition of peer review materials, we agreed these peer review materials need to stand alone from the main article for the purposes of, for example, credit and citation. Ideally, each content item should have its own DOI (as per the recently enhanced Crossref schema). We identified three levels of achieving this, with level 1 being the basic and least preferred option, but probably currently the most achievable and pragmatic:\n\n1. Peer review materials are attached to the article as a single or numerous PDFs. Whether these materials are pulled together into one document or attached as separate documents, there should be some defined mechanism in the JATS XML tagging that would support the capture of any available metadata and identify these files in a machine-readable and interoperable way for publishers to tag this content appropriately.\n\n2. Peer review materials are appended to the article within the full text (so all is machine readable) as a sub-article component of the XML.\n\n3. Peer review materials are full-text XML “articles” or “commentaries” in their own right that link bidirectionally to the main article.\n\n\nRequired metadata versus rich metadata\n\nWhether the material is provided as a PDF(s) attachment to the main article, or as a full-text XML sub-article or separate article, important metadata can be attached to the item in a machine-readable way, and DOIs can be applied. What types of peer review information are available is dependent on the publisher’s peer review policy, for instance whether reviewers and editors are named, whether the peer review material carries the same license as the main article or takes another form, and what items constitute the peer review materials. Additional metadata fields, such as dates of review and date of review publication and the inclusion of ORCID IDs for reviewers and editors will also be subject to publisher policies and workflows. However, all of this material can be added to the item in a machine-readable way. Even if the actual content is not published in full-text XML format, the metadata can (and should) be.\n\nThe topic of licensing of the peer review materials was briefly discussed at the workshop but ultimately left out of the remit of this group because the JATS tagging schema would allow for different licensing information to be added for these items or to retain that of the main article, as a publisher chooses.\n\n\nChallenges in the process\n\nWhile a few of the publishers had processes in place to prepare the peer review content automatically or within a few minutes, others spend 20–40 minutes per article. In such cases, the tasks that are attributed to this time include the following:\n\nRemoving boilerplate text from review reports\n\n“Stitching together” the material from disparate locations in the submission system\n\nEditorial checks:\n\nReviewing the content for sensitive information, e.g., unpublished data additions and confidentiality leaks, as well ensuring the tone of the report is appropriate\n\nRemoving author responses that contain data the author wants to publish in a subsequent paper\n\nArbitration processes for conflicting reviews\n\nWhere time is spent—whether in the editorial or production process—depends on the publisher workflow. Regardless of workflow, the overlap in tasks identified provides evidence of the potential value of updating the infrastructure of submission systems to account for and streamline these efforts. Coordination between publishers and submission systems could minimize the time spent “stitching together” peer review materials into a publishable format.\n\nIn addition to time and workflow hurdles, another major challenge noted by those publishers without their own hosting platforms, was the actual publication process and online hosting of peer review materials. Many publishers identified that some online hosts were not able to manage this new content type. As a result, peer review materials are being captured in supplementary material sections because alternative options are not available. In such cases, it becomes more difficult to capture any relevant associated metadata in a meaningful way for the peer review materials or to make this valuable content easily discoverable.\n\nThese challenges are common for publishers in that most of the established submission systems and hosting platforms were designed and built many years ago and may be slow to accommodate new requirements. Coordinated communication with these platforms regarding the workflows around publishing peer review materials may result in more satisfactory and generic approaches to accommodating publication of peer review materials.\n\nThere are internal challenges of cost control issues that also need to be accounted for, and the publication of a single PDF is often more achievable financially based on current systems than producing full-text XML. However, the attachment of machine-readable metadata to that PDF should be within reach, especially if the submission systems and hosting platforms can build these requirements into their products.\n\n\nImportance of version management\n\nAn additional challenge may be introduced in managing peer review materials in cases where such materials are collected for more than one published version of a paper. The Recommendations of the NISO/ALPSP Journal Article Versions (JAV) Technical Working Group (2008 April), included the following types of article instances:\n\nauthors-original\n\nsubmitted-manuscript-under-review\n\naccepted-manuscript\n\nproof\n\nversion-of-record\n\ncorrected-version-of-record\n\nenhanced-version-of-record\n\nTo this list, the JATS4R working group on “Article publication and history dates” added pre-print. The JATS4R draft recommendation advises that if the publisher publishes a revision of any of these stages, the subsequent revisions should be labelled with suffixes, as follows: “-r1”, “-r2”, etc. (https://jats4r.org/article-publication-and-history-dates).\n\nIf the peer review materials reference content in a specific version of an article, that link between peer review materials and correct version should be captured in the metadata for clarity. Managing the associations between peer review materials and article version is essential for journals that make multiple versions of a paper publicly available, to ensure the archival record is accurate and that the process transparent. For example, if a journal publishes three versions of an article, any related peer review materials should be associated with the appropriate version. It should not be left to a reader to determine if a peer review report or decision letter relates to the first version, the second version, or the third version.\n\n\nJATS XML proposal (designed to aid depositing to Crossref)\n\nIrrespective of the editorial and publisher decisions regarding workflow, we propose the following options regarding JATS XML tagging, designed to also aid metadata registration with Crossref (note we are using the same terms as Crossref where controlled vocabulary is required).\n\nReview documents may be supplied as:\n\n1. sub-articles <sub-article> to the article being reviewed (sub-articles may be full-text XML or XML metadata plus a link to the PDF)\n\n2. independent articles <article> (with the appropriate <related-article> links – Peer Reviews MUST link to the version of the article they are reviewing and Author Replies; Decision Letters MUST link to the version of the article they are passing judgment on; and Author Replies MUST link to each Review/Decision Letter it is addressing)\n\n<sub-article> or <article> MUST have an article-type attribute with one of the values listed in Table 1.\n\nNote: aggregated-review-documents is not currently in the Crossref schema; that schema uses the term aggregate. Crossref has two further attributes to describe the type of content: community-comment and manuscript. The XML sub-group discussed these terms and decided to exclude them as community-comment presumably refers to post-publication comments via systems like Hypothesis and so: a) are not guaranteed to be “peer” comments and are excluded from the criteria of this paper and b) it is unlikely that publishers in the near term would pull that content back into the source JATS XML, post publication. Crossref schema also allows for a stage, pre-publication or post-publication. This is therefore also felt outside of this remit.\n\nThe term manuscript does not map to anything we’ve discussed.\n\nThis is an optional item. Currently there is no corresponding tag in JATS and so would require a request to the JATS Standing Committee.\n\nThere would be a fixed value list, mapped to the Crossref schema:\n\nmajor-revision\n\nminor-revision\n\nreject\n\nreject-with-resubmit\n\naccept\n\naccept-with-reservation\n\nNOTE: There should be no “recommendation” for author-comment type content.\n\nIt is an optional element and should be contained within <contrib>, which should contain a <name> or <anonymous/>.\n\nIf <contrib> is used, it MUST contain @contrib-type that maps to following controlled vocabulary:\n\n○ For Peer Reviews, use @contrib-type=“reviewer”\n\n○ For Decision Letters, use @contrib-type=“editor”\n\n○ For Author Reply, use @contrib-type=“author”\n\nWe intend that the @contrib-type attribute value reflects the contributor’s relationship to the peer review process and not the relationship with the document.\n\nThe <role> tag is optional and can be used for display terms of what publishers may use for their journal (for example variations on the term editor could be Academic Editor, Reviewing Editor, Senior Editor, E-i-C etc.).\n\nNames, affiliations and contributor IDs (such as ORCID), where provided, follow standard JATS tagging (see JATS4R recommendation: https://jats4r.org/authors-and-affiliations).\n\nFollow tagging recommended by JATS4R: https://jats4r.org/conflict-of-interest-statements.\n\nDOIs for peer review materials are optional but strongly encouraged. Use <article-id pub-id-type=“doi”>\n\nEach review document (standalone article or sub-article) SHOULD have license information with a machine-readable license. Review documents supplied as <sub-article> may have their own <license> element or inherit their license information from the parent document as described in the JATS4R Permissions Recommendations (https://jats4r.org/permissions).\n\nEach review document (standalone article or sub-article) MUST have a pub-date and may have other publication information captured as <event>. Review documents supplied as <sub-article> may have their own <pub-date> element or inherit their <pub-date> from the parent document.\n\nThere are some elements that MUST NOT appear in review documents:\n\na. <funding-group>\n\nb. <app>, <app-group>, <ack>, <glossary>, <back>/<sec>\n\nc. <supplementary-material>, <inline-supplementary-material>\n\nd. <bio>\n\ne. <article-version> Once published, review documents will not be “versioned”. If the reviewer(s) write a review on an updated version of the manuscript, the peer review is a new published object.\n\n\nCrossref metadata\n\nAs of November 2017, Crossref supports the scholarly discussions entailed in the publication process as well as those after publication (e.g. “post-publication reviews”). In the same fashion as all content registered with Crossref, peer review metadata is available via the open Crossref APIs and Crossref Metadata Search. For full details and example deposit XML, see the Crossref peer review deposit guide: https://support.crossref.org/hc/en-us/articles/115005255706-Peer-Reviews.\n\n\nDisplay of peer review materials\n\nWe also propose that publisher web platforms and archives display peer review materials (or links to peer review materials) in a clearly labeled peer review section. This practice will help ensure that not only are the journal processes transparent but that the content itself is easy to find and navigate to, regardless of how a journal chooses to make them available.\n\n\nFuture/next steps\n\nThis proposal is intended to lay the groundwork for the publication and archiving of peer review materials across publishers and publication models, providing flexible options to meet different journal needs and workflows. Moving forward, there is a need for continued collaboration and discussion as peer review models and workflows evolve. As the goals of these peer review efforts are more clearly defined across the publishing and academic communities, certain models may lend themselves more readily to supporting those desired outcomes. Continued efforts to identify the most critical needs of each user group should be explored through ongoing efforts such as ASAPbio and FORCE11. In turn, these needs can inform the technical solutions and recommendations going forward.\n\nFurther technical discussions should not be placed on hold in the interim, though. As publishing peer review materials practices grow, there is a pressing need for industry-wide solutions now. We would like to see the XML recommendations from this group be converted to a JATS4R recommendation on the publishing of peer review materials. Similarly, it would be of value to the community for Crossref and JATS to coordinate efforts and ensure some level of metadata alignment for peer review materials that would reduce costs to publishers and minimize barriers to implementation.\n\nThis type of coordination between publishers, archives, and other organizations that support the scholarly communication enterprise is critical to ensuring that the needs of the whole community are being met. Past experience has taught us that making content available is just the first step toward increasing transparency. Doing so in a flexible, consistent, and meaningful way is imperative in making certain that the available material is also discoverable and that long-term preservation of the content can be supported. Implementing the next steps through community-driven recommendations in a sustainable way will be important in increasing transparency and rigor of the scientific record.\n\nIf you are publishing peer review materials, or are not yet and are considering doing so, please comment.\n\n\nData availability\n\nNo data are associated with this article.", "appendix": "Grant information\n\nThis work was supported in part by the Intramural Research Program of the National Library of Medicine, National Institutes of Health. Contributions from AH and MP were supported by Europe PMC. Funding for Europe PMC is provided by 29 European-based funders of life science research: https://europepmc.org/Funders/ under Wellcome Trust grants 098321 and 108758, awarded to EMBL-EBI.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary material\n\nSupplementary File 1. Workshop agenda and notes from the meeting.\n\nClick here to access the data\n\nSupplementary File 2. Current processes of the different journals represented at the meeting.\n\nClick here to access the data\n\n\nReferences\n\nNISO/ALPSP Journal Article Versions (JAV) Technical Working Group: Journal Article Versions (JAV): Recommendations of the NISO/ALPSP JAV Technical Working Group (NISO-RP-8-2008). 2008. Reference Source\n\nPolka JK, Kiley R, Konforti B, et al.: Publish peer reviews. Nature. 2018; 560(7720): 545–547. PubMed Abstract \| Publisher Full Text\n\nRoss-Hellauer T: What is open peer review? A systematic review [version 2; referees: 4 approved]. F1000Res. 2017; 6: 588. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "39586", "date": "22 Oct 2018", "name": "Jessica K Polka", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn “Publishing peer review materials,” Beck and coauthors present the outcome of a workshop that assessed the current state and best practices for the technical representation of peer review material and metadata. The report is well-organized and proposes guidance on XML tagging that the group hopes will be adopted by JATS4R. This article is an important contribution to the conversation about increasing transparency in peer review and will facilitate its development as a valued, linked, and preserved class of scholarship.\nI don’t feel that any of the following comments need to be addressed before the manuscript is indexed in PubMed, but I do believe that the report could be strengthened by attention to these areas:\nPresentation\nThis piece would benefit from more illustration of workshop outputs. For example, is there a file you can point (or embed inline) that demonstrates best practices for tagging? In a similar vein, there is great information in Supplementary File 2. Can the process information be used to generate an in-text table?\nXML proposal Note that I lack practical experience with JATS.\nCould the tag also be used to identify student or postdoc co-reviewers? Is intended to represent the publication date of the review (presumably the same as the article) or the (in this case probably more useful to readers) date that the review was submitted? Or is that captured in ? Clarification would be helpful. Why is disallowed? I would like to imagine a future in which peer review is recognized as a valuable scholarly activity, something that funders would like to know they are supporting.\nText edits\nPlease define “BMA.” “Past experience has taught us that making content available is just the first step toward increasing transparency.” What does this refer to? “The term manuscript does not map to anything we’ve discussed.” What does this mean?\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes", "responses": [] }, { "id": "39584", "date": "29 Oct 2018", "name": "Tony Ross-Hellauer", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\n“Publishing peer review materials” reports the outcomes of a 2018 workshop convened with stakeholders engaged in the publication and dissemination of peer review report, to work towards consensus on “how to represent these materials in the JATS and Crossref data models to facilitate the coordination and discoverability of peer review materials”.\nThe article is valuable for understanding the range of publisher workflows in operation for publishing peer review at present. As such, it both makes the case for the need for standardisation of description and proposes a draft JATS XML proposal designed to aid depositing to Crossref. The proposals are pragmatic and reasonable, reflecting the variety of workflows in operation at present and the possibilities for timely transition towards a more standardised system.\nThis first version clearly indicates that it is a “draft” to “seek feedback on these initial recommendations” – but despite its provisional nature, I believe it is ready to be indexed already in this version, as it is a sufficiently complete record of the discussion to this point.\nThe article makes very clear how much work lies ahead, and that this document is just a starting point. Given its purpose, I don’t believe formal peer review is actually necessary here – as a working proposal released to the community for feedback, the views of each member of the community are equally important. I hope my comments will be received as just one more voice in that conversation, and given equal weight with views shared by other channels.\nI agree with Reviewer 1’s comments regarding useful ways to enhance the level of detail here. To this I would only add that it would be useful to describe why the JATS elements under the sub-heading “Not allowed” are so - why must they not appear in review documents?\nOther than this, I would only note some further thoughts for how these discussions could develop in the future. None of these issues need necessarily be considered within this present article, but I record them here in hopes of continuing the conversation:\n\nLinks to review policies: To understand the context of the review it would be helpful to have some lasting link to review policies under which the review was conducted. For example, what criteria were the reviewer asked to consider? Was it single/double-blind or open identities review? If the latter, were reviewer identities revealed to the author (and if so at what stage)? This information could be made available via descriptive metadata fields in this schema, or via persistent links to policy descriptions. From “is” to “ought”: I understand it is not the scope of this document to prescribe practices, but to ensure pragmatic routes to interoperability of metadata description. Hence, I understand that the “topic of licensing of the peer review materials was discussed at the workshop but ultimately left out of the remit of this group because the JATS tagging schema would allow for different licensing information to be added for these items or to retain that of the main article, as a publisher chooses”. Nonetheless, I think as a community some best-practice guidelines would be helpful to ensure the interoperability and re-usability of this content. This point goes more generally for other best-practice issues (e.g., assignment of DOIs). In the “Future/next steps” section you note that discussions should continue via other fora – here, I wonder if there is a need for a formal working group hosted by, for example, Force11, to work towards consensus on such topics. (This could also potentially link up with the PEERE group’s work on peer review data-sharing). Reviewer roles: We may in future see more stratification of reviewer roles and it would be interesting to have recorded exactly what parts of a manuscript each reviewer had considered (methodology, stats, analysis, etc.). Also, if any computational tools had been applied to check, e.g., statistics. Here it would be useful to have a standard typology of such roles. Other review materials: There are a multitude of venues for “post-publication” review – including journal article comment sections, third-party platforms like publons or pub-peer, or even individual blogs, etc. It would be useful to have formal, persistent links between such materials and the article in question so that readers may more easily place the article in the context of a continuing discussion. I imagine Crossref event data will be useful for this, but I wonder if there would be value in third-party platforms also applying this schema (suitably extended/adapted) to represent these links in a more formal way?\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1655
https://f1000research.com/articles/7-1651/v1	17 Oct 18	{ "type": "Software Tool Article", "title": "KnetMaps: a BioJS component to visualize biological knowledge networks", "authors": [ "Ajit Singh", "Christopher J. Rawlings", "Keywan Hassani-Pak", "Ajit Singh", "Christopher J. Rawlings" ], "abstract": "KnetMaps is a BioJS component for the interactive visualization of biological knowledge networks. It is well suited for applications that need to visualise complementary, connected and content-rich data in a single view in order to help users to traverse pathways linking entities of interest, for example to go from genotype to phenotype. KnetMaps loads data in JSON format, visualizes the structure and content of knowledge networks using lightweight JavaScript libraries, and supports interactive touch gestures. KnetMaps uses effective visualization techniques to prevent information overload and to allow researchers to progressively build their knowledge.", "keywords": [ "knowledge network", "knowledge graph", "network visualisation", "knowledge discovery", "biojs" ], "content": "Introduction\n\nNetworks have been widely used to visually represent complex information in many disciplines, ranging from social sciences (Szell et al., 2010) to engineering, physics, biology, computer science, design and manufacturing (Wang & Alexander, 2015). They fulfil the need to present a system, not only as individual entities but as a whole, by capturing the myriad inter-linked components within the system (Pavlopoulos et al., 2011). Networks are represented as graphs comprising a set of nodes connected by edges. Networks can be homogeneous with all the nodes within the network being of the same type, or heterogeneous with nodes and edges of various types (Sun & Han, 2012). Recently, the term knowledge network or graph has been used frequently in research and business, usually in close association with Semantic Web technologies and linked data. Knowledge networks are increasingly used to model diverse knowledge domains by acquiring and integrating information into an ontology and applying a reasoner to derive new knowledge (Ehrlinger & Wöß, 2016).\n\nA challenge when visualizing knowledge networks is to avoid information congestion and overload that could hinder user experience. The potential richness of data captured in the attributes and density of connections makes it a greater challenge to use standard network visualization tools which often focus on simply visualizing the structure of the network itself (Becker et al., 1995). In molecular biology, there is a wealth of available information, and visualizing all of it at once reduces the value of a visualization or makes it even unusable for analytical purposes, and therefore requires the development of special approaches when visualizing such data (Vehlow et al., 2015).\n\nPreviously, our group developed a web-based tool, Ondex Web (Taubert et al., 2014), for visualising knowledge networks generated with the Ondex data integration platform (Kohler et al., 2006). It supported the Ondex exchange language (OXL) and was predominantly used to visualise a biological knowledge domain. However, being a Java-applet and using legacy web technologies, it constantly led to compliance concerns on different web browsers, which hindered its reusability. The advance of modern JavaScript-based data visualisation libraries such as cytoscape.js (Franz et al., 2016) and jQuery (Benedetti & Cranley, 2011) has made it possible for us to learn from our experience with Ondex Web and to develop a new lightweight and reusable component optimised for the visualisation of content-rich knowledge networks.\n\nIn this paper we describe KnetMaps (Singh & Hassani-Pak, 2018), an interactive BioJS component to visualise integrated knowledge networks. It is well-suited for applications that require scientists to visualise complementary types of evidence in a single interactive view. KnetMaps is an important visualisation component of KnetMiner where it is used for visualising knowledge networks of crop genomes (Hassani-Pak et al., 2016) and supporting scientists to make informed decisions in gene and trait discovery research. It uses a generic design and hence can be readily embedded in other knowledge discovery applications.\n\n\nMethods\n\nThe KnetMaps component has been developed as part of the KnetMiner software suite and follows the standards set by the BioJS registry. KnetMaps employs a variety of network visualization techniques such as interactive controls, information juxtaposition and data filters. Using effective visualization techniques it prevents information overload and allows researchers to progressively explore and reveal the inter-connected entities within the larger knowledge network (Figure 1).\n\n(A1) Display configuration (set network layout, node/edge visibility, label visibility and size). (A2). Network export options (cytoscape JSON, PNG). (B) Interactive network container (pan and zoom the network), (C) Knowledge network rendering (displays nodes and edges of different types). (D) Interactive legend (to overlay more linked entities to the visible network). (E) Network summary statistics (indicates hidden and total nodes/edges). (F) Touch-sensitive context menu (show further information about selected node/edge, hide selected entity or its label, show hidden linked entities). (G) Item Information panel (display content-rich attributes of selected node/edge).\n\nVisualisation of knowledge networks needs to consider two key criteria: i) the heterogeneous and interconnected nature of the network and ii) the content-rich attributes of nodes and edges that cannot be easily displayed as part of the actual network.\n\nNodes in a biological knowledge network represent entities such as genes, proteins, phenotypes, pathways, publications and ontology terms; connected by edges of various types such as “encodes”, “published_in” and “ortholog” (Figure 1C). KnetMaps visualises each node type using a customized combination of shape and colour. Edge types are rendered using a combination of distinct size and colour attributes. The position of nodes and the length of edges is calculated using a force-directed layout that enables connectedness, separation and pattern-based clustering of closely inter-linked entities (Dogrusoz et al., 2009). Labels can be added to nodes and/or edges to enable easier understanding of the underlying data.\n\nTo view the potentially rich set of key-value attributes on nodes and edges, we have developed the Item Information panel (Figure 1G). It displays all textual (e.g. abstract, title of a publication) and numeric properties (e.g. accessions, scores and weights) of a selected node or edge in table format, including annotations, detailed descriptions, secondary labels and links to external websites and databases about the selected entity. Users can also use the information displayed in this panel to customize the rendered visualization of node and edge labels to their needs.\n\nOn relevant devices, KnetMaps supports basic touch gestures such as tap, tap-hold, tap-drag and pinch and zoom for user interaction. Users can interact with individual nodes and edges in the rendered network by using standard mouse or touch gestures such as click or tap gesture on a specific node or edge to get a summary of its properties or use the mouse wheel or pinch gesture to manipulate the zoom settings on the network.\n\nUsers can right-click or tap-hold on a node or edge to activate a radial context menu (Figure 1F) that provides a range of easy-to-use mechanisms for exploring or manipulating the selected entity. Users can click or tap on a node or edge and view further information such as type, description and annotations, summarised in a dialog box. Users can also tap-drag individual nodes or edges to re-align them within the visible network or tap-hold and reposition the rendered network as a whole.\n\nThe visualized network can be further explored and exported (Figure 1A1) using a variety of menu functions. For example, networks can be exported from KnetMaps as images (in png format) or as cytoscape-compatible JSON which can then be opened in the Cytoscape desktop application (Kohl et al., 2011) for further downstream analysis.\n\nKnetMaps controls information overload in the visible network by providing means to overlay data and extend it in incremental steps, thereby adopting a progressive approach where a subnetwork of interest from the underlying knowledge network is initially visualized and end-users are given the means to add more related information to the visible network. The subnetwork of interest is determined by the application using KnetMaps and passed to it through a “display” attribute in the API/JSON. KnetMaps generates a summary of the number of visible/hidden entities in the knowledge network so that end-users have an overview of what information might be present in the knowledge network but is currently hidden in the visible network. This information is automatically updated each time the user reveals or hides entities from the visible network (Figure 1E).\n\nThe first way of adding additional information to the network is by using the interactive legend that gives a summary of all node types present in the network, along with a numerical count of the total number of nodes of each type (Figure 1D). For example, clicking on a “Publication” symbol in the legend will add publications linked to visible gene and protein nodes, thereby enabling users to expand the visible knowledge network in real-time.\n\nThe second way to add or hide information is by using the context menu (Figure 1F). It allows users to hide individual nodes and edges, or hide all nodes and edges of a particular type. This can be useful for removing irrelevant or noisy information from the visible network. Additionally, it allows in- and out-going relations to be added to a selected node when these were initially hidden. This can be useful when a node acts as a knowledge hub, but only a small subset is initially visualized to intentionally prevent information overload, or if a node is part of a larger, more intricate knowledge pathway. In such cases, users can rapidly overlay connected entities within the selected node’s neighbourhood onto the visible network to effectively connect the dots and explore the myriad relationships between the network entities.\n\n\nImplementation\n\nKnetMaps leverages CytoscapeJS v.2.4.7 and jQuery v1.11.2 to visualise knowledge networks. It has been designed in a modular fashion and made available in NPM and BioJS, making it a reusable plug-and-play component within dynamic web applications.\n\nKnetMaps loads JSON input data (streamed or locally stored) and renders it as a knowledge network. It uses the cytoscapeJS JSON format specification in which the network is modeled as nested “nodes” and “edges” array objects. Each node or edge entity has a set of required properties such as colour, shape, size, identifier, label, border and visibility. We have extended the cytoscapeJS schema with an additional JSON object to store optional node/edge properties, e.g. abstract and title of a “Publication” node. The separation of required visual properties and optional data specific information, provided a more efficient way of rendering the general network while displaying node/edge specific information on demand, e.g. when the user clicks on a node.\n\nNetworks are rendered using a cytoscapeJS-based network stylesheet that maps the set of required JSON properties to the network object. The KnetMaps generator stylesheet sets the shape and colour of a node based on parameters provided for it in the JSON input dataset, e.g., ‘shape: data(conceptShape)’ and ‘background-color: data(conceptColor)’ where ‘conceptShape’ and ‘conceptColor’ are properties with set values in the input dataset. Developers can customize the stylesheet to replace the supported static cytoscapeJS shapes (such as triangle, round-rectangle, ellipse, pentagon and star) with images. CytoscapeJS selectors have been incorporated in the network stylesheet to filter nodes and edges based on these interactions and add functions that toggle their visual attributes such as highlighting a node or edge when selected and toggling visibility of labels accordingly.\n\nKnetMaps provides various features for interactive and incremental data exploration by incorporating useful Javascript libraries, such as the cytoscape.js-cxtmenu widget and various force-directed layout libraries to render the knowledge network, including the CoSE layout (Dogrusoz et al., 2009), which is the default network layout used in KnetMaps. Other layouts that can be used by end-users include the physics-based force layout, the CoSE-Bilkent layout that provides additional network topology and geometrical constraints, or static in-built cytoscapeJS layouts, such as the pattern-based circular layout or the concentric layout. KnetMaps packages cytoscapeJS-compatible extensions to these layouts within the application distribution and incorporates optimised settings for each layout within the application itself.\n\nNetworks of up to 1000 nodes and up to 3000 edges can be visualized in KnetMaps without significant performance degradation or visual delay in layout animations. Visualizing much larger networks (i.e. networks with over 10,000 nodes) increases the initialisation time and can cause jerky or delayed layout animation effects. Some of the rich visual styles used by KnetMaps can be somewhat expensive to render by cytoscapeJS, for example, rendering bezier curved edges.\n\nThe KnetMaps code addresses this by providing developers with flexible options to reduce the rendering complexity of the networks. All visual display settings have been made fully customizable to allow developers to tweak element styles such as node shape, edge curve and node border. Network container settings such as pixel ratio and motion blur can also be similarly easily altered, as can layout parameters such as reducing animation time, decreasing the number of layout iterations to run and disabling animation when rendering very large networks. The default parameters and settings work well in KnetMiner, based on the average sizes of the biological networks (between 300–1000 nodes) that it visualizes. However, customizing these parameters to employ simpler visual settings for larger networks can mitigate performance degradation during rendering.\n\nKnetMaps has been published to NPM and the BioJS (Gómez et al., 2013) registry, which provides a centralized portal of JavaScript tools and widgets used to analyse and visualize biological data, making it easy for research software users to install KnetMaps and embed within the HTML of their own web pages. The minimum system requirement is a PC with npm (part of Node.js) installed, a modern web browser with JavaScript enabled and a JSON sample file (see knetmaps/sampleFiles). Once bower for managing front-end components and gulp for bundling and distribution have been installed to the system (npm install bower gulp), following steps are needed to install and bundle KnetMaps:\n\n\n\nThe process will take a few minutes and create the dist/ folder, containing further img/, js/, and css/ subfolders which need to be copied to a web page’s root directory. Now, a simple index.html can be created to load and visualise the sample JSON dataset:\n\n\n\n\nUse cases\n\nKnetMaps is used as a network visualization component within tools and platforms that visualize biological knowledge as an interactive network, such as the KnetMiner (Hassani-Pak, 2017) and Daisychain. In KnetMiner, there is a need to visualise query related subsets of a genome-scale knowledge network (Hassani-Pak et al., 2016) in the web-browser. KnetMaps is one of the key components in KnetMiner to visualize and explore integrated information of inter-linked biological entities and processes to help in hypothesis generation and validation, and to accelerate candidate gene discovery. KnetMaps is also part of the KnetMiner Web API and therefore enables collaborators to view knowledge networks for specific genes and keywords from their own applications.\n\nDaisychain is a web application that links genome annotations, aiming to enable researchers working on certain species genes to investigate homologs in other published assemblies via a web interface called Daisychain-Web. The application can be queried using keywords or FASTA sequences with statistical cut-offs and the search results, i.e., links between genes and annotations across similar or identical species or cultivars are visualized as a network using KnetMaps. Daisychain uses KnetMaps out-of-the-box for rendering and visualization and adds further annotation and filtering options to the Item Information panel.\n\n\nConclusion\n\nVisualizations are a useful mechanism employed in many disciplines to present information in an intuitive representation that enhances user cognition and helps identify unique patterns and important trends in data. Network formalisms are becoming an increasingly popular means to combine data from inter-connected sources into a concise representation for easier and intuitive exploratory analysis. KnetMaps has been implemented as a fast and lightweight touch-friendly tool for visualizing content-rich, heterogeneous knowledge networks. The implementation uses cytoscapeJS, jQuery and JavaScript extensions for interactive functionality to ensure that low-memory, touch-compatible networks can be rendered in web browsers without the need to write extensive and unwieldy server-side code. Usage of JavaScript ensures rendering compatibility with most web browsers without the need to install any additional software, e.g., Java Applet or Adobe Flash. KnetMaps provides an interactive means to display, filter and overlay networked knowledge, and visually traverse the relationships connecting information within the rendered network. It incorporates a host of visualization techniques such as juxtaposition and superposition to encourage a step-by-step exploration of larger volumes of disparate data, thereby enabling end-users to investigate and analyse inter-linked knowledge in an incremental and intuitive manner.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.\n\n\nSoftware availability\n\nLatest source code available from: https://github.com/Rothamsted/knetmaps.js.\n\nPackaged distribution at BioJS: http://biojs.net/#/component/knetmaps.\n\nPackaged distribution at NPM: https://www.npmjs.com/package/knetmaps.\n\nArchived source code as at the time of publication: https://zenodo.org/record/1434144 (Singh & Hassani-Pak, 2018).\n\nExample demonstration of KnetMaps: http://knetminer.rothamsted.ac.uk/KnetMaps/.\n\nLicense: GNU General Public License v3.0.", "appendix": "Grant information\n\nThis work was funded by the Biotechnology and Biological Sciences Research Council (BBSRC) grants Designing Future Wheat (DFW) (BB/P016855/1) and DiseaseNetMiner (BB/N022874/1).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe are grateful to Max Franz for fruitful discussions on cytoscape.js integration, network rendering and other user interaction elements.\n\n\nReferences\n\nBecker RA, Eick SG, Wilks AR: Visualizing Network Data. IEEE Transactions on Visualization and Computer Graphics. 1995; 1(1): 16–28. Publisher Full Text\n\nBenedetti R, Cranley R: Head First JQuery. O’Reilly Media, Inc. 2011. Reference Source\n\nDogrusoz U, Giral E, Cetintas A, et al.: A Layout Algorithm for Undirected Compound Graphs. Inf Sci. 2009; 179(7): 980–94. Publisher Full Text\n\nEhrlinger L, Wöß W: Towards a Definition of Knowledge Graphs. Semantics and Pragmatics. 2016. Reference Source\n\nFranz M, Lopes CT, Huck G, et al.: Cytoscape.js: a graph theory library for visualisation and analysis. Bioinformatics. 2016; 32(2): 309–11. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGómez J, García LJ, Salazar GA, et al.: BioJS: an open source JavaScript framework for biological data visualization. Bioinformatics. 2013; 29(8): 1103–4. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nHassani-Pak K: KnetMiner-An Integrated Data Platform for Gene Mining and Biological Knowledge Discovery. 2017. Reference Source\n\nHassani-Pak K, Castellote M, Esch M, et al.: Developing integrated crop knowledge networks to advance candidate gene discovery. Appl Transl Genom. 2016; 11: 18–26. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nKöhler J, Baumbach J, Taubert J, et al.: Graph-based analysis and visualization of experimental results with ONDEX. Bioinformatics. 2006; 22(11): 1383–90. PubMed Abstract \| Publisher Full Text\n\nKohl M, Wiese S, Warscheid B: Cytoscape: software for visualization and analysis of biological networks. Methods Mol Biol. 2011; 696: 291–303. PubMed Abstract \| Publisher Full Text\n\nPavlopoulos GA, Secrier M, Moschopoulos CN, et al.: Using graph theory to analyze biological networks. BioData Min. 2011; 4(1): 10. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSingh A, Hassani-Pak K: Rothamsted/knetmaps.js: v1.0.1 (Version v1.0.1). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1434144\n\nSun Y, Han J: Mining Heterogeneous Information Networks: Principles and Methodologies. Morgan & Claypool Publishers. 2012. Publisher Full Text\n\nSzell M, Lambiotte R, Thurner S: Multirelational organization of large-scale social networks in an online world. Proc Natl Acad Sci U S A. 2010; 107(31): 13636–13641. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTaubert J, Hassani-Pak K, Castells-Brooke N, et al.: Ondex Web: web-based visualization and exploration of heterogeneous biological networks. Bioinformatics. 2014; 30(7): 1034–35. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVehlow C, Kao DP, Bristow MR, et al.: Visual analysis of biological data-knowledge networks. BMC Bioinformatics. 2015; 16: 135. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWang L, Alexander CA: Big Data in Design and Manufacturing Engineering. Am J Eng Appl Sci. 2015; 8(2): 223–32. Publisher Full Text" }	[ { "id": "39702", "date": "01 Nov 2018", "name": "Ian Dunham", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a useful article and software package to help apply cytoscape.js to heterogeneous knowledge graphs. KnetMaps.js adds several features surrounding the cytoscape.js graph visualization, such as the legend, information panel and download functionality. Following the setup guide to run the sample application locally was fairly straightforward.\nThe software is available online to test at both http://knetminer.rothamsted.ac.uk/KnetMaps/ and http://daisychain.appliedbioinformatics.com.au/. Developers can obtain the source through npm and biojs, where the package is named knetmaps. KnetMaps.js requires that input data be in a specific format and there are examples of this in the setup guide. The format is easy to understand, but combines data and styling information per node or edge. Separation of these concerns would add clarity and conciseness.\nA user wishing to display heterogeneous data, i.e. data containing different node and edge types, might find KnetMaps.js could save them development time. Several features that one might want to add on top of cytoscape.js are provided out of the box, such as the interactive legend, node/edge information panel, PNG/JSON export functionality, filtering and a variety of layout algorithms. However, for any substantial deviation from the UI design provided, a developer might be tempted to start with cytoscape.js directly, which has many online examples and is well documented.\nMinor revisions:\nSoftware\n\nPNG export. This functionality does not currently include the interactive legend, which would be useful to explain the static image. File extensions. The sample files have misleading file extensions. For example, ara2.json, which has JavaScript content, should really be a .js file. Automated build step. The running of gulp optimize could be done prior to publishing on npm. This would make it even easier for an end user to embed the necessary .js and .css files. Descriptions for example datasets. There are several example datasets available at http://knetminer.rothamsted.ac.uk/KnetMaps/. It would be helpful to provide a brief description of where the datasets came from and what they contain.\n\nManuscript\n\nUse of bower. I did not need to use bower to set up the sample application and would suggest removing the reference in the paper.\n\nLonger term suggestions for improvement:\nAdd some unit tests. This would also double as documentation for developers. Use npm to install third party libraries. Several libraries, such as cytoscape, cytoscape-cose-bilkent are not installed using npm. Upgrading them would be more straightforward if the package manager were used.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [] }, { "id": "42477", "date": "24 Jan 2019", "name": "Chia-Yi Cheng", "expertise": [ "Reviewer Expertise bioinformatics software user and tester" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis article describes a package KnetMaps that provides a JavaScript-based tool to visualize content-rich biological data. KnetMaps is suited for a website with heterogeneous biological information by integrating results in a network format. The examples listed on the demo page (http://knetminer.rothamsted.ac.uk/KnetMaps/) allows users to taste the flavor of KnetMaps. The interface is intuitive and straightforward. Below are suggestions to further enhance the clarity and completeness of the user experience:\nIt would be beneficial for both the developers and end users if the author could provide a step-by-step guide to reproduce one example dataset as on the demo page. The ‘interactive legend’ feature allows users to add an additional layer to the default network. It is not completely clear to me: i) how was the visible/invisible edges/nodes set in the first place? Ii) is it possible to remove the once added information? For example, clicking on a ‘Domain’ symbol in the legend will add ‘Domains’ linked to visible nodes. A user may review and find the information not needed and want to hide those edges. I may have missed it but I did not find a way to cherry pick the symbols once the associated edges displayed.\n\nThe PNG is working on the demo page (http://knetminer.rothamsted.ac.uk/KnetMaps/) but not the use case pages (http://knetminer.rothamsted.ac.uk/Zea_mays/) (http://daisychain.appliedbioinformatics.com.au/). A blank window popped up when I hit the PNG icon.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [] }, { "id": "42478", "date": "25 Jan 2019", "name": "Nicholas J. Provart", "expertise": [ "Reviewer Expertise Data visualization", "bioinformatics", "gene regulatory network analysis", "interaction network visualization", "cyberinfrastructure" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nIn general the concept of the KnetMaps BioJS component is very welcome. I like the functionality that the authors have developed (at least as described in the paper), but I think they could take it up a notch to use additional “visualization techniques to prevent information overload”, especially taking advantage of Cytoscape.js’s ability to group nodes via the compound node feature – this way, different types of nodes could be grouped (which might be useful for larger networks where things might become quite overloaded). Nodes and edges might receive some additional treatment.\n\nThe authors write: “Developers can customize the stylesheet to replace the supported static cytoscapeJS shapes (such as triangle, round-rectangle, ellipse, pentagon and star) with images.” It would be nice to actually include a set of image-based icons for the different nodes in the KnetMaps release. Given the multiple types of nodes possible, perhaps a node icon that looks like a page with a few lines on it could symbolize publications, a node icon with an stylized alpha helix could represent the protein etc. I’d also explore using different kinds of edges (straight, curved, dashed) instead of a just a colour scheme.\nSample.json did not seem to be in NPM (as described in the code in the article)? I tried getting KnetMaps to work within Codepen.io and unkpg.com/knetmaps/ but wasn’t quite able to. It would be nice to have this set up so readers could try out the package simply within a browser (by making the “pen” public and sharing the link in the F1000 article). It was almost working, except the JSON file (e.g. Arabidopsis flower network) couldn’t be loaded because of a MIME type error and even when I included the JSON object in the script it wasn’t working (see code here). Another reason for providing a Codepen link is that there is no preview on the BioJS site.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Partly", "responses": [] }, { "id": "42475", "date": "29 Jan 2019", "name": "Ramil Mauleon", "expertise": [ "Reviewer Expertise Bioinformatics", "molecular genetics", "genomics" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe functionality provided by KnetMaps is very useful for molecular geneticists (and the bioinformaticians who support them), whose aim is to identify the best candidate genes to use in trait improvement (I am coming from the bioinformatician/geneticist side). Using heterogenous evidences (omics experiment results, literature, function by association inference from sequence identity, GWAS results) is a common exercise in order to get a more informed list of candidate genes to work with. This is the strength of the visualization of Knetmaps: the ability to display the evidences incrementally, and to visualize the connection of evidences into an intuitive network display, which is very helpful for biologists in order to make decisions on which genes to select for further (expensive) experimental validation. Network rendering and response is fast, thus providing an enjoyable end user experience. I do share one reviewer's feedback on how to revert the network view back to the original state prior to clicking on an icon in the interactive legend, I also seem to miss the step on how to do this.\nA suggestion on the implementation section, it would be nice to have a graphical block diagram that shows the steps and inputs required for KnetMaps to be installed in an end-user's system. A statement reporting on installation and functionality of the system under 3 dominant OSs would be appreciated (Linux, Unix/MacOS, MSWindows?), as upfront knowledge for users who wish to install this in their own systems.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1651
https://f1000research.com/articles/7-1650/v1	17 Oct 18	{ "type": "Research Article", "title": "Metagenomic evaluation of a Utah tar sand microbiota suggests the predominant hydrocarbonoclastic role of Actinobacteria", "authors": [ "Dawn E. Lewis", "Ashish Pathak", "Cynthia B. Jones", "Charlemagne Akpovo", "Ashvini Chauhan", "Dawn E. Lewis", "Ashish Pathak", "Cynthia B. Jones", "Charlemagne Akpovo" ], "abstract": "Background: Occurrences of tar sands have been reported in 22 states in the United States; however, the largest deposit is located in southwestern Utah. It has been suggested that tar sands were created by the microbial degradation of immobile subsurface oil over several million years; however, little is known about the indigenous microbial communities in the bituminous tar sands. Methods: This study identified Utah tar sand microbiota using next-generation sequencing technology and characterized the functional diversity using community-level physiological profile (CLPP). Results: Microbiota identified in these tar sands are mainly affiliated with the Gram-positive Actinobacteria and representatives of genera that have also been previously shown to degrade aromatic hydrocarbons, including Arthrobacter, Dietzia, Janibacter, Nocardioides, Microbacterium, Agrococcus and Salinibacterium, suggesting that these microbes likely play roles in the biodegradation of oil-hydrocarbons. CLPP analysis revealed less than 24 h was needed for the first color development in the microplate wells containing the polymers, whereas the duration of the lag phase of the carboxylic acids was prolonged. Conclusions: The quick utilization of the polymers suggests that the indigenous microbial community, especially the actinomycetes in the tar sand habitat, are poised and primed to degrade these recalcitrant compounds.", "keywords": [ "Tar", "Sand", "Microbiota" ], "content": "Introduction\n\n‘Tar sands’ is a term used to describe a combination of bituminous sand, sandstone, oil-impregnated rock, oil sand and rock asphalt1. The tar sands is a rapidly developing source of unconventional petroleum. The bitumen, an oil-rich residue, can be extracted from the tar sands and refined into crude oil. In the United States, occurrences of tar sands have been reported in 22 states; however, one of the largest deposit is located in southwestern Utah, with estimated recoverable oil ranging from 12 to 20 barrels2.\n\nIt has been suggested that tar sands were created by the microbial degradation of immobile subsurface oil over several million years3. Little is known about the origin of the microorganisms in the tar sands of Utah, although the indigenous microbial communities exist in extreme conditions in the bituminous tar sands and are limited by harsh conditions such as low moisture and oxygen, recalcitrant hydrocarbons and a high concentrations of toxic metals4. However, a paucity of data exists on the indigenous microbial communities of the tar sands of Utah.\n\nTraditionally, the diversity of the bacterial communities in the tar sands has been investigated using the isolation and cultivation approaches5. However, only 1.5% of bacteria in the soil can be readily cultured6,7. Most cultural conditions cannot mimic the specific microhabitats that many prokaryotes thrive within. Therefore, new methods for the analysis of whole microbial community structure and metabolic function in the bituminous tar sands have been developed. One of the methods developed to provide a dynamic tool for the assessment of microbial community structure and function is the BIOLOG™ Community Level Physiological Profile (CLPP). This method is based upon the preferential metabolism of 31 different carbon sources on a microtiter plate. These carbon sources include a wide range of chemical classes, such as carbohydrates, esters, polymers, carboxylic acids, alcohols, amides, phosphorylated chemicals, amino acids, aromatic chemicals, and amines8. Each well on the BIOLOG™ Ecoplate contains a unique carbon source, peptone, and a 2,3,5-triphenyltetrazolium chloride (TTC) dye. NADH is produced from the respiration of the specific carbon sources. The final electron acceptor TTC is irreversibly reduced to formazan, a red pigment that can be quantified visually by use of a microplate reader8,9. The intensity of the color change correlates to the amount of metabolism of the carbon source in that well. The net intensity of the color change is calculated by subtracting the absorbance of the non-carbon-source control well. The oxidation of the carbon substrate oxidized by the microbe can be considered to be its metabolic fingerprint. However, Windig10 reported that the BIOLOG™ system excludes strict anaerobes and bacteria that lack certain electron transport enzymes. Yao et al.11 advocated that CLPP is selective and favors microbes that grow quickly or those with a high inoculum density in the initial sample. Hence, culture-independent analysis is the method of choice for the investigation of the bituminous tar sands.\n\nThe objective of this study was to identify and characterize the indigenous bacterial communities of the Utah tar sands using next-generation sequencing technology and to assess the functional diversity using the CLPP system. The overall goal of this project is significant because studies have demonstrated that naturally occurring microbes can be harnessed for the degradation of recalcitrant polyaromatic hydrocarbons, heavy metals, and naphthenic acids. The combination of the functional diversity and the characterization of the indigenous microbiota will advance our understanding of the fate of tar-associated potentially toxic compounds by environmental microbiota.\n\n\nMethods\n\nThe University of Alberta Geotechnical Centre, Canada, kindly provided the tar sands sample. All samples were shipped on ice to Florida A&M University where they were processed for metagenomics and CLPP analyses. Genomic DNA was extracted from the samples using the PowerSoil DNA Isolation kit (MoBio Laboratories, Inc., Carlsbad, CA, USA). Bar-coding pyrosequencing of the 16S rRNA gene V4 region was performed using the Earth Microbiome Project (EMP) standard protocols (http://www.earthmicrobiome.org/emp-standard-protocols) and sequenced on a Roche 454 FLX instrument (Roche, Indianapolis, IN, USA) with titanium reagents, following manufacturers recommended procedures. Sequences that passed quality controls were uploaded to MOTHUR v.1.33.3, where tags, low-quality sequences and chimeric reads were removed.\n\nCLPP of the tar sand sample was analyzed using BIOLOG™ Ecoplates, as described previously8,12. Triplicate samples of the BIOLOG™ Ecoplate were used for this experiment. A total of three sets of 1 g samples of the tar sand were shaken in 99 ml of distilled sterile water for 20 min at 20°C. The samples were then incubated at 4°C for 30 min and then centrifuged for 10 min at 500g. The supernatant of the tunes were pooled and thoroughly mixed. In total, 150 μl of the supernatant was inoculated into each well of the Biolog EcoPlates. This ensured that each of the 96-well plate contained three replicates of each sample exposed to 31 different carbon sources. Water blanks were included with each plate and the plates were incubated at 25°C without shaking. Substrate utilization was monitored by measuring absorbance at 595 nm using a BIOTEK UQuant Microplate Spectrophotometer, (BIO-TEK Instruments, Inc., USA). The first measurement was made immediately after inoculation and subsequent readings were taken after every 24 h interval for 7 days. The measurements of individual substrates were corrected for background absorbance by subtracting the absorbance of the control (water) samples. If a negative number was obtained, it was manually set to zero. A well was considered positive if the mean OD595nm exceeded that of 0.25.\n\nIn this study the 31 carbon sources were organized into groups 1–5 as described by Zak et al.13. Carbon sources originally grouped as miscellaneous by Zak et al.13 or those new to the BIOLOG ™ EcoPlate, were grouped into one of the other five categories - carbohydrates, carboxy acids, amino acids, esters and polymers. Grouping the data into 5 guilds compresses a 31-dimensional data set into 5 dimensions, significantly reducing the complexity of the data and subsequent interpretation14.\n\nThe average well color development (AWCD) was used to assess the microbial response for all the carbon sources. The AWCD was determined as follows:\n\n\n\nwhere ODi is optical density value from each well, corrected subtracting the blank well (inoculated with water) values from each plate well8,15. The 96-h data was used for statistical analysis of CLPPs.\n\nThe diversity of substrate utilization was calculated using the Shannon’s diversity index (H) and evenness (E)16. The Shannon-Weaver index is a measure of the capacity of the bacterial community to degrade the carbon sources in the well and can be considered as an index of the physiological diversity of the bacterial community in the tar sands. Higher values of H indicate the ability of the microbial community to degrade the substrates with a high efficiency16:\n\n\n\nwhere Pi = (OD reading of well I)/(sum of all wells)], based on the OD in the Ecoplates.\n\nThe Shannon evenness (E) is a measure of how dissimilar the abundances of the species in a community are from each other17 and was calculated as:\n\nE – H/lnS\n\nwhere S (Shannon richness) is the number of substrates used by the microbial communities. Further, the cluster analysis of the substrate utilization pattern was constructed using the nearest neighbor method with Euclidian distance to form linkage dendrogram.\n\n\nResults and discussion\n\nIn this study, the compositions of bacterial community of the tar sands was investigated by a pyrosequencing-based analysis of the 16S rRNA gene sequences. The gram-positive Actinobacteria (50%) were dominant in the tar sand followed by Betaproteobacteria (27%), Alphaproteobacteria (7%), Gammaproteobacteria (7%) and Acidobacteria (2%) (Figure 1). Actinobacteria are known for their role in the biodegradation of a variety of different pollutants including petroleum hydrocarbons18,19. The predominant genus identified in these tar sands was Arthrobacter, followed by Dietzia, Janibacter, Nocardioides, Microbacterium, Agrococcus and Salinibacterium (Figure 2). Most of these genera have also been previously shown to degrade aromatic hydrocarbons, indicating that tar sands are a very active repository of hydrocarbonoclastic microorganisms. When the taxonomic affiliation of the obtained metagenomic sequences were investigated, we found that the gram-positive Arthrobacter spp. from the Actinobacteria phyla were predominant, which comprised approximately half of the total microbial community assemblage in this particular Utah tar sand. Arthrobacter species possess a significant hydrocarbonoclastic potential as demonstrated by previous studies and we recommend that Actinobacteria, native to the tar sand habitats, should be targeted for future research on this area.\n\nThe gram-positive Actinobacteria were dominant in the tar sand followed by Betaproteobacteria, Alphaproteobacteria, Gammaproteobacteria and Acidobacteria.\n\nThe gram-positive Arthrobacter spp. from the Actinobacteria phyla were dominant in the tar sand microbiota.\n\nBiolog® Ecoplates were used to evaluate the functional diversity of the microbial community of the tar sands. The AWCD revealed sigmoidal relationships between the OD590 and time for all carbon sources (polymers, carbohydrates, amino acids, esters) except the carboxylic acids (Figure 3). This pattern for the polymers, carbohydrates, amino acids, and esters is fundamentally similar to the bacterial growth curve. Yao et al20 suggested that color development reflected species metabolic activity and the ability of the bacterial community to respond to substrates. The AWCD for all the substrates in the Biolog® Eco-microplates was 0.25. Less than 24h was needed for the first color development the microplate wells containing the polymers whereas the duration of the lag phase of the carboxylic acids was prolonged. The first color development of the carboxy acids (AWCD0.2) was recorded at 48 h (Figure 3). The highest utilization was observed in the microplate wells containing the polymers (AWCD0.78) after 96 hours. The observed differences in the substrate utilization pattern of five major carbon groups (Figure 3) might be due to the presence of different functional groups, such as carbohydrates R-C=O), amino acids (-NH2 and -COOH), carboxylic acid (-COOH), esters, amines and phosphorylated acids (-COOR’, -NH2), and polymers (–(CH2-CH2)–(R)n). The high utilization of the recalcitrant polymers suggest that the polymers could be more easily utilized by the indigenous microbial community in the tar sands. After 96 h, there was a decrease in the rate of AWCD; thus, the analysis of functional diversity of microbial community of tar sands was completed at 96 h. The decreased rate may be due to the change of metabolic activity of active microbial communities utilizing the C-substrates. Cluster analysis of the substrates revealed a systematic grouping of the substrates based on their utilization pattern (Figure 4).\n\nThe Shannon-Wiener index is an indication of the spread or distribution of carbon source utilization by the microbial community21. Typical values of the Shannon-Wiener Index usually lies between 1.5 and 3.5, and rarely exceed 4.022. In this study, the Shannon-Weiner index, H, of the microbial community in the tar sands was 3.072 and the Shannon evenness, E, was 1.009. These values indicate a more diverse microbial community and an even distribution of the species within the tar sands.\n\n\nConclusion\n\nThis study investigated the indigenous microbial communities of the tar sands of Utah in an effort to understand the structure and the functional diversity of the microbial community. The substrate utilization patterns of resident microbial communities and the identification of the hydrocarbonoclastic microorganisms in the tar sands provide valuable baseline information than can be used for hydrocarbon bioremediation and for devising biotechnological approaches to tar sands bioremediation efforts.\n\n\nData availability\n\nDataset 1. Raw data obtained from the community-level physiological profile, 10.5256/f1000research.16126.d21890923.\n\nThe DNA sequences from this metagenomic project are available from the Sequence Read Archive/European Nucleotide Archive, accession number SRR1699470.", "appendix": "Grant information\n\nThis work was partially supported by the Department of Energy (DOE) Minority Serving Institution Partnership Program (MSIPP) managed by the Savannah River National Laboratory under SRNS contract DE-AC09-08SR22470 (Task Order Agreements [TOA] 0000272350 and 0000332983).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nThe University of Alberta Geotechnical Centre, Canada, is acknowledged for the provision of the tar sand samples.\n\n\nReferences\n\nSpieker EM: Contributions to economic geology (short papers and preliminary reports), 1930, Part II, Mineral fuel. Bituminous sandstone near Vernal, Utah. U.S. Geological Survey Bulletin. 1930; 822: 77–98. Publisher Full Text\n\nBaxter SR: Tar Sands: Worth the Energy? An Analysis of the Future of Utah's Tar Sands. J Land, Resources & Envtl L. 2007; 27(2): 323–344. Reference Source\n\nHarner NK, Richardson TL, Thompson KA, et al.: Microbial processes in the Athabasca Oil Sands and their potential applications in microbial enhanced oil recovery. J Ind Microbiol Biotechnol. 2011; 38(11): 1761–75. PubMed Abstract \| Publisher Full Text\n\nKim JS, Crowley DE: Microbial diversity in natural asphalts of the Rancho La Brea Tar Pits. Appl Environ Microbiol. 2007; 73(14): 4579–91. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDemeter MA, Lemire J, George I, et al.: Harnessing oil sands microbial communities for use in ex situ naphthenic acid bioremediation. Chemosphere. 2014; 97: 78–85. PubMed Abstract \| Publisher Full Text\n\nAmann RI, Ludwig W, Schleifer KH: Phylogenetic identification and in situ detection of individual microbial cells without cultivation. Microbiol Rev. 1995; 59(1): 143–169. PubMed Abstract \| Free Full Text\n\nHead IM, Saunders JR, Pickup RW: Microbial Evolution, Diversity, and Ecology: A Decade of Ribosomal RNA Analysis of Uncultivated Microorganisms. Microb Ecol. 1998; 35(1): 1–21. PubMed Abstract \| Publisher Full Text\n\nGarland JL, Mills AL: Classification and characterization of heterotrophic microbial communities on the basis of patterns of community-level sole-carbon-source utilization. Appl Environ Microbiol. 1991; 57(8): 2351–2359. PubMed Abstract \| Free Full Text\n\nBochner BR: Sleuthing out bacterial identities. Nature. 1989; 339(6220): 157–158. PubMed Abstract \| Publisher Full Text\n\nWinding A: Fingerprinting bacterial soil communities using Biolog microtitre plates. In: Ritz K, Dighton J, Giller K (eds.) Beyond the Biomass, Wiley and Sons, Chichester, 1994; 85–94. Reference Source\n\nYao H, He Z, Wilson MJ, et al.: Microbial Biomass and Community Structure in a Sequence of Soils with Increasing Fertility and Changing Land Use. Microb Ecol. 2000; 40(3): 223–237. PubMed Abstract\n\nInsam H: A new set of substrates proposed for community characterization in environmental samples. In: Insam, H., Rangger, A. (Eds.), Microbial Communities. Springer-Verlag, Berlin/Heidelberg, Germany, 1997; 259–260. Publisher Full Text\n\nZak JC, Willig MR, Moorhead DL, et al.: Functional diversity of microbial communities: A quantitative approach. Soil Biol Biochem. 1994; 26(9): 1101–1108. Publisher Full Text\n\nWeber KP, Legge RL: One-dimensional metric for tracking bacterial community divergence using sole carbon source utilization patterns. J Microbiol Methods. 2009; 79(1): 55–61. PubMed Abstract \| Publisher Full Text\n\nWeber KP, Grove JA, Gehder M, et al.: Data transformations in the analysis of community-level substrate utilization data from microplates. J Microbiol Methods. 2007; 63(3): 461–9. PubMed Abstract \| Publisher Full Text\n\nHan XM, Wang RQ, Liu J, et al.: Effects of vegetation type on soil microbial community structure and catabolic diversity assessed by polyphasic methods in North China. J Environ Sci (China). 2007; 19(10): 1228–1234. PubMed Abstract \| Publisher Full Text\n\nSmith B, Wilson J: A consumer’s guide to evenness indices.Oikos. 1996; 76(1): 70–82. Publisher Full Text\n\nYeung CW, Woo M, Lee K, et al.: Characterization of the bacterial community structure of Sydney Tar Ponds sediment. Can J Microbiol. 2011; 57(6): 493–503. PubMed Abstract \| Publisher Full Text\n\nPathak A, Green SJ, Ogram A, et al.: Draft Genome Sequence of Rhodococcus opacus Strain M213 Shows a Diverse Catabolic Potential. Genome Announc. 2013; 1(1): pii: e00144-12. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYao HY, Xu JM, Huang CY: Substrate utilization pattern, biomass and activity of microbial communities in a sequence of heavy metal-polluted paddy soils. Geoderma. 2003; 115(1–2): 139–148. Publisher Full Text\n\nHarch BD, Correll RL, Meech W, et al.: Using the Gini coefficient with BIOLOG substrate utilisation data to provide an alternative quantitative measure for comparing bacterial soil communities. J Microbiol Methods. 1997; 30(1): 91–101. Publisher Full Text\n\nMay RM: Patterns of species abundance and diversity. In Ecology and Evolution of Communities (ed. M. L. D. Cody, J. M.). Cambridge, Mass.: Harvard University Press. 1975. Reference Source\n\nLewis DE, Pathak A, Jones CB, et al.: Dataset 1 in: Metagenomic evaluation of a Utah tar sand microbiota suggests the predominant hydrocarbonoclastic role of Actinobacteria. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16126.d218909" }	[ { "id": "41171", "date": "18 Feb 2019", "name": "Chunkai Huang", "expertise": [ "Reviewer Expertise Biological wastewater treatment", "biofilms", "oil sands process-affected water (OSPW)" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis manuscript documented microbial communities in Utah tar sand. The methods are well described and results are clearly presented. The performance of these communities to consume 31 different carbon sources was compared. Conceptually, I like the idea to compare the degradation rate of 31 different carbon sources using Utah tar sand indigenous microbial communities at micro plate. However, I believe the microbial community characterization, especially the high-throughput data, are not vigorously examined. I have a few minor concerns as follows.\n\nMajor concerns:\n\n1. As you mentioned in the Introduction part, the BIOLOG system excludes strict anaerobes and bacteria that lack certain electron transport enzymes. That means the indigenous microbes of Utah sand were selected after inoculation and cultivation at 96-well plates. Some of the Utah sand anaerobes and other microbes were inhibited. It is possible that those inhibited Utah sand anaerobes are efficient for the degradation of other recalcitrant compounds.\n\n2. It is necessary to mention the primer pairs and the regions of the 16S rRNA genes, because it is essential to ensure full reproducibility by others.\n\n3. All raw pyrosequencing data that was obtained from this study should be deposited to the NCBI Sequence Read Archive, and the accession no. should be provide in the method.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "60547", "date": "25 Mar 2020", "name": "Marilene H. Vainstein", "expertise": [ "Reviewer Expertise Biofilms", "oil degradation", "bioremediation." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe manuscript \"Metagenomic evaluation of a Utah tar sand microbiota suggests the predominant hydrocarbonoclastic role of Actinobacteria\" describes the Utah sand microbiota using NGS and CLPP. Although the experiments are well-designed, the authors should boost the \"Results and Discussion\" section to create a more informative manuscript, exploring better the obtained results. Furthermore, some minor corrections are needed:\nFigure 1: There are four Acidobacteria groups (Gp16, Gp7, Gp6, and Gp4). What is the difference between these groups?\nFigure 1: Figure and data explanation through the text does not match the figure data. For example, the authors described that \"The gram-positive Actinobacteria (50%) were dominant in the tar sand followed by Betaproteobacteria (27%), Alphaproteobacteria (7%), Gammaproteobacteria (7%) and Acidobacteria (2%)\", but in the figure, Actinobacteria correspond to 48% and Betaproteobacteria 26%, while Alphaproteobacteria 8%.\n\"Actinobacteria are known for their role in the biodegradation of a variety of different pollutants including petroleum hydrocarbons18,19.\": The authors can explore better this statement, maybe they can describe scientific and commercial applications of Arthrobacter, Dietzia, Janibacter, Nocardioides, Microbacterium, Agrococcus and Salinibacterium.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [] } ]	1	https://f1000research.com/articles/7-1650
https://f1000research.com/articles/7-99/v1	23 Jan 18	{ "type": "Research Article", "title": "A fruit fly model for studying paclitaxel-induced pain", "authors": [ "Zina Hamoudi", "Thang Manh Khuong", "Tiffany Cole", "G. Gregory Neely", "Zina Hamoudi", "Thang Manh Khuong", "Tiffany Cole" ], "abstract": "Background: Paclitaxel-induced peripheral neuropathy is a common and limiting side effect of an approved and effective chemotherapeutic agent. The cause of this nociception is still unknown. Methods: To uncover the mechanism involved in paclitaxel-induced pain, we developed a Drosophila thermal nociceptive model to show the effects of paclitaxel exposure on third instar larvae. Results: We found that paclitaxel increases pain perception in a dose-dependent manner, without overt morphological changes. Conclusions: Our simple, high throughput model can be combined with genomics approaches to identify regulators of chemotherapy-induced pain to eliminate its adverse side effects.", "keywords": [ "Drosophila", "fruit fly", "paclitaxel", "nociception", "pain", "CIPN" ], "content": "Introduction\n\nChemotherapy-induced peripheral neuropathy (CIPN) is a dose-limiting side effect of many effective cancer treatments (Burton et al., 2007), and can have a lasting impact on the quality of life of cancer survivors (Hausheer et al., 2006 and Shimozuma et al., 2012). A meta-analysis of 31 studies from over 4000 chemotherapy-treated patients revealed that CIPN was prevalent in 68.1% of patients in the first month following chemotherapy, in 60% of patients at 3 months, and in 30% at 6 months or more (Seretny et al., 2014).\n\nPaclitaxel has a potent ability to cause CIPN (Addington & Freimer, 2016; Reyes-Gibby et al., 2009). Derived from the bark of the western yew, Taxus brevifolia, it is an approved and effective treatment against breast, ovarian, lung and Kaposi sarcoma (Chang et al., 1993; Gill et al., 1999; Holmes et al., 1991; McGuire et al., 1989; Wani et al., 1971). Patients treated with paclitaxel experience side effects as early as one to three days following treatment (Lipton et al., 1989; Reyes-Gibby et al., 2009). Common symptoms are hyperalgesia, hypoalgesia, allodynia, tingling, numbness, and shooting pain (Boland et al., 2010). Paclitaxel has a direct effect on Schwann cells, promotes axonal degeneration, and can cause mitochondrial damage (André et al., 2000; Cavaletti et al., 1995; Sahenk et al., 1994), however the molecular mechanisms causing pain are still largely unknown.\n\nWhile much knowledge has been gained about the genetics of pain from vertebrate systems, high throughput dissection of pain is possible using the fruit fly Drosophila melanogaster (Neely et al., 2010). When challenged with a noxious thermal stimulus, third instar larvae exhibit an aversive escape response that has been utilised to identify conserved genes required for nociception (Babcock et al., 2009; Neely et al., 2010; Tracey et al., 2003). This nociceptive response is a result of activating class IV multidendritic-dendritic arborisation (md-da) sensory neurons at the site of stimulation (Hwang et al., 2007). Previously in Drosophila, paclitaxel has been reported to be toxic in somatic cells, and causes loss of axons in peripheral nerves. (Bhattacharya et al., 2012; Cunha et al., 2001). However, its effects on nociception have not yet been evaluated. Here, we examined the effects of paclitaxel exposure on the fruit fly larval nociception system, and observed a robust and dose-dependent increase in pain perception. This system is amenable to high throughput screening and genetic manipulation, and may help define why chemotherapies such as paclitaxel cause pain.\n\n\nMethods\n\nAll flies were reared at 25°C and 65% humidity over a 12-hour light-dark cycle. Six female and two male Canton S Drosophila melanogaster were mated on food medium (5.4% sucrose, 3.6% yeast, 1% agar, 1.2% nipagin, and 0.6% propionic acid) treated with ethanol, 0 µM, 0.1 µM, 0.5 µM, 2.5 µM, 5 µM or 10 µM paclitaxel (Taxol®; Catalog No. A4393) purchased from ApexBio (Houston, USA). A stock of 1000 µM paclitaxel in ethanol was prepared and diluted in food medium accordingly to create the different drug concentrated food. F0 Flies were discarded two days after mating and F1 larvae were left to grow for another three days. On the fifth day, early third instar were collected to assess nociception or dendritic morphology.\n\nFor the thermal nociceptive assay (Tracey et al., 2003), distilled water was added to experimental vials to soften the food and release the foraging third instar larvae. The softened, liquid food was then passed through mesh to catch the larvae to be transferred to a 100mm petri dish sprayed with distilled water. The larvae were touched laterally on abdominal segments four to six with a heat probe (soldering iron with narrow tip) set to 42°C or 46°C. The rolling response was measured in seconds with a cut-off of 10 seconds. For each drug concentration, five repeats were performed, with 30–40 larvae per repeat.\n\nThird instar larvae (w;ppk-CD4-tdGFP) were collected, washed, and placed onto a sylgard dissecting plate (Sylgard184 silicone elastomer kit, Dow Corning, Midland, MI). At least 10 larvae of each group were then dissected in HL-3 saline solution (70 mM NaCl, 5 mM KCl, 10 mM NaHCO3, 5 mM HEPES, 115 mM sucrose, 5 mM threalose, 20 mM MgCl2, pH 7.2). Dissected larvae were fixed with 4% paraformaldehyde solution, and then washed in PBS four times. Larvae of the same genotype were then transferred to microscope slides to be mounted using Vectashield (Vector Laboratories, Burlingame, CA).\n\nGFP-expressing class IV md-da sensory neurons were imaged on a Leica DMI-6000 inverted microscope and SP8 Basic Confocal system using 40x oil immersion lens (Leica, Wetzlar, Germany). Class IV neuron at abdominal segment 5 (A5) of each larva was imaged. 488 nm lasers for excitation were used, emission signal was captured at 500–550nm wave length, scanning speed of 400 Hz, zoom factor of 0.7, and gain of 600. Images of Z-stack sections were captured at 1024 × 1024 pixel resolution with section thickness of 0.4 µm. Z-stack images were then rendered into a maximum intensity projection using ImageJ. Total dendritic surface area was quantified using ImageJ. The experiment was conducted in a blinded manner.\n\nData are presented as mean ± SEM and compared to vehicle control. Analysis was done using GraphPad Prism 5. Statistical analysis for response time was done using Krustal-Wallis ANOVA, followed by Dunn’s pairwise test for multiple comparisons. Statistical analysis for dendritic area percentage was done using Student’s t-test.\n\n\nResults\n\nOur goal here was to develop a reproducible paradigm to investigate the effects of paclitaxel on nociception in the fly larvae. Based on previous studies for toxicity (Bhattacharya et al., 2012; Cunha et al., 2001), we selected paclitaxel doses below the lethal limit (Figure 1A), and then tested larval nociception using a heat probe set to a low intensity noxious heat (42°C; Figure 1B), which is mildly nociceptive to fly larvae (Babcock et al., 2009). Our dose-response study revealed 2.5 µM paclitaxel was sufficient to induce significant hyperalgesia, with a maximal hyperalgesia effect observed at 10 µM (Figure 1C). Concentrations higher than 10 µM paclitaxel were 100% lethal (not shown). Interestingly, paclitaxel did not significantly alter heat nociception latency to a 46°C heat stimulus across any of the doses (Figure 1D). Vehicle (ethanol) control and normal (no ethanol) control showed a response time of 5.71 sec (±0.23 SEM; n=173) and 5.62 sec (±0.20 SEM, n=180, not shown), respectively (42°C; Figure 1E). At low concertation’s of 0.1 µM (5.21 sec ± 0.23 SEM; n=150) and 0.5 µM (5.44 sec ± 0.26 SEM; n=131) paclitaxel did not affect response profiles, however, concentrations of 2.5 µM paclitaxel (4.22 sec ± 0.19 SEM; n=180; p<0.001) and higher altered response distribution and significantly enhanced nociceptive latency (42°C; Figure 1E). The fastest latency response was observed at 10 µM paclitaxel (3.84 sec ± 0.24 SEM; n=140; p<0.001) with a 36.6% increase in response time relative to vehicle control (Figure 1C).\n\nSchematic representation of the A) experimental design and B) thermal nociceptive assay in third instar Drosophila larvae. C–D) Average nociceptive latency (in seconds) in response to a 42°C or 46°C thermal stimulus, respectively. Increased paclitaxel concentration significantly induces heat-hyperalgesia in third instar larvae at 42°C. Note concentrations higher than 10 µM paclitaxel were 100% lethal. E) Percentage response to each time point in seconds to 42°C thermal stimulus. All values represent mean ± SEM. p values were generated using Krustal-Wallis ANOVA, followed by Dunn’s pairwise test for multiple comparisons. Significance is relative to vehicle control. Five repeats were performed for each drug concentration with roughly 30 larvae each (n = 130–180 animals).\n\nTo evaluate if paclitaxel exposure caused robust morphological differences in peripheral pain sensing neurons, we fed genetically labeled (w;ppk-CD4-tdGFP) larvae paclitaxel and imaged the sensory neuron structure (Figures 2A–B). Surprisingly, treating larvae with 10 µM paclitaxel did not affect dendritic area percentage compared to vehicle control (Figure C) despite significantly enhancing nociceptive sensitivity (not shown). Thus we establish that paclitaxel sensitizes larvae to heat pain via enhancing sensory neuron or higher order nociception, and not via inducing overt morphological changes.\n\nA–B) Confocal images of third instar ppk-CD4-tdGFP larvae following (A) vehicle control or (B) 10 µM paclitaxel treatment. Images are of class IV md-da neurons at abdominal segment A5. Images are at 40x magnification and 0.7 zoom factor. Scale bar represents 40 µm. C) Quantification of dendritic area percentage of class IV md-da sensory neurons. Values represent mean ± SEM (n = 10–11 animals).\n\n\nConclusions\n\nMany approved and effective anti-cancer therapeutic agents cause severe pain as a side effect. This can limit our ability to eradicate tumors, and often leaves cancer patients and survivors in intense, and untreatable pain. In this study, we describe a simple system to provide a high throughput dissection of the mechanisms involved in paclitaxel-induced pain. By combining these techniques with genomic approaches to identify regulators of chemotherapy pain, we can better understand how the pain arises, and potentially avoid these severe side effects while more effectively targeting the underlying disease.\n\n\nData availability\n\nDataset 1: Larval response time in seconds to 42°C heat stimulus. Paclitaxel fed larvae were touched with a 42°C heat probe and their response time was measured in seconds with a cut-off of 10 seconds. Different treatments were tested: food control, ethanol control, 0.1 µM, 0.5 µM, 2.5 µM, 5 µM, and 10 µM paclitaxel. Five repeats were performed (n = 130 - 180). DOI, 10.5256/f1000research.13581.d191022 (Hamoudi et al., 2018a).\n\nDataset 2: Larval response time in seconds to 46°C heat stimulus. Paclitaxel fed larvae were touched with a 46°C heat probe and their response time was measured in seconds with a cut-off of 10 seconds. Different treatments were tested: food control, ethanol control, 0.1 µM, 0.5 µM, 2.5 µM, 5 µM, and 10 µM paclitaxel. Five repeats were performed (n = 130 - 180). DOI, 10.5256/f1000research.13581.d191023 (Hamoudi et al., 2018b).\n\nDataset 3: Dendritic morphology of third instar ppk-CD4-tdGFP. Confocal images of vehicle control and 0.1 µM paclitaxel treated larvae. Images represent class IV md-da neurons at abdominal segment A5. Images are at 40x magnification and 0.7 zoom factor. Scale bar represents 40 µm. DOI, 10.5256/f1000research.13581.d222138 (Hamoudi et al., 2018c).", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported in part through NHMRC project grants APP1026310, APP1029672, APP1028887, APP1046090, APP1042416, APP1086851, and by a NHMRC career development fellowship II CDF1111940.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nAddington J, Freimer M: Chemotherapy-induced peripheral neuropathy: an update on the current understanding [version 1; referees: 2 approved]. F1000Res. 2016; 5(F1000 Faculty Rev): 1466. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nAndré N, Braguer D, Brasseur G, et al.: Paclitaxel induces release of cytochrome c from mitochondria isolated from human neuroblastoma cells’. Cancer Res. 2000; 60(19): 5349–53. PubMed Abstract\n\nBabcock DT, Landry C, Galko MJ: Cytokine signaling mediates UV-induced nociceptive sensitization in Drosophila larvae. Curr Biol. 2009; 19(10): 799–806. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBhattacharya MR, Gerdts J, Naylor SA, et al.: Taxanes: the genetic toxicity of paclitaxel and docetaxel in somatic cells of Drosophila melanogaster. Mutagenesis. 2012; 16(1): 79–84.\n\nBoland BA, Sherry V, Polomano RC: Chemotherapy-induced peripheral neuropathy in cancer survivors. Oncol Nurse Edn. 2010; 24(2): 33–38, 42–43. Reference Source\n\nBurton AW, Fanciullo GJ, Beasley RD, et al.: Chronic pain in the cancer survivor: a new frontier. Pain Med. 2007; 8(2): 189–198. PubMed Abstract \| Publisher Full Text\n\nCavaletti G, Tredici G, Braga M, et al.: Experimental peripheral neuropathy induced in adult rats by repeated intraperitoneal administration of taxol. Exp Neurol. 1995; 133(1): 64–72. PubMed Abstract \| Publisher Full Text\n\nChang AY, Kim K, Glick J, et al.: Phase II study of taxol, merbarone, and piroxantrone in stage IV non-small-cell lung cancer: The Eastern Cooperative Oncology Group Results. J Natl Cancer Inst. 1993; 85(5): 388–94. PubMed Abstract \| Publisher Full Text\n\nCunha KS, Reguly ML, Graf U, et al.: Taxanes: the genetic toxicity of paclitaxel and docetaxel in somatic cells of Drosophila melanogaster. Mutagenesis. 2001; 16(1): 79–84. PubMed Abstract \| Publisher Full Text\n\nGill PS, Tulpule A, Espina BM, et al.: Paclitaxel is safe and effective in the treatment of advanced AIDS-related Kaposi's sarcoma. J Clin Oncol. 1999; 17(6): 1876–83. PubMed Abstract \| Publisher Full Text\n\nHamoudi Z, Khuong TM, Cole T, et al.: Dataset 1 in: A fruit fly model for studying paclitaxel-induced pain. F1000Research. 2018a. Data Source\n\nHamoudi Z, Khuong TM, Cole T, et al.: Dataset 2 in: A fruit fly model for studying paclitaxel-induced pain. F1000Research. 2018b. Data Source\n\nHamoudi Z, Khuong TM, Cole T, et al.: Dataset 3 in: A fruit fly model for studying paclitaxel-induced pain. F1000Research. 2018c. Data Source\n\nHausheer FH, Schilsky RL, Bain S, et al.: Diagnosis, management, and evaluation of chemotherapy-induced peripheral neuropathy. Semin Oncol. 2006; 33(1): 15–49. PubMed Abstract \| Publisher Full Text\n\nHolmes FA, Walters RS, Theriault RL, et al.: Phase II trial of taxol, an active drug in the treatment of metastatic breast cancer. J Natl Cancer Inst. 1991; 83(24): 1797–805. PubMed Abstract \| Publisher Full Text\n\nHwang RY, Zhong L, Xu Y, et al.: Nociceptive neurons protect Drosophila larvae from parasitoid wasps. Curr Biol. 2007; 17(24): 2105–2116. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLipton RB, Apfel SC, Dutcher JP, et al.: Taxol produces a predominantly sensory neuropathy. Neurology. 1989; 39(3): 368–373. PubMed Abstract \| Publisher Full Text\n\nMcGuire WP, Rowinsky EK, Rosenshein NB, et al.: Taxol: A unique antineoplastic agent with significant activity in advanced ovarian epithelial neoplasms. Ann Intern Med. 1989; 111(4): 273–9. PubMed Abstract \| Publisher Full Text\n\nNeely GG, Hess A, Costigan M, et al.: A genome-wide Drosophila screen for heat nociception identifies α2δ3 as an evolutionarily conserved pain gene. Cell. 2010; 143(4): 628–638. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nReyes-Gibby CC, Morrow PK, Buzdar A, et al.: Chemotherapy-induced peripheral neuropathy as a predictor of neuropathic pain in breast cancer patients previously treated with paclitaxel. J Pain. 2009; 10(11): 1146–50. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSahenk Z, Barohn R, New P, et al.: Taxol neuropathy. Electrodiagnostic and sural nerve biopsy findings. Arch Neurol. 1994; 51(7): 726–729. PubMed Abstract \| Publisher Full Text\n\nSeretny M, Currie GL, Sena ES, et al.: Incidence, prevalence, and predictors of chemotherapy-induced peripheral neuropathy: A systematic review and meta-analysis. Pain. 2014; 155(12): 2461–70. PubMed Abstract \| Publisher Full Text\n\nShimozuma K, Ohashi Y, Takeuchi A, et al.: Taxane-induced peripheral neuropathy and health-related quality of life in postoperative breast cancer patients undergoing adjuvant chemotherapy: N-SAS BC 02, a randomized clinical trial. Support Care Cancer. 2012; 20(12): 3355–64. PubMed Abstract \| Publisher Full Text\n\nTracey WD Jr, Wilson RI, Laurent G, et al.: painless, a Drosophila gene essential for nociception. Cell. 2003; 113(2): 261–73. PubMed Abstract \| Publisher Full Text\n\nWani MC, Taylor HL, Wall ME, et al.: Plant antitumor agents. VI. The isolation and structure of taxol, a novel antileukemic and antitumor agent from Taxus brevifolia. J Am Chem Soc. 1971; 93(9): 2325–7. PubMed Abstract \| Publisher Full Text" }	[ { "id": "30150", "date": "01 Mar 2018", "name": "Adam Claridge-Chang", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors characterize the thermal nociception in Drosophila larvae that have been cultured on a range of paclitaxel concentrations. Using a heat probe to elicit the rolling defense behavior, they find that while paclitaxel has no effect on response times with a 46°C probe, it shortens probe response times when the larvae have been grown on 2.5 µm paclitaxel or above.\nThe authors and readers might like to consider the following comments on and questions about the 23 Jan 2018 version.\nThe Title describes a model for studying paclitaxel-induced pain, however the assay uses a heat probe to induce pain, and paclitaxel lowers the sensitivity to the probe, thus modeling the hyperalgesia component of paclitaxel CIPN. Would the Title better serve the reader if edited to focus on this side-effect specifically? In Abstract–Results, the authors write: \"We found that paclitaxel increases pain perception in a dose-dependent manner, without overt morphological changes.\" Changing \"perception\" to \"sensitivity\" would eliminate the baggage of the former word. In Abstract–Conclusions: \"Our simple, high throughput model can be combined with genomics approaches to identify regulators of chemotherapy-induced pain to eliminate its adverse side effects.\" However, they have not established that this is high-throughput by most common definitions of the term, nor do they show anywhere in the paper that it can be combined with genomics. The Conclusions would be improved if rephrased to better reflect what the data show. \"High-throughput\" is a phrasal adjective that requires a hyphen. In Introduction it says \"This system is amenable to high throughput screening\" however, this is not shown in the present manuscript nor is a reference cited in support of the statement. \"Krustal-Wallis ANOVA.\" Correct to \"Kruskal–Wallis.\" I encourage the authors to use estimation statistics instead of significance testing. This would involve presenting and discussing the effect sizes. For example, in Figure 1c, it looks like 2.5 µm paclitaxel has the effect of reducing response time by ~1.5 s. It would be nice to calculate standardized effect sizes (e.g. Cohen's d) of the paclitaxel effects; this would allow the authors and readers to estimate sample sizes needed for a screen (and thus possible throughput rates). In Results, the authors write \"Thus we establish that paclitaxel sensitizes larvae to heat pain via enhancing sensory neuron or higher order nociception, and not via inducing overt morphological changes.\" Is it true that enhancing sensory neuron or higher-order nociception are the only two alternatives? If not, this sentence should be rephrased. The Conclusions section reads more like an overview of future plans for the assay. Could it be rewritten to more closely address the paper's findings?\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? No", "responses": [ { "c_id": "4047", "date": "16 Oct 2018", "name": "Greg Neely", "role": "Author Response", "response": "Thank you for your comments.Response to comment #1: That’s a reasonable point. Since first submission, we have new data that shows our model also involves peripheral neuropathy, so taken together, we have updated the title to capture this aspect and address the reviewer’s comment. The new title is “A fruit fly model for studying paclitaxel-induced peripheral neuropathy and hyperalgesia”. We hope this is acceptable.Response to comment #2: Done.Response to comment #3: We have revised this section and now state “Our simple system can be applied to identify regulators of chemotherapy-induced pain”.Response to comment #4: Done.Response to comment #5: We have now included a reference for this statement.Response to comment #6: Done.Response to comment #7: We thank the reviewers for their comment and we have incorporated the estimation statistics into our analysis and added all the data points. Response to comment #8: We have calculated the effect size, added it to the graphs (1C and 1D), and we have now also mentioned this in the results section. Moreover, we have changed the data representation to show all the data points.Response to comment #9: Good point, this has been revised as suggested.Response to comment #10: We have now added a discussion section." } ] }, { "id": "30145", "date": "22 Mar 2018", "name": "Salahadin Abdi", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis study examined the effects of Paclitaxel exposure on Drosophila larval nociception system and the authors propose that their model is suitable for high throughput screening and further mechanistic studies. The study is overall an interesting and clearly written, however, I do have the following concerns: 1. The dose response effect of thermal stimulation was only at 42 degrees. There was no discussion or explanation why this effect was not seen at 46 degrees. 2. The behavior experiment was based on thermal stimulation. I would be interested why mechanical stimulation was not chosen since mechanical sensitivity is common among patients who develop Paclitaxel induced peripheral neuropathy? 3. It is not clear to me what the timeline is between the exposure of the larvae with paclitaxel and performing microscopic studies. 4. There should be at least a short discussion about the result.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Partly\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nI cannot comment. A qualified statistician is required.\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Partly", "responses": [ { "c_id": "4046", "date": "16 Oct 2018", "name": "Greg Neely", "role": "Author Response", "response": "Thank you for your comments. Response to comment #1: This effect is not seen at 46°C. At this temperature intensity, larvae respond rapidly (~1.5 seconds) and it is difficult to see even faster responses. To look for hyperalgesia, we instead lowered the heat stimulus intensity to 42°C, which is at the threshold for nociception in this system, and where nociceptive responses take on average ~5 seconds to elicit. Response to comment #2: The type IV multidendritic nociceptor neurons that transduce heat nociception also transduce mechanical nociception, as these neurons are multimodal. We have tried on numerous occasions to generate reproducible data for mechanical nociception but so far in our hands this assay does not work well enough for us to feel comfortable publishing. Given the multimodal nature of type IV multidendritic nociceptor neurons, we reasoned that thermal hyperalgesia is a good readout for the overall sensitization of these sensory neurons. Response to comment #3: The animals are born into paclitaxel containing food, and then early third instar are collected at day 6 to assess nociception or dendritic morphology. This information was provided in the methods, however we have further clarified this aspect. Response to comment #4: We have now written a short discussion, please see discussion section." } ] } ]	1	https://f1000research.com/articles/7-99
https://f1000research.com/articles/7-876/v1	22 Jun 18	{ "type": "Research Article", "title": "Fuchs heterochromic iridocyclitis-associated glaucoma: a retrospective comparison of primary Ahmed glaucoma valve implantation and trabeculectomy with mitomycin C", "authors": [ "Hamed Esfandiari", "Nils A. Loewen", "Kiana Hassanpour", "Ali Fatourechi", "Shahin Yazdani", "Chao Wang", "Mehdi Yaseri", "Mohammad Pakravan", "Hamed Esfandiari", "Kiana Hassanpour", "Ali Fatourechi", "Shahin Yazdani", "Chao Wang", "Mehdi Yaseri", "Mohammad Pakravan" ], "abstract": "Background: The aim of this study was to compare the safety and efficacy of primary trabeculectomy with mitomycin C and Ahmed glaucoma valve (AGV) implantation in patients with Fuchs heterochromic iridocyclitis (FHIC)-related glaucoma, a rare complication of an uncommon form of uveitis. Methods: In this retrospective comparative case series, 26 FHIC-associated glaucoma patients received trabeculectomy (n=12) or an AGV (n=14). Primary outcome measures were surgical success, defined as intraocular pressure (IOP) ≤21 mmHg, decreasing ≥20% from baseline, and no secondary glaucoma surgery. Secondary outcome measures were the number of glaucoma medications, complications, best corrected visual acuity (BCVA), and IOP. Results: The follow-up was 34.0±17.7 months in patients that received trabeculectomy and 33.4±18.6 months in AGV (P= 0.837). The cumulative probability of success rate was 41.7% for trabeculectomy and 85.7% for AGV, with no significant difference in complications (P>0.05). The IOP in patients that received trabeculectomy dropped from 23.4±3.3 mmHg to 21.6±5.2 mmHg at the final visit (P= 0.041). In patients that received AGV, the IOP decreased from 24±7.8 to 17.1±2.6 mmHg (P= 0.003). The number of glaucoma medications at baseline were 3.3±0.5 in those that received trabeculectomy and 3±0.6 in those that received AGV (P=0.233), and decreased to 2.4±1.0 (P=0.008) and 1.7±0.6 (P=0.002), respectively. BCVA was equal in both groups and did not change (P>0.05). Conclusion: Primary AGV had a higher success rate than trabeculectomy, with patients also needing fewer medications for the management of FHIC-associated glaucoma.", "keywords": [ "Fuchs heterochromic iridocyclitis", "glaucoma drainage implant", "trabeculectomy", "uveitic glaucoma." ], "content": "Introduction\n\nFuchs heterochromic iridocyclitis (FHIC) is a rare form of uveitis. While the incidence of all forms of uveitis is approximately 0.035% of the population1, the incidence of FHIC is only about 0.00105% (3% of all uveitis cases)2,3 and occurs in both eyes in 10% of patients4. It is characterized by low-grade intraocular inflammation, small stellate keratic precipitates, and iris stromal atrophy5. Recent evidence points towards an association between rubella and FHIC6,7, but an association between FHIC and toxoplasmosis and toxocariasis has also been reported8,9. Affected patients are often asymptomatic for years and mostly present with symptoms of a cataract or floaters during the third or fourth decade of life. Because the presentation is often variable, FHIC is among the most underdiagnosed conditions in ophthalmology10. Since there is an average 3.7-year delay in diagnosing FHIC, it should be considered as a differential diagnosis for any young patient with unilateral low-grade uveitis and good visual acuity5. Although FHIC is frequently complicated by cataract formation in two-thirds of patients, the outcome of phacoemulsification and intraocular lens implantation is excellent and comparable to that in normal eyes11. Older age and a cataract can put patients with FHIC at risk of glaucoma12 which occurs in 15 to 59%13,14.\n\nSince anterior and posterior synechiae are uncommon in this condition, angle-closure mechanisms do not play an important role in the development of glaucoma. Abnormal angle vessels, physical obstruction of trabecular meshwork by inflammatory cells, disruption of uveal and juxtacanalicular structures, trabecular meshwork fibrosis and steroid-induced ocular hypertension are all contributing causes13,15.\n\nFHIC often responds poorly to medical management, requiring a surgical intervention to control intraocular pressure (IOP)11,13,14,16. There is a paucity of literature regarding the best initial surgical approach in the management of FHIC-associated glaucoma. The purpose of this study was to compare the outcomes of the two most common surgical interventions, glaucoma drainage device implantation and trabeculectomy, for glaucoma caused by FHIC. We hypothesized that Ahmed glaucoma drainage devices, a valved implant, would have a lower failure rate but at the expense of a higher average pressure as seen in other glaucomas with these modalities17.\n\n\nMethods\n\nThis study was approved by the ethics committee and the Institutional Review Board (IRB) at the Ophthalmic Research Center of Shahid Beheshti University of Medical Sciences (Tehran, Iran, protocol number: IR.SBMU.ORC.REC.1391.2) and followed the tenets of the Declaration of Helsinki. The IRB waived patient consent for the use of their medical records in this retrospective chart review. The chart review occured at the Labbafinejad Medical Center, Tehran, Iran, and included charts from May 2001 to September 2017, yielding 26 patients with FHIC-associated glaucoma that either had mitomycin C (MMC)-augmented trabeculectomy or a primary Ahmed glaucoma valve (AGV) implantation. Inclusion criteria were age equal to or above 18 years of age and a diagnosis of FHIC-associated glaucoma. FHIC-associated glaucoma was defined as cases of previously known FHIC or diagnosed as FHIC at the time of presentation accompanied by uncontrolled IOP and progressive glaucomatous optic neuropathy. Exclusion criteria consisted of prior glaucoma surgery, need to combined either surgery with cataract extraction, ocular or systemic comorbidities that could affect the procedure and study outcomes including immunodeficiency, connective tissue disease and uncontrolled diabetes. Patients were not formally matched across demographics.\n\nDemographic and baseline data, including age, gender, baseline best corrected visual acuity (BCVA), IOP, number of medications, anterior chamber cells18, type of surgery, and surgical details were recorded. In all cases, surgery was only performed when the eyes were not more than 0.5 inflamed18. Primary outcome measures were surgical success defined as IOP ≤21 mmHg and decreased ≥20% from the baseline, no secondary glaucoma surgery, and no loss of light perception.\n\nSecondary outcome measures were the rate of complications, cataract development, number of medications, IOP reduction and inflammation. Hypotony was defined as an IOP <6 mmHg at any postoperative visit, and hypertensive phase following AGV implantation was defined as an IOP >21 mmHg during the first 3 months after the surgery (with or without medications)19. All postoperative data for each surgery were documented until the last follow-up visit or when a secondary glaucoma surgery was performed.\n\nTrabeculectomy. A 7-0 silk traction suture was passed through the superior cornea. A conjunctival peritomy was performed at the supranasal quadrant followed by Tenon’s dissection. Wet-field cautery was used to stop episcleral vessels bleeding. A 4×3 mm trapezoidal half-thickness scleral flap was created, followed by lamellar dissection to the peripheral cornea. Sponges soaked in 0.04% MMC were applied for 3 minutes. After creating a sideport, a keratome was used to enter the anterior chamber underneath the flap, and a block of clear cornea was removed using a Kelly punch. The scleral flap was closed relatively tightly with two releasable sutures so that spontaneous drainage was minimal. The conjunctiva was closed with 10-0 nylon sutures. At the conclusion of surgery betamethasone and cefazolin were injected into the subtenon space away from the site of operation. The postoperative regimen consisted of chloramphenicol 0.5% eye drops (Sina Darou Lab. Co., Tehran, Iran) four times a day for 1 week and betamethasone 0.1% eye drops (Sina Darou Lab. Co., Tehran, Iran) six times a day, which was tapered to 4, 3, 2, 1 times a day every two weeks.\n\nAhmed glaucoma valve implantation. A 7-0 silk traction suture was placed through the superior clear cornea. The conjunctiva was opened 4 mm posterior to the limbus in the supratemporal quadrant, and a blunt dissection of the Tenon was performed using Westcott scissors to provide space for the plate insertion. The device (Ahmed glaucoma drainage implant, model FP7, New World Medical, Rancho Cucamonga, CA, USA) was primed with 2 ml of buffered saline solution (BSS) and gently pushed into the subtenon space. The plate was secured to the sclera 10 mm posterior to the limbus using 7-0 silk sutures. The tube was trimmed bevel-up with an estimated intracameral length of 2 mm. A 23-gauge needle was inserted into the anterior chamber bevel-up, parallel to the iris and 1 mm posterior to the limbus. The tube was passed through the tunnel into the anterior chamber and secured to the sclera with a 10-0 nylon suture. A 5×8 mm scleral patch graft was placed over the tube. Tenon’s capsule and the conjunctiva were closed using a running 10-0 nylon mattress suture. At the end of the surgery, 0.5 ml of subtenon triamcinolone (40 mg/ml) was injected next to the plate in four patients. Betamethasone (4 mg) and cefazolin (50 mg) were injected into the inferior subconjunctival space upon conclusion of the surgery. The postoperative regimen consisted of chloramphenicol 0.5% eye drops four times a day for 1 week and betamethasone 0.1% eye drops six times a day, which was tapered to 4, 3, 2, 1 times a day every two weeks.\n\nTo test for a difference between the two groups at baseline, we used the t-test, Mann-Whitney, chi-square and Fisher’s exact test. We used a general linear model and ordinal logistic regression to compare the groups adjusted for the baseline. Changes within groups were evaluated using paired t-test and Wilcoxon signed rank test. A P-value less than 0.05 was considered statistically significant. All statistical analyses were performed with SPSS software (IBM Corp. Released 2016. IBM SPSS Statistics for Windows, Version 24.0. Armonk, NY) Data was described as frequency (percent), mean ± standard deviation or median and range.\n\n\nResults\n\nA total of 26 patients were included for the final analysis, of whom 14 were male (53.8%). There were 12 trabeculectomies and 14 AGV surgeries. All cases were primary surgeries with no history of glaucoma surgery. There was no significant difference regarding sex, age, IOP, BCVA, and numbers of glaucoma medications at baseline (Table 1). The mean age at the time of surgery for trabeculectomy was 47.5±6.1 years and for AGV was 45.9±9.3 years (P= 0.608). In total, 10 patients (83.3%) in the trabeculectomy group were phakic and 14 patients (100%) in the AGV group were phakic (P=0.203). Preoperatively, the angle was open in all patients upon gonioscopy. In the trabeculectomy group, two patients had a phacoemulsification and lens implantation in the same session. The mean follow-up time was 34±17.7 months in the trabeculectomy group and 33.4±18.6 months in the AGV group (P= 0.837).\n\nBCVA, best corrected visual acuity; IOP, intraocular pressure. †Using t-test. ‡Using Mann–Whitney U-test. Using chi-squared test. Using Fisher’s exact test.\n\nSurgical success at the final follow-up was 41.7% for trabeculectomy surgery and 85.7% in AGV (P= 0.025). IOP decreased significantly from 24±7.8 mmHg at baseline to 17.14±2.6 mmHg at the final follow-up in AGV (P= 0.003). The corresponding numbers for trabeculectomy were 23.4±3.3 and 21.58±5.2 mmHg, respectively (P= 0.041; Table 2). AGV had a significantly lower average IOP at the final follow-up visit compared to trabeculectomy (P= 0.018). There were three patients in the trabeculectomy group and one in the AGV group that needed a surgical revision specifically to control high IOP. AGV was used as a secondary glaucoma surgery in all these cases. The number of glaucoma medications decreased significantly from 3±0.6 at baseline to 1.71±0.6 at the final follow-up visit in the AGV group (P= 0.002). The medications in the trabeculectomy group were 3.3±0.5 at baseline and 2.41±1.01 at the conclusion of the study, respectively (P= 0.008).\n\n†Using t-test. ‡Using Mann–Whitney U-test. Using Fisher’s exact test. § Adjusted for the baseline, based on General linear model. ¥Adjusted for the baseline, based on ordinal logistic regression. □Using Wilcoxon signed rank test. $ Using paired sample t-test.\n\nPatients in the AGV group needed fewer glaucoma medications at the final follow-up (P= 0.041). Kaplan–Meier survival curves for the two groups are shown in Figure 1. The estimated mean survival time of the surgery was 20.8 months for those in the AGV group and only 12.7 months for those in the trabeculectomy group (P= 0.002). The reason for failure of trabeculectomy was bleb fibrosis. Five patients (37.5%) in the AGV group experienced an early hypertensive phase.\n\nLog-rank P = 0.002. Estimated mean survival time of trabeculectomy was 12.7 months (95% confidence interval, 8.5–16.9). Estimated mean survival time for glaucoma drainage device implant surgery is 20.8 months (95% confidence interval, 17.2–24.4).\n\nTriamcinolone had no impact on IOP (P= 0.320). The most frequent complication in both groups was hyphema (Table 3). In total, five patients in the trabeculectomy group (41.6%) and three patients in the AGV group (21.4%) developed hyphema (P= 0.292) which could be managed conservatively. There was one patient in the AGV group and three in the trabeculectomy group that exhibited established choroidal effusions that had to be drained. The anterior cell reaction did not exceed 0.5 during the preoperative or postoperative exam and there were no significant difference between the AGV and trabeculectomy groups (p=0.871 and 0.9, respectively). One patient in each group developed endophthalmitis. The endophthalmitis in the patient that underwent AGV was preceded by tube exposure. The patient underwent vitrectomy and the device was removed. A new AGV was implanted in the infranasal location in the same session. Although the endophthalmitis in trabeculectomy could be controlled by an injection of intravitreal antibiotics and a corticosteroid (vancomycin (25 mg in 0.5 ml), ceftazidime (100 mg in 0.5 ml) and dexamethasone (6 mg in 0.25 ml) injected as a bolus), a glaucoma drainage device was needed. In the AGV group, two patients experienced endothelial touch, and one of them underwent tube shortening due to early corneal decompensation. Hypotony was observed in two cases in the trabeculectomy group in the early postoperative period, which resolved without a surgical intervention within 1 month. There was no significant difference between the rate of complications between the two groups (Table 3). None of the listed complications were significant factors for surgical failure in AGV or trabeculectomy. A cataract extraction was indicated in five patients in the trabeculectomy group and in only one patient in the AGV group. The mean time between trabeculectomy and cataract surgery was 9.1±4.3 months.\n\nN/A, not applicable.\n\n\nDiscussion\n\nIn this retrospective study, we evaluated the outcome of two common surgeries for FHIC-associated glaucoma, a valved tube shunt (AGV) and trabeculectomy with MMC. Although FHIC is rare, occurring in about 0.00105% of the population2,3, and the course is typically mild, almost 50% of patients with FHIC develop glaucoma12–14,21 and require aggressive management. We found that AGV had a significantly higher success rate than trabeculectomy, confirming our hypothesis. Unexpectedly, patients also needed fewer glaucoma medications in AGV, while the complication rate was similar.\n\nMost glaucoma patients exhibit open-angle configuration on gonioscopy. Decreased outflow is instead caused by inflammatory cells, fibrotic changes of the trabecular meshwork, and long-term steroid use13–15. The management of FHIC-associated glaucoma is challenging22. In a study by Liesegang, 66% of patients with FHIC-associated glaucoma needed surgical intervention and did so earlier in life than individuals with primary open-angle glaucoma16. Laser trabeculoplasty is contraindicated because it can exacerbate the inflammation, cause bleeding from neovascularization of the angle and induce peripheral anterior synechiae23.\n\nWhen the uveitis is only mildly active, trabeculectomy can be performed to quickly lower IOP, including in FHIC24, despite the risk of bleb failure25 but the success rate is less than 30% at 5 years26,27, far worse than in primary open-angle glaucoma28. Although FHIC is not typically characterized by severe inflammation, trabeculectomy outcomes have been reported to be worse13,14. The high rate of hyphema in our series likely contributed to this because blood can reduce the bleb size in trabeculectomy29,30 but not in tube shunts. Hyphema commonly occurs in FHIC because of the angle neovascularization in FHIC31 and rupture of these fragile vessels following IOP reduction5. Cataract is another common occurrence in FHIC and has an increased incidence after trabeculectomy. In our study, five out of ten patients required cataract surgery. The high rate of cataract formation after trabeculectomy appears to be an under-reported risk of failure of trabeculectomy in FHIC-associated glaucoma. For this reason, same-session cataract removal should be considered because modern phacoemulsification at the time of glaucoma surgery may have only a negligible impact on IOP outcomes32\n\nThe reported intermediate success rate for glaucoma drainage devices in uveitic glaucoma is between 66% and 85%33–35. In a study by Tan et al., the short- and long-term success rate of non-valved Baerveldt implants in uveitic glaucoma was 89% and 75%, respectively36. In another study, Satana et al. used valved Ahmed implants in 14 patients with uveitic glaucoma secondary to Behcet disease and reported the cumulative probability of surgical success rate of 90.9% at 18 months follow-up34. Kwon et al. examined the outcome of AGV implantation in 28 patients with uveitic glaucoma including FHIC and reported a success rate of 75% during the 2-year follow-up37. Voykov and colleagues assessed the short and intermediate-term success rate of AGV implantation in 17 patients with FHIC- associated glaucoma38. Qualified success defined as 6 mmHg ≤ IOP ≤21 mmHg was achieved in 58.3% of patients after 1 year and only 38.4% after 3 years, although 88% of patients had conjunctival scarring from prior procedures, a known risk factor39. This may also explain the rate of complications (23% tube exposure, 23% device exposure, 6% endophthalmitis, 6% diplopia, 11.7% hypotony) in the mentioned study.\n\nOur results indicate that primary AGV has a higher cumulative probability of success in FHIC. Regardless, consistent with prior studies35,40, our complication rate was high. This highlights how challenging and unpredictable uveitis is, even though FHIC is a relatively mild form of uveitis. Given the high rate of complications and the patients’ age, microincisional procedures should be considered for FHIC that have been proven to be safe and effective in other mild-to-moderate forms of uveitic glaucoma41,42, and appropriate for a range of glaucoma severity43. The occurrence of a hypertensive phase in the current study is lower than the 47% incidence rate reported by Voykov et al.38. The lower rate could partially be explained by the modulation of encapsidation and inflammation by triamcinolone used in several patients here, although contradictory results have been reported44,45. Bleb vascularity is a recently identified risk factor for bleb failure28 that we did not examine here. An intensified treatment for this problem can include bevacizumab, which can be used subconjunctivally46 instead of intravitreally47. Another explanation for the lower incidence of early hypertension may be the start of aqueous suppressant to reduce fibroblast stimulation from stretch and cytokines48.\n\nLimitations of this study are the relatively small patient number dictated by the overall rareness of FHIC, a retrospective design and the use of triamcinolone in some patients. Although the anterior chamber cell reaction in FHIC is not as prominent as in other uveitic glaucomas, we could have systematically measured the anterior chamber cell reaction to see whether it was associated with survival of each operation.\n\nIn conclusion, this study shows that Ahmed glaucoma drainage devices are superior to trabeculectomy in FHIC-associated glaucoma. The relatively high complication rate is consistent with prior reports and highlights the considerable risk associated with this relatively mild form of uveitis.\n\n\nData availability\n\nDataset 1. Raw data collected from all study participants. DOI: http://doi.org/10.5256/f1000research.15244.d20741720.", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nWe acknowledge support from the Initiative to Cure Glaucoma of the Eye and Ear Foundation of Pittsburgh; NIH CORE Grant P30 EY08098 to the Department of Ophthalmology, and from an unrestricted grant from Research to Prevent Blindness, New York, NY.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nA pre-print of article is available on the University of Pittsburgh Institutional Repository (http://d-scholarship.pitt.edu/33292/).\n\n\nReferences\n\nWakefield D, Chang JH: Epidemiology of uveitis. Int Ophthalmol Clin. 2005; 45(2): 1–13. PubMed Abstract \| Publisher Full Text\n\nBonfioli AA, Curi AL, Orefice F: Fuchs’ heterochromic cyclitis. Semin Ophthalmol. 2005; 20(3): 143–6. PubMed Abstract \| Publisher Full Text\n\nVan Gelder RN, Prasad AG: Review of Uveitis. SLACK Incorporated. 2008; 189. Reference Source\n\nJones NP: Fuchs' Heterochromic Uveitis: a reappraisal of the clinical spectrum. Eye (Lond). 1991; 5( Pt 6): 649–61. PubMed Abstract \| Publisher Full Text\n\nMohamed Q, Zamir E: Update on Fuchs’ uveitis syndrome. Curr Opin Ophthalmol. 2005; 16(6): 356–63. PubMed Abstract \| Publisher Full Text\n\nSuzuki J, Goto H, Komase K, et al.: Rubella virus as a possible etiological agent of Fuchs heterochromic iridocyclitis. Graefes Arch Clin Exp Ophthalmol. 2010; 248(10): 1487–91. PubMed Abstract \| Publisher Full Text\n\nde Groot-Mijnes JD, de Visser L, Rothova A, et al.: Rubella virus is associated with fuchs heterochromic iridocyclitis. Am J Ophthalmol. 2006; 141(1): 212–4. PubMed Abstract \| Publisher Full Text\n\nGritz DC, Wong IG: Incidence and prevalence of uveitis in Northern California; the Northern California Epidemiology of Uveitis Study. Ophthalmology. 2004; 111(3): 491–500; discussion 500. PubMed Abstract \| Publisher Full Text\n\nToledo de Abreu M, Belfort R Jr, Hirata PS: Fuchs’ heterochromic cyclitis and ocular toxoplasmosis. Am J Ophthalmol. 1982; 93(6): 739–44. PubMed Abstract \| Publisher Full Text\n\nBrancaleoni A, Sekkat L, Bouchenaki N, et al.: Delay in the diagnosis of Fuchs’ uveitis and its deleterious consequences. iovs.arvojournals.org. 2003. Reference Source\n\nRam J, Kaushik S, Brar GS, et al.: Phacoemulsification in patients with Fuchs’ heterochromic uveitis. J Cataract Refract Surg. 2002; 28(8): 1372–8. PubMed Abstract \| Publisher Full Text\n\nTugal-Tutkun I, Güney-Tefekli E, Kamaci-Duman F, et al.: A cross-sectional and longitudinal study of Fuchs uveitis syndrome in Turkish patients. Am J Ophthalmol. 2009; 148(4): 510–5.e1. PubMed Abstract \| Publisher Full Text\n\nToniolo JT, Hall AJ, Smith JG, et al.: Risk Factors for Glaucoma in a Cohort of Patients with Fuchs Heterochromic Iridocyclitis. Ocul Immunol Inflamm. 2017; 25(6): 753–759. PubMed Abstract \| Publisher Full Text\n\nJones NP: Glaucoma in Fuchs’ Heterochromic Uveitis: aetiology, management and outcome. Eye (Lond). 1991; 5(Pt 6): 662–7. PubMed Abstract \| Publisher Full Text\n\nAbraham S, George R: Glaucoma in Uveitis. In: Uveitis: An Update. Springer, New Delhi. 2016; 49–55. Publisher Full Text\n\nLiesegang TJ: Clinical features and prognosis in Fuchs’ uveitis syndrome. archopht.jamanetwork.com. Arch Ophthalmol. 1982; 100(10): 1622–1626. Publisher Full Text\n\nWilson MR, Mendis U, Paliwal A, et al.: Long-term follow-up of primary glaucoma surgery with Ahmed glaucoma valve implant versus trabeculectomy. Am J Ophthalmol. 2003; 136(3): 464–70. PubMed Abstract \| Publisher Full Text\n\nJabs DA, Nussenblatt RB, Rosenbaum JT, et al.: Standardization of uveitis nomenclature for reporting clinical data. Results of the First International Workshop. Am J Ophthalmol. 2005; 140(3): 509–16. PubMed Abstract \| Publisher Full Text\n\nNouri-Mahdavi K, Caprioli J: Evaluation of the hypertensive phase after insertion of the Ahmed Glaucoma Valve. Am J Ophthalmol. 2003; 136(6): 1001–8. PubMed Abstract \| Publisher Full Text\n\nEsfandiari H, Loewen NA, Hassanpour K, et al.: Dataset 1 in: Fuchs heterochromic iridocyclitis-associated glaucoma: a retrospective comparison of primary Ahmed glaucoma valve implantation and trabeculectomy with mitomycin C. F1000Research. 2018. Data Source\n\nLa Hey E, de Vries J, Langerhorst CT, et al.: Treatment and prognosis of secondary glaucoma in Fuchs’ heterochromic iridocyclitis. Am J Ophthalmol. 1993; 116(3): 327–40. PubMed Abstract \| Publisher Full Text\n\nSalim S: Fuchs Heterochromic Iridocyclitis, Glaucoma. In: Schmidt-Erfurth U, Kohnen T, editors. Springer Berlin Heidelberg. Encyclopedia of Ophthalmology. 2016; 1–3.\n\nZhou Y, Aref AA: A Review of Selective Laser Trabeculoplasty: Recent Findings and Current Perspectives. Ophthalmol Ther. 2017; 6(1): 19–32. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCeballos EM, Beck AD, Lynn MJ: Trabeculectomy with antiproliferative agents in uveitic glaucoma. J Glaucoma. 2002; 11(3): 189–96. PubMed Abstract \| Publisher Full Text\n\nNoble J, Derzko-Dzulynsky L, Rabinovitch T, et al.: Outcome of trabeculectomy with intraoperative mitomycin C for uveitic glaucoma. Can J Ophthalmol. 2007; 42(1): 89–94. PubMed Abstract\n\nStavrou P, Murray PI: Long-term follow-up of trabeculectomy without antimetabolites in patients with uveitis. Am J Ophthalmol. 1999; 128(4): 434–9. PubMed Abstract \| Publisher Full Text\n\nTowler HM, McCluskey P, Shaer B, et al.: Long-term follow-up of trabeculectomy with intraoperative 5-fluorouracil for uveitis-related glaucoma. Ophthalmology. 2000; 107(10): 1822–8. PubMed Abstract \| Publisher Full Text\n\nEsfandiari H, Pakravan M, Loewen NA, et al.: Predictive value of early postoperative IOP and bleb morphology in Mitomycin-C augmented trabeculectomy. F1000Res. 2017; [cited 2017 Oct 27]. 6: 1898. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWise JB: Treatment of chronic postfiltration hypotony by intrableb injection of autologous blood. Arch Ophthal. 1993; 111(6): 827–30. PubMed Abstract \| Publisher Full Text\n\nNakatake S, Yoshida S, Nakao S, et al.: Hyphema is a risk factor for failure of trabeculectomy in neovascular glaucoma: a retrospective analysis. BMC Ophthalmol. 2014; 14: 55. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVelilla S, Dios E, Herreras JM, et al.: Fuchs’ heterochromic iridocyclitis: a review of 26 cases. Ocul Immunol Inflamm. 2001; 9(3): 169–75. PubMed Abstract \| Publisher Full Text\n\nParikh HA, Bussel II, Schuman JS, et al.: Coarsened Exact Matching of Phaco-Trabectome to Trabectome in Phakic Patients: Lack of Additional Pressure Reduction from Phacoemulsification. PLoS One. 2016; 11(2): e0149384. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCeballos EM, Parrish RK 2nd, Schiffman JC: Outcome of Baerveldt glaucoma drainage implants for the treatment of uveitic glaucoma. Ophthalmology. 2002; 109(12): 2256–60. PubMed Abstract \| Publisher Full Text\n\nSatana B, Yalvac IS, Sungur G, et al.: Ahmed Glaucoma Valve Implantation for Uveitic Glaucoma Secondary to Behçet Disease. J Glaucoma. 2015; 24(8): 607–12. PubMed Abstract \| Publisher Full Text\n\nPapadaki TG, Zacharopoulos IP, Pasquale LR, et al.: Long-term results of Ahmed glaucoma valve implantation for uveitic glaucoma. Am J Ophthalmol. 2007; 144(1): 62–9. PubMed Abstract \| Publisher Full Text\n\nTan AN, Cornelissen MF, Webers CAB, et al.: Outcomes of severe uveitic glaucoma treated with Baerveldt implant: can blindness be prevented? Acta Ophthalmol. 2017; 96(1): 24–30. PubMed Abstract \| Publisher Full Text\n\nKwon HJ, Kong YXG, Tao LW, et al.: Surgical outcomes of trabeculectomy and glaucoma drainage implant for uveitic glaucoma and relationship with uveitis activity. Clin Experiment Ophthalmol. 2017; 45(5): 472–80. PubMed Abstract \| Publisher Full Text\n\nVoykov B, Doycheva D, Deuter C, et al.: Outcomes of Ahmed Glaucoma Valve Implantation for Glaucoma Secondary to Fuchs Uveitis Syndrome. Ocul Immunol Inflamm. 2017; 25(6): 760–766. PubMed Abstract \| Publisher Full Text\n\nEibschitz-Tsimhoni M, Schertzer RM, Musch DC, et al.: Incidence and management of encapsulated cysts following Ahmed glaucoma valve insertion. J Glaucoma. 2005; 14(4): 276–9. PubMed Abstract \| Publisher Full Text\n\nRachmiel R, Trope GE, Buys YM, et al.: Ahmed glaucoma valve implantation in uveitic glaucoma versus open-angle glaucoma patients. Can J Ophthalmol. 2008; 43(4): 462–7. PubMed Abstract \| Publisher Full Text\n\nAnton A, Heinzelmann S, Neß T, et al.: Trabeculectomy ab interno with the Trabectome® as a therapeutic option for uveitic secondary glaucoma. Graefes Arch Clin Exp Ophthalmol. 2015; 253(11): 1973–8. PubMed Abstract \| Publisher Full Text\n\nKaplowitz K, Loewen NA: Trabectome-Mediated Ab Interno Trabeculectomy for Secondary Glaucoma or as a Secondary Procedure. In: Aref AA, Varma R, editors. Advanced Glaucoma Surgery. Springer International Publishing. (Essentials in Ophthalmology). 2015; 15–29. Publisher Full Text\n\nLoewen RT, Roy P, Parikh HA, et al.: Impact of a Glaucoma Severity Index on Results of Trabectome Surgery: Larger Pressure Reduction in More Severe Glaucoma. PLoS One. 2016; 11(3): e0151926. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTuralba AV, Pasquale LR: Hypertensive phase and early complications after Ahmed glaucoma valve implantation with intraoperative subtenon triamcinolone acetonide. Clin Ophthalmol. 2014; 8: 1311–6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nYazdani S, Doozandeh A, Pakravan M, et al.: Adjunctive triamcinolone acetonide for Ahmed glaucoma valve implantation: a randomized clinical trial. Eur J Ophthalmol. 2017; 27(4): 411–6. PubMed Abstract \| Publisher Full Text\n\nNilforushan N, Yadgari M, Kish SK, et al.: Subconjunctival bevacizumab versus mitomycin C adjunctive to trabeculectomy. Am J Ophthalmol. 2012; 153(2): 352–7.e1. PubMed Abstract \| Publisher Full Text\n\nRamezani A, Esfandiari H, Entezari M, et al.: Three intravitreal bevacizumab versus two intravitreal triamcinolone injections in recent onset central retinal vein occlusion. Acta Ophthalmol. 2014; 92(7): e530–9. PubMed Abstract \| Publisher Full Text\n\nPakravan M, Rad SS, Yazdani S, et al.: Effect of early treatment with aqueous suppressants on Ahmed glaucoma valve implantation outcomes. Ophthalmology. 2014; 121(9): 1693–8. PubMed Abstract \| Publisher Full Text" }	[ { "id": "35657", "date": "13 Aug 2018", "name": "Syril K. Dorairaj", "expertise": [ "Reviewer Expertise Glaucoma" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is an interesting study where authors retrospectively compared the safety and efficacy of primary trabeculectomy with Mitomycin C and Ahmed glaucoma valve (AGV) implantation in patients with Fuchs heterochromic iridocyclitis (FHIC)-related glaucoma. Esfandiari et al hypothesized that Ahmed glaucoma drainage devices, would have a lower failure rate compared to trabeculectomy and based on their results, authors successfully conclude that Primary AGV implantation had a higher success rate than trabeculectomy, with patients also needing fewer medications for the management of FHIC-associated glaucoma.\n\nThough there is a general trend towards utilizing minimally invasive angle based surgeries, authors have done a timely job in evaluating trabeculectomy and AGV since there is greater risk of failure of glaucoma surgery consequent of trabecular meshwork fibrosis and steroid response in FHIC associated glaucoma. This study gives insight into designing a randomized control trials comparing device-assisted trabeculectomy with conventional trabeculectomy and glaucoma drainage device in the future.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [] }, { "id": "37324", "date": "29 Aug 2018", "name": "Naveed Nilforushan", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nGeneral comment: The authors are trying to compare the results of Trabx and AGV in FHI patients, but the main concern and cofounding factor are the significant difference between the rates of phaco surgery during the follow up which is known as a risk factor for the failure of Trabx. This effect could be more prominent when another risk factor such as chronic uveitis is present. The main cause of such difference between 2 groups, in terms of success, with sharp downward slope after month 6 is most probably related to cataract surgery. This needs to be discussed more.\n\nSpecific comments:\nThe hypothesis mentioned in the last paragraph of introduction which has been stressed again in the first paragraph of the discussion needs more justification and explanation. What does it mean: “having lower failure rate but at the expense of higher average pressure in AGV” According to the cited article “Lower IOPs were noted for the trabeculectomy group during the first year. With longer follow-up, the IOPs and the cumulative probabilities of success were comparable between the two groups”. Higher success rate for AGV in other studies such as TVT was due to higher rates of hypotonia in Trabx which was not the cause of failure in the present study.\n\nCombined glaucoma surgery and phaco was among the exclusion criteria, but 2 cases in Trabx have had a combined surgery. These cases should have been excluded.\n\nBy considering which criteria, in some cases Trabx and in others AGV were performed?\n\nThe definition for hypertensive phase is not precise and cannot differentiate those cases with non functioning tube from hypertensive phase. Please give some explanations and give the rate of hypertensive phase.\n\nHow many surgeons did perform those surgeries?\n\nWere there cases that needed postop needling or subconjunctival antifibrotic injections?\n\nWas the grading of inflammation on those with failed surgery during follow up different from the others?\n\nAlthough the following 2 references have used old fashion type of surgery (No antifibrotic or 5FU) but still according to reference 26 and 27 the 5 year success rate of Trabx was 78% and 67% respectively. This has been mentioned less than 30% in the article. Please correct or give explanation.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4040", "date": "15 Oct 2018", "name": "Nils Loewen", "role": "Author Response", "response": "General comment: The authors are trying to compare the results of Trabx and AGV in FHI patients, but the main concern and confounding factor are the significant difference between the rates of phaco surgery during the follow up which is known as a risk factor for the failure of Trabx. This effect could be more prominent when another risk factor such as chronic uveitis is present. The main cause of such difference between 2 groups, in terms of success, with a sharp downward slope after month 6 is most probably related to cataract surgery. This needs to be discussed more. Authors: Thank you for this suggestion. We have added the following: “Another risk factor for bleb failure is the higher rates of cataract extraction in T. Cataract formation or progression is common even after uneventful trabeculectomy with a range of 6 to 58 %. The fact that trabeculectomy tends to enhance cataract progression while cataract extraction can reduce the success rate of trabeculectomy limits its use in the management of glaucoma in FHIC.” Specific comments: 1. The hypothesis mentioned in the last paragraph of introduction which has been stressed again in the first paragraph of the discussion needs more justification and explanation. What does it mean: “having lower failure rate but at the expense of higher average pressure in AGV” According to the cited article “Lower IOPs were noted for the trabeculectomy group during the first year. With longer follow-up, the IOPs and the cumulative probabilities of success were comparable between the two groups”. The higher success rate for AGV in other studies such as TVT was due to higher rates of hypotonia in Trabx which was not the cause of failure in the present study. Authors: We appreciate and agree and have deleted this statement. We now state: “There is a paucity of literature regarding the best initial surgical approach in the management of FHIC-associated glaucoma. The purpose of this study was to compare the outcomes of the two most common surgical interventions, glaucoma drainage device implantation, and trabeculectomy, for glaucoma caused by FHIC.” 2. Combined glaucoma surgery and phaco were among the exclusion criteria, but 2 cases in Trabx have had a combined surgery. These cases should have been excluded. Authors: Thank you for pointing this out. We have corrected the description of exclusion criteria. They consisted of prior glaucoma surgery, ocular or systemic comorbidities that could affect the procedure and study outcomes including immunodeficiency, connective tissue disease and uncontrolled diabetes. Patients were not formally matched across demographics. 3. By considering which criteria, in some cases, Trabx and in others AGV were performed? Authors: There were no strict criteria. Based on the surgeon’s preference and comfort as well as patients’ decision, one of these procedures was performed. 4. The definition for the hypertensive phase is not precise and cannot differentiate those cases with a non-functioning tube from the hypertensive phase. Please give some explanations and give the rate of hypertensive phase. Authors: We added the following in the Methods and Results section, respectively, to be more clear about the hypertensive phase: “A hypertensive phase following AGV implantation was defined as an IOP rising above 21 mmHg during the first 3 months with or without medications after an initial IOP that was lower than 21 mmHg.” “The reason for trabeculectomy failure was bleb fibrosis. Five patients (37.5%) in the AGV group experienced an early hypertensive phase. All patients in the AGV group had early postoperative IOPs of less than 21 mmHg indicating that all valved implants were functioning properly.” 5. How many surgeons did perform those surgeries? Authors: We added to Methods: “Three surgeons performed the procedures.” 6. Were there cases that needed postop needling or subconjunctival antifibrotic injections? Authors: We added to Results: “Eight trabeculectomy patients needed bleb needling with MMC. Needling was done during the postoperative period for impending failure from a contracting bleb. Thirty minutes after injecting 0.1 ml of 0.02% Mitomycin-C into the bleb-adjacent subtenon space a 27 gauge needle was used to reform the bleb and dissect adhesions at the slit lamp.” 7. Was the grading of inflammation on those with failed surgery during follow up different from the others? Authors: We now mention in the Discussion: “Additionally, in this retrospective study the anterior cell reaction was not assessed systematically and objectively enough to test for a formal correlation to the outcome of surgeries. While the anterior chamber cell reaction in FHIC is not as prominent as in other uveitic glaucomas and remained at or below a grading of 0.5, it would be interesting to examine this aspect in more detail to try to understand why AGV patients did better.” 8. Although the following 2 references have used old fashion type of surgery (No antifibrotic or 5FU) but still according to reference 26 and 27 the 5 years success rate of Trabx was 78% and 67% respectively. This has been mentioned less than 30% in the article. Please correct or give an explanation. Authors: Thank you. We changed that paragraph to now read: “When the uveitis is only mildly active, a trabeculectomy can be performed to quickly lower IOP, including in FHIC24 even though the risk of bleb failure is relatively high in uveitic glaucoma25. Although the trabeculectomy success rate in uveitis is above 50% at 5 years26,27, it is lower the one reported for epibulbar glaucoma drainage implants even in primary open-angle glaucoma28." } ] } ]	1	https://f1000research.com/articles/7-876
https://f1000research.com/articles/7-1185/v1	03 Aug 18	{ "type": "Opinion Article", "title": "The Congress Impact Factor: A proposal from board members of the World Society of Emergency Surgeons.it (WSES) and Academy of Emergency Medicine and Care (AcEMC)", "authors": [ "Belinda De Simone", "Luca Ansaloni", "Micheal Denis Kelly", "Federico Coccolini", "Massimo Sartelli", "Salomone Di Saverio", "Michele Pisano", "Gianfranco Cervellin", "Gianluca Baiocchi", "Fausto Catena", "Luca Ansaloni", "Micheal Denis Kelly", "Federico Coccolini", "Massimo Sartelli", "Salomone Di Saverio", "Michele Pisano", "Gianfranco Cervellin", "Gianluca Baiocchi", "Fausto Catena" ], "abstract": "Many scientific congresses and conferences are held every year around the world. The aim of the World Society of Emergency Surgeons.it (WSES) and Academy of Emergency Medicine and Care (AcEMC) was to develop a simple mathematical parameter as an indicator of academic quality and scientific validity of a congress. In this opinion article, a new metric, the Congress Impact Factor (IFc), is proposed taking into consideration the widely used Impact Factor as an indicator of journals’ prestige and using H-index analysis. The IFc is derived from the mathematical ratio between the mean H-index of invited lecturers normalized for lecture topic and number of lectures in the conference. In case of multiple sessions, the mean of all IFc is calculated along with its standard deviation. We conclude that the IFc can be a useful measure for evaluating and comparing congress prestige, and may also represent a potentially useful parameter for improving academic curriculum and helping participants to choose the more prestigious meetings for their education.", "keywords": [ "Congress Impact Factor", "HIndex", "Educational Program", "Scientific Quality", "Academic Curriculum" ], "content": "Introduction\n\nMany scientific congresses, meetings and conferences are organized each year around the world. Each congress can be promoted by a scientific society, which supports and organizes scientific sessions choosing topics and inviting national and international scientists as discussants, speakers or chairs. The choice of attending a specific congress is largely based on personal preferences, scientific area of interest and/or research, or simply as a desire to investigate, update and discuss topics of scientific relevance within the scientific community. The identification of the most useful and prestigious congresses and conferences organized by scientific societies is challenging, especially for young doctors who have not yet garnered a sufficient level of expertise. The scientific impact of a congress can only be valuable when supported by a good scientific program; the lectures delivered by experts in the field are essential for analyzing and discussing different medical and surgical topics1.\n\nThe journal Impact Factor (IF), originally conceived by Irving H. Sher and Garfield in the early 1960s, is a bibliometric parameter aimed to evaluate journals’ prestige. It is usually calculated by dividing the number of citations in the previous two years to the number of citable items published in the same period2. Therefore, a journal IF is based on two elements: the numerator, which is the number of citations in the current year to items published by the journal in the previous two years; the denominator, which is the number of citable items published in the previous two years3,4. Information about citations is obtained from a database now maintained by Clarivate Analytics (formerly by the Institute for Scientific Information). The list of journals’ IF is then published in the InCites Journal Citation reports, which is hence a useful means for establishing the absolute and relative (i.e., within a specific scientific field) prestige of a journal. Notably, albeit originally conceived for evaluating journals’ prestige, the IF is occasionally used also for evaluating scientists according to the number of articles published in high-IF journals5–7.\n\nUnlike the IF, the H-index is a different metric used to evaluate scientists’ prestige according to the number of citations https://scholar.googleblog.com/2012/04/google-scholar-metrics-for-publications.html8. The H-index was suggested in 2005 by Jorge E. Hirsch as a tool for determining theoretical physicists' relative quality9 and is sometimes called the Hirsch index or Hirsch number. The definition of the H-index is that a scholar with an index of x has published x papers each of which has been cited in other papers at least x times10. Consequently it involves the number of publications and the number of citations for publication to evaluate the scientific activity of a researcher and not only the total number of citations or publications. The limit is that the H-index can only be properly used for comparing scientists working in the same field.\n\n\nThe congress impact factor\n\nThe aim of this opinion article is to present a mathematical coefficient to assess the quality and the academic validity of a scientific congress, using the IF formula and H-index calculation to create a useful tool: the Congress Impact Factor (IFc).\n\nWe propose that the IFc is calculated using the following formula:\n\nMean H-index of Lecturers normalized for lecture topic was calculated using Google Scholar by Publish or Perish Harzing.com. For example, to obtain a mean H index normalized for lecture topic by Publish or Perish program is very easy: you have to choose to send your query by Google Scholar, searching for the “Name” and “Surname” of the author; automatically you will obtain your H index. Then you narrow down the search field to lecture topic and obtain H-index normalized for topic, for that author. All results should be analyzed checking for the right scientist, excluding non-relevant ones.\n\nSubsequently, the mean H-index of all lecturers at the congress, normalized for lecture topic, is calculated to obtain a mean H-index plus a standard deviation. This value is divided by the number of lectures given in the congress obtaining the IFc.\n\nThen the mean of all standard deviations must be calculated.\n\nConsiderations:\n\n- The Chair’s H-index is always excluded because they do not give lectures.\n\n- Only invited lectures should be considered.\n\n- Free paper presenters are excluded because their academic value is too unpredictable and variable as we do not know how much they can influence the literature in the future: will they be published? In which journal? Will they be cited? How many times will they be cited?\n\n- In case of a multi-session congress, a mean of all sessions plus standard deviation should be calculated.\n\nMethods. As an example, we calculated the IFc for the first day of the Open Abdomen International Consensus Conference held in Dublin on July 2016. This was a consensus conference on critical surgical abdomen that produced guidelines on indications and benefits of open abdomen in non-trauma patients, which were published in the World Journal of Emergency Surgery11. There were no other published proceedings of this conference. To create a comparison, we performed the H-index calculation for the same lecturers normalized for a different topic, “acute” “leukemia”, where none of the lecturers had specific expertise. The following mesh-words were used by Publish or Perish to calculate the H-index for every lecturer and mean H index in the two different topics (Table S1): \"Name Surname\" and \"open abdomen\" and for the other evaluation \"Name Surname\" and \"acute leukemia”. The comparison was made by the Student's t-test Statistical analysis was performed using IBM SPSS Statistics 22. P<0.05 was considered significant.\n\nResults. Invited speakers attending the two sessions of first day were 14 international emergency and trauma surgeons with a specific expertise in the open abdomen field. Table S1 shows the results of the IFc calculation based on normalized H-index for topic. The mean normalized H-index for open abdomen was 13.57 (SD 8.033), and the IFc was 0.96. The mean normalized H-index for the same speakers with a topic outside their expertise (acute leukemia) is 1.85 (SD 1.80; Table S1). The IFc for this hypothetical congress was 0.13. The difference between normalized H-index calculated between these two congresses was statistically significant (p=0.0001).\n\n\nDiscussion\n\nIn evaluating the quality and quantity of publications, two major categories of bibliometric indicators are available: quantitative indicators that measure the research productivity of a researcher; performance indicators that evaluate the quality of publications12. The H-index is one of many available bibliometric indicators and is the most popular one to evaluate the academic and scientific activity of a researcher6. In 2005, physicist Jorge E. Hirsch developed this index as a process for quantifying the output of an individual researcher. Hirsch stated: “I propose the index H, defined as the number of papers with citation number ≤ h, as a useful index to calculate the scientific output of a researcher”9.\n\nThe H-index can be very useful in conceiving the IFc as a parameter to assess scientific quality of countless congresses and conferences that are proposed every year by scientific societies. The scientific impact of a congress is measured by a scientific program worthy of attention. We propose this simple indicator to measure quality of a Congress program based on the quality of its invited lecturers. The IFc involves the H-index combining it with IF calculation principles “to dilute” citation parameter with number of published articles. For IFc “the dilution” is performed with the number of lectures planned at the congress. We use the scientific potential given by the H-index of Lecturers invited/called to participate at the congress, normalized for the specific topic, avoiding the possibility that a highly cited scientist could give a lecture on a field outside their expertise, decreasing their educational effect. Dividing normalized H-index with the number of lectures, we can achieve a real-time picture of the quality of the educational meeting with clear evidence of the congress’s scientific impact. Only a limited number of good quality lectures is the source of a significant IFc with effective education of congress participants.\n\nIFc is based on the H-index, which is a strong indicator of scientific quality, and the IF philosophy. Currently they are both used to evaluate the strength of a scientist and of a scientific journal. IFc is able to describe the scientific expertise of lecturers on a specific topic with a quantitative evaluation of the quality of the meeting. For validation, we calculated the IFc for the WSES Consensus Conference on Open Abdomen: this was a high level meeting on a particular topic (open abdomen) where international experts are invited. The results of the validation of IFc suggest that the IFc can be an effective qualitative/quantitative metric for assessing congresses.\n\nOne limitation of IFc is that it would be difficult to calculate IFc in cases of a very large and heterogeneous congresses (e.g. American College of Surgeons). This is because many different symposiums have to be evaluated but the final IFc could be the mean of all these different IFcs using standard deviation to analyze the dispersion.\n\nTo the best of our knowledge there is nothing like the formal IF for conferences. In the past, conference proceedings publications were used to rate “lower quality” as compared to other “higher quality” congresses, especially if articles were published in peer reviewed international journals that are included in Thomson Reuters Journal Citation Reports http://wokinfo.com/products_tools/multidisciplinary/webofscience/cpci/. However with this system it is possible to have a retrospective and quite delayed information which is not so useful for choosing a congress prospectively. In other cases, conference proceedings were ranked in Thomson Reuters using “Conference Proceeding Citation Index”, but this is not comparable with an IF, and in this case you have retrospective and inaccurate information (evaluation of the congress is done a posteriori and without taking in consideration the lecturers). There is also the CORE Conference/Journal Ranking http://www.scimagojr.com/journalsearch.php?q=conference&tip=jou; http://arnetminer.org/page/conference-rank/html/All-in-one.htm, but again it is not a parameter based on strong indicators. There are other sources that could prove useful as an estimate of conference quality. Google Scholar lists top venues mixing journals and conferences in their listings. They list H-index of the venue instead of an IF, but again this is a misleading information (a high H-index venue can organize a Congress with low H-index lecturers).\n\nChoosing the best Congress to attend can be difficult, and especially so for young attendees. Residents, scientific researchers and students require an ideal metric system to use as an indicator of scientific quality of a congress, so they can have the possibility to join congresses with high scientific impact and build on a competitive academic curriculum.\n\nWe believe that the IFc is an effective evaluation tool for a scientific meeting and it can become a valid instrument of education to develop a competitive academic curriculum vitae, i.e. reporting in the curriculum vitae the different conferences attended with their respective IFc.\n\n\nConclusions\n\nBibliometric indicators are essential to evaluate scientific activity both of a researcher and an institution, or a journal.\n\nMany congresses are organized and held every year and analysis of the congress programs shows that not all have a high scientific quality, despite being sponsored by international scientific societies and biomedical companies. In addition fees are requested to participate, and consequently it is very important to attend the best meetings that can improve one’s knowledge of a specific topic. It is important to be able to have a measurement of the quality of any given conference. We propose the IFc as the mathematical ratio between the mean H-index of invited lecturers normalized for lecture topic and the number of lectures at the conference. We believe that the IFc can be a useful metric system to assess the scientific validity of a congress, helping attendees to choose the best quality meeting to attend.\n\n\nData availability\n\nAll data underlying the results are available as part of the article and no additional source data are required.", "appendix": "Author contributions\n\n\n\nWSES board member: BDS, LA, MDK, FCo, MS, SDS, MP, FCa\n\nAcEMC board member: GC\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nSupplementary material\n\nTable S1: Example of IFc calculation for Open Abdomen Congress 2016, in comparison with IFc for hypothetic Acute Leukemia Congress with the same authors.\n\nClick here to access the data.\n\n\nReferences\n\nCatena F, Moore F, Ansaloni L, et al.: Emergency surgeon: \" last of the mohicans\" 2014-2016 editorial policy WSES- WJES: position papers, guidelines, courses, books and original research; from WJES impact factor to WSES congress impact factor. World J Emerg Surg. 2014; 9(1): 14. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGarfield E: The history and meaning of the journal impact factor. JAMA. 2006; 295(1): 90–93. PubMed Abstract \| Publisher Full Text\n\nSeglen PO: Why the impact factor of journals should not be used for evaluating research. BMJ. 1997; 314(7079): 498–502. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSaha S, Saint S, Christakis DA: Impact factor: a valid measure of journal quality? J Med Libr Assoc. 2003; 91(1): 42–46. PubMed Abstract \| Free Full Text\n\nGarfield E: The meaning of the impact factor. Revista internacional de psicología clínica y de la salud=International journal of clinical and health psychology. 2003; 3(2): 363–369. Reference Source\n\nLippi G, Borghi L: A short story on how the H-index may change the fate of scientists and scientific publishing. Clin Chem Lab Med. 2014; 52(2): e1–3. PLoS Medicine Editors. \"The impact factor game.\" PLoS medicine 3.6 (2006): e291. PubMed Abstract \| Publisher Full Text\n\nMeral UM, Alakus U, Urkan M, et al.: Publication Rate of Abstracts Presented at the Annual Congress of the European Society for Surgical Research during 2008-2011. Eur Surg Res. 2016; 56(3–4): 132–40. PubMed Abstract \| Publisher Full Text\n\nJones T, Huggett S, Kamalski J: Finding a way through the scientific literature: indexes and measures. World Neurosurg. 2011; 76(1–2): 36–8. PubMed Abstract \| Publisher Full Text\n\nHirsch JE: An index to quantify an individual's scientific research output. Proc Natl Acad Sci U S A. 2005; 102(46): 16569–72. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nde Meijer VE, Knops SP, van Dongen JA, et al.: The fate of research abstracts submitted to a national surgical conference: a cross-sectional study to assess scientific impact. Am J Surg. 2016; 211(1): 166–71. PubMed Abstract \| Publisher Full Text\n\nCoccolini F, Montori G, Ceresoli M, et al.: The role of open abdomen in non-trauma patient: WSES Consensus Paper. World J Emerg Surg. 2017; 12: 39. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nJoshi MA: Bibliometric indicators for evaluating the quality of scientifc publications. J Contemp Dent Pract. 2014; 15(2): 258–62. PubMed Abstract \| Publisher Full Text" }	[ { "id": "36778", "date": "30 Aug 2018", "name": "Francesco Azzaroli", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nI reviewed with interest the paper entitled “The Congress Impact Factor: a proposal….”and, as far as my knowledge goes, this is the first time that a metric evaluation of a medical congress is proposed.\nThe authors propose to measure an impact factor based on the mean H-index of invited lecturers normalized for lecture topic (i.e. the H-index of an author limited to the topic of the invited lecture) related to the number of invited lectures.\nObviously, this metric has several limitations that come both from the intrinsic original defects of the H-index and from the complexity of evaluating the quality of a conference and of the speakers.\nIn fact, the H-index reflects only the number of papers that have received a certain number of citations and does not include any information about the real contribution of that author to the manuscript nor the number of self citations. Furthermore, it tends to increase with time with increasing number of citations even though that author is no more productive.\nBecause of these limitations several attempts have been made to improve the H-index trying to take into account the contribution of that author to the paper or the period of activity of the researcher adjusting for the number of years since the first publication. Nonetheless, there still is no perfect index to measure the quality and quantity of research which may be affected by so many factors 1-4. In fact, some researchers that have deeply impacted the world of science do not have impressive H-index 2,3.\nDealing with the world of medicine there is another point to consider that is practical expertise. The professionalism of a physician is not represented by the H-index. We all know that being scientifically very productive does not always correspond to being an “hands on” expert and to measure the practical expertise is an even more challenging task. The implementation of such an index could significantly impact the choice of speakers and may leave out non productive “hands on” experts.\nFinally, the metric may be affected by the number of speakers; i.e. a small conference may see its H-index rise if just a few authors with high H-index are invited. In such a case, the median with the range may better reflect the overall composition of invited speakers.\nDespite all these observations, I believe this paper deserves publication in order to start a serious discussion about scientific conferences. However, I believe the road to develop an acceptable measure of the quality of a conference is still long and rough.\nComing specifically to the paper I have the following comments:\nPage 4, last paragraph before the discussion section: it should not be “between these two congresses…” but “…topics…”\n\nThe discussion section should be partially rewritten taking into account the comments I made above and the fact that the H-index is not so robust. The authors should acknowledge the limitations of the metric and the possible drawbacks.\n\nIn the last paragraph of the discussion the authors state that the conference impact factor can become a valid instrument of education to develop a competitive academic curriculum vitae. I disagree with this concept since participating to a conference does not necessarily correspond to an improvement of the professional knowledge. In this view the CME program is more close to this concept than the IF of a conference that does not measure learning. I would erase this sentence limiting the conclusion to the fact that the IFc may represent the first step in developing a simple tool to evaluate scientific conferences.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Partly\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Partly", "responses": [ { "c_id": "4038", "date": "15 Oct 2018", "name": "belinda de simone", "role": "Author Response", "response": "Dear Professor Azzaroli,Thank you to have review our opinion paper. We agree with you in highlighting limitations of the H index. We followed your suggestions and modified the manuscript considering all the issues you reported and upgrading the references, as you can check in the updated version of the paper.Aware of all the criticism existing in literature about the real value of the H index but at the present there is no other indicator that can substitute it." } ] }, { "id": "37590", "date": "17 Sep 2018", "name": "Luca Luzzi", "expertise": [ "Reviewer Expertise Thoracic Surgery", "Lung transplantation", "Robotic surgery" ], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe paper from Ansaloni and co writers, touch on some crucial aspects of scientific divulgation. The first is about the quality of teaching: especially for young surgeons or others specialists the meetings are probably the first and most important instruments of updating. However, meanwhile some international meetings represents the masterpiece in their fields, many others give a less impact or appear redundant with a lower quality. Having a standardized method to rank the meetings you can address young to achieve the best quality information avoiding waste of money and time. On the other hand a sort of scientific meeting ranking could help company to address their investments to the best offers of scientific training. Only a final consideration to take in account, the meeting ranking should consider that also local up to date meetings have still a value to improve the medical or scientific culture in the periphery because more accessible respect an international one. Could be reasonable to cure this aspect with a different standardised event classification, for example; international meetings, nationals, inter-study groups or up to date.\nIn my opinion the paper is for sure worthy of publication and then open for discussion in the scientific arena.\n\nIs the topic of the opinion article discussed accurately in the context of the current literature? Yes\n\nAre all factual statements correct and adequately supported by citations? Yes\n\nAre arguments sufficiently supported by evidence from the published literature? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes", "responses": [ { "c_id": "4037", "date": "15 Oct 2018", "name": "belinda de simone", "role": "Author Response", "response": "Dear Colleague,Thank you for your opinion and suggestions. You have hit our aim in the proposition of the IFc as a tool to evaluate the scientific impact both of an international congress and of a meeting organized with the intent to update knowledge about a specific research's field. By the IFc, we can select the best congress/meeting to upgrade our academic curriculum." } ] } ]	1	https://f1000research.com/articles/7-1185
https://f1000research.com/articles/7-1030/v1	09 Jul 18	{ "type": "Research Article", "title": "Do funding applications where peer reviewers disagree have higher citations? A cross-sectional study", "authors": [ "Adrian G Barnett", "Scott R. Glisson", "Stephen Gallo", "Scott R. Glisson", "Stephen Gallo" ], "abstract": "Background: Decisions about which applications to fund are generally based on the mean scores of a panel of peer reviewers. As well as the mean, a large disagreement between peer reviewers may also be worth considering, as it may indicate a high-risk application with a high return. Methods: We examined the peer reviewers' scores for 227 funded applications submitted to the American Institute of Biological Sciences between 1999 and 2006. We examined the mean score and two measures of reviewer disagreement: the standard deviation and range. The outcome variable was the relative citation ratio, which is the number of citations from all publications associated with the application, standardised by field and publication year. Results: There was a clear increase in relative citations for applications with a higher mean. There was no association between relative citations and either of the two measures of disagreement. Conclusions: We found no evidence that reviewer disagreement was able to identify applications with a higher than average return. However, this is the first study to empirically examine this association, and it would be useful to examine whether reviewer disagreement is associated with research impact in other funding schemes and in larger sample sizes.", "keywords": [ "meta-research", "research funding", "peer review", "citations", "research impact" ], "content": "Introduction\n\nWinning funding is an important stage of the research process and researchers spend large amounts of their time preparing applications1. Applications are typically assessed using relatively small panels of 3 to 12 peer reviewers, sometimes including external reviewers with additional expertise, which is similar to the journal process of editors and reviewers. Given the importance of funding processes to researchers’ careers and the progress of science, there is surprisingly little research on whether funding systems reliably identify the best research. A recent systematic review found there are many unanswered questions in funding peer review, and concluded, “there is a need for open, transparent experimentation and evaluation of different ways to fund research”2. A prior systematic review similarly concluded that studies to examine the accuracy and soundness of funding peer review are “urgently needed”3. Whilst a systematic review of innovations for effectiveness and efficiency in peer review funding found only eight studies and called for more studies4.\n\nThe majority of funding systems rank applications using the mean score from the review panel, and award funding from the highest to the lowest ranked applications, stopping when the budget is exhausted (exceptions are sometimes made for applications below the funding line because of national research priorities). An interesting recent idea is that an application’s mean score may not be the only statistic worth considering, and that the standard deviation in peer reviewers’ scores may also be a useful statistic for ranking applications5. A zero standard deviation means all panel members gave the same score. Larger standard deviations indicate more disagreement between panel members, and this disagreement may be useful for identifying high-risk research that may also have a higher return.\n\nUsing the mean score for ranking may allow panel members to “sink” an application by awarding a low score that pulls the mean below the funding line. Including a measure of reviewer disagreement in funding could ameliorate such “sinking” and allow applications that have strong support from a few reviewers to be supported. This may also increase the diversity in what kinds of applications are funded. Some peer review systems already recognise this issue by giving each panel member a wildcard which allows them to “float” an application above the funding line regardless of other panel members’ scores. At least one funding scheme also includes patient and stakeholder reviewers to increase the diversity of viewpoints6.\n\nA recent systematic review found “suggestive” evidence that funding peer review can have an anti-innovation bias2 and that innovation and risk may not often be sufficiently addressed in review feedback7. There is evidence that riskier cross-disciplinary research has lower success rates8. Some researchers feel they need to write conservative applications that please all members of the panel to achieve a good average score9. However, supporting risky research can have huge benefits for society when it pays off10. In a survey of Australian researchers, 90% agreed with the statement: “I think the NHMRC [the main Australian funding body for health and medical research] should fund risky research that might fail but, if successful, would change the scientific field”11.\n\nPrevious studies have investigated the association between an application’s mean score (or ranking based on the mean) and subsequent citations, where citations are used as a measure of success. Many studies using large sample sizes found either no association or only a weak association between the mean score and the citations of subsequent publications12–16. Other studies have shown a positive association between higher mean peer review scores and increased citations17,18, including a study that used the same data analyzed here19. To our knowledge, no previous study has empirically estimated how the disagreement in peer reviewers’ scores may also predict citations.\n\n\nMethods\n\nWe examined 227 successful grant applications submitted to the American Institute of Biological Sciences between the years 1999 to 2006. These successful applications came from 2,063 total applications (overall 11% success rate). The applications covered a wide range of biomedical research areas, including vision, drug abuse, nutrition, blood-related cancer, kidney disease, autoimmune diseases, malaria, tuberculosis, osteoporosis, arthritis and autism. Applications were assessed by between 2 and 18 peer reviewers, with a median of 10 reviewers. Panels evaluated an average of 25 applications over two days. Ninety percent of applications were reviewed by on-site panels with an average size of 10 reviewers, and 10% of applications were reviewed via teleconferences of 3 reviewers. Further details on the funding process is available in a previous study of how the applications’ mean scores predicted citations19.\n\nOur key predictor is the peer reviewers’ scores. Individual peer reviewers, who were not conflicted, scored applications between 1.0 (best) to 5.0 (worst) in 0.1 increments. To determine funding, the score was averaged across all reviewers. In this study we also consider statistics that measure within-panel disagreement, which are the standard deviation and the range (largest minus smallest score).\n\nThe primary outcome is the citation counts from publications associated with the successful application. The publication data for the funded applications were taken from the mandatory final reports submitted by the applicants. On average, these reports were submitted 5 years after the application’s peer review. Publications were produced from 1 to 8 (average 4.3) years after the review date. Only peer-reviewed publications were counted, confirmed through PubMed and Web of Knowledge searches. Publications listed in the final report as “submitted” or “in preparation”, were included if they could be found as peer-reviewed published papers. Citations were counted in 2014 using Web of Knowledge.\n\nThis analysis used 20,313 citations from 805 peer reviewed publications. The total citation level per funded application was the cumulative citations of all publications. As citations are time-dependent they were standardized using the average citation level of all publications by scientific field and year, using data from a published calculation using the Thomson Reuters Essential Science Indicators database20. These published average rates were determined for 2000–2010 by scientific field, assessed in 2011 and displayed a linear relationship with time (e.g., R2 = 0.99 for the field of molecular biology). We chose molecular biology because it was the highest cited field and in general was the field most applicable to the funded applications.\n\nBecause the Reuters curve was assessed in 2011, we extended the curve for 2014, back calculating using a linear fit which had a very high R2 of 0.99. In this way, we could most accurately standardize the data for the relationship between publication date and citation level and could calculate the Total Relative Citation per application. A total relative citation of 1 meant the application achieved the average number of citations, whereas values above 1 meant a higher than average number of citations.\n\nWe note that a recent study that used both unadjusted citation counts and relative citations, found the two measures gave similar results when used as the key outcome variable21.\n\nThe total relative citations were modelled using a multiple regression model. We ran two models with the three application variables:\n\n1. Review year, mean score, score standard deviation\n\n2. Review year, mean score, score range\n\nOur aim was to examine two measures of panel disagreement: the standard deviation and range. We included mean score because it has already been shown to be an important predictor for these data19 and we were interested in the additional value of a measure of disagreement. We adjusted for review year (1999 to 2006) because there was a difference in the application score statistics over time, and because year was associated with citation numbers, hence it was a potential confounder.\n\nThe citations were first base e log-transformed because of their positive skew. We added a small positive constant of 0.1 before using the log-transform because some citations were zero. The estimates were back-transformed and plotted to show the results on the original relative citations scale. Using equations the multiple regression model was:\n\nwhere Y are the citations and γ are random intercepts to adjust for the potential within-panel correlation in citations (where p(i) is the panel number for application i and M is the total number of panels). The mean (µ) had a constant (β0) and the three application predictors (X) which were first transformed using a fractional polynomial function.\n\nAssociations between the score statistics and citation outcomes could be non-linear, for example a larger difference in citations for a change in mean score from 1.0 to 1.1 compared with a change from 2.0 to 2.1. To model this potential non-linearity we used fractional polynomials to examine a range of non-linear associations between the scores and citations22. The fractional polynomial function is:\n\nWe examined the eight transformations of: P = {−2, −1, −0.5, 0, 0.5, 1, 2, 3} and chose the optimal P using the deviance. The optimal P was chosen for each of the three predictors, meaning we examined 83 = 512 models in total. We only present results for the best model with the smallest deviance. We checked the distribution of the model residuals for multi-modality and outliers, and used Cook’s distance to find influential observations.\n\nSixteen (7%) observations were missing the score standard deviation and 32 (14%) observations were missing the score range because the individual peer reviewer scores were no longer available for some applications at the time of this retrospective analysis. These missing observations were imputed using linear regression with the application variables: review year, mean score, score standard deviation, minimum score, maximum score and range. We used five multiple imputations.\n\nAll analyses were made using R version 3.4.423 with the imputations using the “MICE” package24. The code and anonymized data are available here: http://doi.org/10.5281/zenodo.129938425.\n\nWe report our results using the STROBE guidelines for observational research26.\n\n\nResults\n\nThe histograms in Figure 1 show the distributions of total relative citations and the application score statistics: mean, standard deviation, minimum, maximum and range (maximum minus minimum). There was a strong positive skew in citations with one outlying relative citation of 104; the next largest citation was 34. To counter this positive skew we used a base e log transform in later analyses. There was also a positive skew in the score standard deviation and range, and we also log-transform these predictors.\n\nThe lower the mean score, the better the application did in peer review.\n\nThe scatter-plots in Figure 2 show the associations between the application score statistics. There was a strong correlation between the standard deviation and maximum (0.80), but not between the standard deviation and minimum (0.05). This indicates the largest disagreement is where at least one panel member has given a poor score (remembering lower scores are better). Applications where there was one dissenting panel member with a good score were unlikely to be funded as their mean score would not be competitive, and hence are not in this sample. There was a strong positive correlation between the two measures of panel disagreement, the standard deviation and range (0.93).\n\nThe numbers in the bottom-left diagonal of the plot matrix are the Pearson correlations.\n\nThe scatter-plots in Figure 3 show the association between total relative citations and the score statistics. We used the log-transformed citations and the standard deviation and range to remove the skew and so show a clearer association. The variance in log-citations appears relatively stable over the score statistics, somewhat confirming the validity of log-transforming the citations27. The points along the bottom of the y-axis are the 74 applications (33%) with no citations. Some association between mean score and citations is visible, with a generally downward pattern in citations for increasing score. There is no clear association between citations and either the standard deviation or range.\n\nThe results from the multiple linear regression models are in Table 1. Only the application’s mean score had a statistically significant association with citation numbers.\n\nFP = fractional polynomial.\n\nThe predictions from the multiple linear regression models are in Figure 4. There was a reduction in citations for applications with a higher (worse) mean score. The mean lines are flat for both the standard deviation and range, indicating no association between these score statistics and citation numbers. The 95% confidence intervals for the standard deviation are very wide for large standard deviations. The application with the largest standard deviation was influential according to Cook’s statistic, and removing this application had little impact on the mean line but did reduce these wide intervals (see additional results at https://github.com/agbarnett/funding.disagree). The model residuals had an approximately symmetric and unimodal distribution with no outliers.\n\nThe solid lines are the means and the grey areas are 95% confidence intervals. The dotted horizontal line at 1 represents the average citations, so values above this are better than average.\n\n\nDiscussion\n\nWe found a statistically significant association between an application’s mean score and subsequent citation counts, with the result in the expected direction because applications with better scores had more citations (on average). The largest size of the increase is also practically significant as the highest scores have a mean of 2 (Figure 4), meaning double the average citations.\n\nWe found no association between the two measures of reviewer disagreement and citations counts. It appears that any disagreement between peer reviewers did not indicate an application with a potentially high return.\n\nDisagreements between reviewers about an application can stem from different sources. Disagreements about the proposed methods may mean the study is not viable and would struggle to produce valid results and/or publish papers. Disagreements about the application’s goals may reflect a difference of opinion about the potential impact, and it is these disagreements that are likely more subjective and hence where a higher return is possible if one reviewer is right. Disagreements between reviewers can also occur for more trivial reasons such as the dynamics of the panel and personal disagreements28. A more sophisticated measure of panel disagreement to those used here may be more predictive of the benefits of the research, but such measures would need to be well-defined and prospectively recorded at the panel meeting. Some reviews already breakdown scores into separate areas, such as track record and innovation, and reviewer disagreement could be examined using these separate scores. Each reviewer brings their own experience and biases to the funding process and such intellectual differences influence application scores29,30. Indeed, some recent research has recently indicated that there is more variability in score across reviewers than across proposals31 and previous results indicate inter-rater reliability as very low32. An application’s average score is somewhat due to the “luck of the draw” of what reviewers were selected33. Variations can also occur because of the way the application is summarised at the panel meeting28,34. If a larger disagreement leads to high-risk returns then we might expect an increase in the variance in citations for larger score standard deviations and ranges. This is because there might be more “failures” with zero citations, but also more big returns. Our multiple regression models only examined a change in mean citations and a different statistical model would be needed to examine a change in variance. However, the scatter-plots in Figure 3 show no sign of an increasing variance in citations for higher standard deviations or ranges.\n\nSome have argued that studies like ours are invalid because: 1) they only consider funded applications and do not include unfunded studies, 2) an application’s score is not the only criteria used to award funding (e.g., applications with low scores awarded funding because of national priorities), and 3) because budgets are frequently cut, meaning the actual research may differ from the application35. Studies that follow funded and unfunded fellowship applicants are possible, e.g., 36, but this is very difficult when examining projects that need specific funding37. We believe, despite the limitations of bibliometric measures, it is reasonable to expect a dose-response association between scores and citations within funded applications. Samples that include applications that were funded for reasons other than their mean score, such as national priorities, increase the variance in the key predictors of application scores statistics and hence increase statistical power. Cuts to the budget are important and can hinder the planned research. However, assuming the reviewers believed the study was still viable with the reduced budget, such studies still test the ability of a panel to predict what research will have the greatest return.\n\nCitations are an imperfect measure of the impact of research because many citations have little worth and scientists often report that their most highly cited work is not their best38. Studies that examine more detailed outcomes such as translation into practice or cost-benefits would be incredibly useful, however these studies would themselves require funding as it would involve further data collection, analyses and interviews of the applicants. Our results may not be generalisable to other funding schemes, especially as there are large differences between fields in their perceptions of what makes a good application28. It would be useful to examine whether reviewer disagreement is associated with research impact in other funding schemes and in larger sample sizes.\n\n\nConclusions\n\nWe found no association between two measures of reviewer disagreement when assessing an application and the subsequent research impact of that application as measured by citation counts.\n\n\nData availability\n\nThe code and anonymized data are available here: http://doi.org/10.5281/zenodo.129938425.", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nAB receives funding from the Australian National Health and Medical Research Council (APP1117784).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgements\n\nWe thank the AIBS SPARS staff for their work in implementing these reviews for nearly a decade.\n\n\nReferences\n\nHerbert DL, Barnett AG, Clarke P, et al.: On the time spent preparing grant proposals: an observational study of Australian researchers. BMJ Open. 2013; 3(5): pii: e002800. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGuthrie S, Ghiga I, Wooding S: What do we know about grant peer review in the health sciences? [version 2; referees: 2 approved]. F1000Res. 2018; 6: 1335. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDemicheli V, Di Pietrantonj C: Peer review for improving the quality of grant applications. Cochrane Database Syst Rev. 2007; (2): MR000003. PubMed Abstract \| Publisher Full Text\n\nShepherd J, Frampton GK, Pickett K, et al.: Peer review of health research funding proposals: A systematic map and systematic review of innovations for effectiveness and efficiency. PLoS One. 2018; 13(5): e0196914. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLinton JD: Improving the peer review process: Capturing more information and enabling high-risk/high-return research. Research Policy. 2016; 45(9): 1936–1938. Publisher Full Text\n\nFleurence RL, Forsythe LP, Lauer M, et al.: Engaging patients and stakeholders in research proposal review: the patient-centered outcomes research institute. Ann Intern Med. 2014; 161(2): 122–130. PubMed Abstract \| Publisher Full Text\n\nGallo S, Thompson L, Schmaling K, et al.: Risk evaluation in peer review of grant applications. Environment Systems and Decisions. 2018; 38(2): 216–229. Publisher Full Text\n\nBromham L, Dinnage R, Hua X: Interdisciplinary research has consistently lower funding success. Nature. 2016; 534(7609): 684–687. PubMed Abstract \| Publisher Full Text\n\nFang FC, Casadevall A: NIH peer review reform--change we need, or lipstick on a pig? Infect Immun. 2009; 77(3): 929–932. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBraben DW: Promoting the Planck Club: How Defiant Youth, Irreverent Researchers and Liberated Universities Can Foster Prosperity Indefinitely. Wiley, 2014; ISBN 9781118546383. Reference Source\n\nBarnett AG: Ask the researcher: The experience of applying for health and medical research funding in Australia Survey results. 2017. Reference Source\n\nScheiner SM, Bouchie LM: The predictive power of NSF reviewers and panels. Frontiers in Ecology and the Environment. 2013; 11(8): 406–407. Publisher Full Text\n\nKaltman JR, Evans FJ, Danthi NS, et al.: Prior publication productivity, grant percentile ranking, and topic-normalized citation impact of NHLBI cardiovascular R01 grants. Cir Res. 2014; 115(7): 617–624. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLauer MS, Danthi NS, Kaltman J, et al.: Predicting productivity returns on investment: Thirty years of peer review, grant funding, and publication of highly cited papers at the National Heart, Lung, and Blood Institute. Circ Res. 2015; 117(3): 239–243. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDoyle JM, Quinn K, Bodenstein YA, et al.: Association of percentile ranking with citation impact and productivity in a large cohort of de novo NIMH-funded R01 grants. Mol Psychiatry. 2015; 20(9): 1030–1036. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nFang FC, Bowen A, Casadevall A: NIH peer review percentile scores are poorly predictive of grant productivity. eLife. 2016; 5. pii: e13323. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDanthi N, Wu CO, Shi P, et al.: Percentile ranking and citation impact of a large cohort of National Heart, Lung, and Blood Institute-funded cardiovascular R01 grants. Circ Res. 2014; 114(4): 600–606. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLi D, Agha L: Research funding. Big names or big ideas: do peer-review panels select the best science proposals? Science. 2015; 348(6233): 434–8. PubMed Abstract \| Publisher Full Text\n\nGallo SA, Carpenter AS, Irwin D, et al.: The validation of peer review through research impact measures and the implications for funding strategies. PLoS One. 2014; 9(9): e106474. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nTimes Higher Education: Citation averages, 2000-2010, by fields and years. 2011. Reference Source\n\nLindner MD, Torralba KD, Khan NA: Scientific productivity: An exploratory study of metrics and incentives. PLoS One. 2018; 13(4): e0195321. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRoyston P, Ambler G, Sauerbrei W: The use of fractional polynomials to model continuous risk variables in epidemiology. Int J Epidemiol. 1999; 28(5): 964–974. PubMed Abstract \| Publisher Full Text\n\nR Core Team: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria, 2018. Reference Source\n\nvan Buuren S, Groothuis-Oudshoorn K: mice: Multivariate imputation by chained equations in R. J Stat Softw. 2011; 45(3): 1–67. Publisher Full Text\n\nBarnett AG: agbarnett/funding.disagree: First release of funding disagreement code and data (version v1.0). 2018. Publisher Full Text\n\nvon Elm E, Altman DG, Egger M, et al.: The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. PLoS Med. 2007; 4(10): e296. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nManning WG: The logged dependent variable, heteroscedasticity, and the retransformation problem. J Health Econ. 1998; 17(3): 283–95. PubMed Abstract \| Publisher Full Text\n\nCoveney J, Herbert DL, Hill K, et al.: ‘Are you siding with a personality or the grant proposal?’: observations on how peer review panels function. Res Integr Peer Rev. 2017; 2(1): 19. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBoudreau KJ, Guinan EC, Lakhani KR, et al.: Looking Across and Looking Beyond the Knowledge Frontier: Intellectual Distance, Novelty, and Resource Allocation in Science. Manage Sci. 2016; 62(10): 2765–2783. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGallo SA, Sullivan JH, Glisson SR: The Influence of Peer Reviewer Expertise on the Evaluation of Research Funding Applications. PLoS One. 2016; 11(10): e0165147. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPier EL, Brauer M, Filut A, et al.: Low agreement among reviewers evaluating the same NIH grant applications. Proc Natl Acad Sci U S A. 2018; 115(12): 2952–2957. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCicchetti DV: The reliability of peer review for manuscript and grant submissions: A cross-disciplinary investigation. Behavioral and Brain Sciences. 1991; 14(1): 119–135. Publisher Full Text\n\nGraves N, Barnett AG, Clarke P: Funding grant proposals for scientific research: retrospective analysis of scores by members of grant review panel. BMJ. 2011; 343: d4797. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGregorius S, Dean L, Cole DC, et al.: The peer review process for awarding funds to international science research consortia: a qualitative developmental evaluation [version 3; referees: 2 approved]. F1000Res. 2018; 6: 1808. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLindner MD, Nakamura RK: Examining the Predictive Validity of NIH Peer Review Scores. PLoS One. 2015; 10(6): e0126938. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBornmann L, Wallon G, Ledin A: Does the committee peer review select the best applicants for funding? An investigation of the selection process for two European molecular biology organization programmes. PLoS One. 2008; 3(10): e3480. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nDecullier E, Huot L, Chapuis FR: Fate of protocols submitted to a French national funding scheme: A cohort study. PLoS One. 2014; 9(6): e99561. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nIoannidis JP, Boyack KW, Small H, et al.: Bibliometrics: Is your most cited work your best? Nature. 2014; 514(7524): 561–562. PubMed Abstract \| Publisher Full Text" }	[ { "id": "36742", "date": "20 Sep 2018", "name": "Jonathan Shepherd", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nWe thank the authors for this very interesting study. We have some comments and suggestions which we think will enhance the manuscript.\n\n1. The primary outcome measure is the citation counts from publications associated with the successful application. Publications were produced from 1 to 8 years after the peer review date (average 4.3 years). This does not appear to take account of varying time since the projects were funded (i.e. 7 year gap between projects that were funded from 1999 to 2006). Thus, older studies would have had more time for publications to be produced and cited. We therefore suggest a more meaningful outcome measure would be either the number of citations per year per study, or the total number of citations in, say, the 5 years following the final project report, or some other standardised project milestone. Adding the review year to the model does not seem to adequately control for this factor (though qualified statistical advice is needed to clarify this). We think this is the issue that likely affects the interpretation of the results the most in our critique.\n\n2. The number of citations was standardised by academic field – i.e. molecular biology. Was there any variation in study designs within this field that might also change the expected number of citations (e.g. systematic reviews may attract more citations than primary experimental studies in some fields)? It would be useful for context if there could be a table with some basic aggregate details of the funded studies, such as types of study design, molecular biological application, study sample characteristics, duration of study etc. This would help to put the results into context.\n\n3. The impact of a piece of research on which referees could not agree might be either lower or higher than those on which they could. So, it would be useful to plot the standard deviation of the citations against the standard deviation of the peer review score.\n\n4. As well as using multiple imputation to correct for missing data, a sensitivity analysis in which cases with missing data are omitted would be useful.\n\n5. Some measure of the model fit would be useful, eg. adjusted R squared\n\n6. There was a wide variation in the number of reviewers per article from 2 to 18. Was this due to differences in the kind of research, amount of funding requested or some other perceived risk on behalf of the funder? Could this artificially influence the standard deviation of the score, confounding any association with citations?\n\n7. Fractional polynomial model results are only presented for the best fitting model with the smallest deviance. This is acceptable in principle, but it would be useful for the authors to comment on whether there was any variation in the results according to the other fractional polynomial transformations (if available). This will provide confidence in the robustness of the findings.\n\n8. Reference is made in the first paragraph to a “recent systematic review” by Guthrie et al (20181) (and also in the third paragraph). We note that this publication doesn’t refer to itself as being a systematic review, and indeed, it is an update of a 2009 review which describes itself as a non-systematic review. We would suggest using the tern “non-systematic review”, or just “literature review”.\n\n9. Thank you for citing our own recent systematic review on peer review of grants in health. You mention that the review included eight studies and called for further research in this area, which is correct. However, the review focused specifically on studies aiming to improve the effectiveness and efficiency of peer review. These were drawn from a wider set of 83 studies on peer review which we systematically mapped. In the map there were some studies which focused on assessing the impact of funded research, eg. In terms of bibliometrics. Thus, there is a body of evidence on this topic, though we didn't systematically review it in detail. We are happy to provide you with a list of these studies.\n\n10. The sentence on page 3 beginning “A recent systematic review found “suggestive” evidence that funding peer review can have an anti-innovation bias2 and that innovation and risk may not often be sufficiently addressed in review feedback7” needs re-wording as not only is the Guthrie et al paper a non-systematic review, but the second reference cited in that sentence by Gallo et al is a survey (i.e. not a review at all). The way the sentence is phrased implies that it is a systematic review.\n\n11. Suggest amending the sentence on page 3 “Many studies using large sample sizes found either no association or only a weak association between the mean score and the VOLUME of citations of subsequent publications”\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Partly\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nPartly\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4056", "date": "15 Oct 2018", "name": "Adrian Barnett", "role": "Author Response", "response": "1. The primary outcome measure is the citation counts from publications associated with the successful application. Publications were produced from 1 to 8 years after the peer review date (average 4.3 years). This does not appear to take account of varying time since the projects were funded (i.e. 7 year gap between projects that were funded from 1999 to 2006). Thus, older studies would have had more time for publications to be produced and cited. We therefore suggest a more meaningful outcome measure would be either the number of citations per year per study, or the total number of citations in, say, the 5 years following the final project report, or some other standardized project milestone. Adding the review year to the model does not seem to adequately control for this factor (though qualified statistical advice is needed to clarify this). We think this is the issue that likely affects the interpretation of the results the most in our critique.Thanks for the comment. Actually, the citation values are normalized per publication year, thus older studies would not have an advantage as the resultant citations from publications are normalized by year. It is true that more time post peer review date would allow more publications, but all of these products were derived from the final reports, which were submitted on average five years after peer review. Thus, whatever is reported by the principal investigator in their final report is what was measured, which is what is presented by funding agencies as the output of the grant.2. The number of citations was standardised by academic field – i.e. molecular biology. Was there any variation in study designs within this field that might also change the expected number of citations (e.g. systematic reviews may attract more citations than primary experimental studies in some fields)? It would be useful for context if there could be a table with some basic aggregate details of the funded studies, such as types of study design, molecular biological application, study sample characteristics, duration of study etc. This would help to put the results into context.This is interesting, but would likely need to be a much larger, separate study to go back into each application and characterize the study design, sample characteristics, etc., and is out of the current scope. That said, the applications included in this analysis were submitted to a 4-year, R01-style support mechanism. In every program year there was a significant proportion of both applied and basic research applications, with many applications encompassing varying degrees of both basic and applied research in their aims. There was a wide variety of research topic areas, including vision, drug abuse, nutrition, blood-related cancer, kidney disease, autoimmune diseases, malaria, tuberculosis, osteoporosis, arthritis, and autism research, among others. This is described in more detail in our original publication: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0106474.3. The impact of a piece of research on which referees could not agree might be either lower or higher than those on which they could. So, it would be useful to plot the standard deviation of the citations against the standard deviation of the peer review score.We have now plotted the inter-quartile range in citations against the application score mean and standard deviation. We used the inter-quartile range instead of the standard deviation because of the strong skew in citations, and the standard deviation was strongly influenced by the application with the highest citations.The new results show moderate support for a greater return for applications with a higher agreement between peer reviewers. See new Figure 4, with new explanatory paragraph in the “Statistical Methods” section, and an updated discussion.4. As well as using multiple imputation to correct for missing data, a sensitivity analysis in which cases with missing data are omitted would be useful.We have added a complete case analysis to the more detailed results available on github (https://github.com/agbarnett/funding.disagree). All the results were very similar to the imputed data.5. Some measure of the model fit would be useful, eg. adjusted R squaredWe are not certain that the R-squared statistic would be useful here. The adjusted R-squared can be useful for comparing between alternative models, or where the goal is accurate prediction. Here we were just looking for any signal between application scores and citations, assuming that the detection of any signal could lead to improvements in the design of funding systems. We have investigated the residuals to look for poor model fit.6. There was a wide variation in the number of reviewers per article from 2 to 18. Was this due to differences in the kind of research, amount of funding requested or some other perceived risk on behalf of the funder? Could this artificially influence the standard deviation of the score, confounding any association with citations?The majority (90%) of applications were reviewed in panels (roughly 10 to 15 reviewers). Some were reviewed via a mail review mechanism (2 to 3 reviewers). The determination of review mechanism largely depended on the topic area. We don’t believe this should be a large confounder of the results.We have now included a sensitivity analysis using just the larger panels and excluding the mail review panels. The results were similar to the analysis using all panels. These new results are available on github (https://github.com/agbarnett/funding.disagree).7. Fractional polynomial model results are only presented for the best fitting model with the smallest deviance. This is acceptable in principle, but it would be useful for the authors to comment on whether there was any variation in the results according to the other fractional polynomial transformations (if available). This will provide confidence in the robustness of the findings.We have examined the predictions for the best five models and the average results were similar in terms of the mean. However, the confidence intervals for the second best model had a much narrow confidence interval for citation predictions using the application score standard deviation. In hindsight we prefer this model, as it has only a slightly poorer deviance, but a much more believable confidence interval. It would have been better to have a multi-criteria decision for the “best” model, that used both the deviance and average confidence interval width. This has been a useful lesson for future studies, but we continue to present the model with the wider CI in the main results because we feel we should stick to our original protocol.These new results are available on github (https://github.com/agbarnett/funding.disagree).8. Reference is made in the first paragraph to a “recent systematic review” by Guthrie et al (2018 1) (and also in the third paragraph). We note that this publication doesn’t refer to itself as being a systematic review, and indeed, it is an update of a 2009 review which describes itself as a non-systematic review. We would suggest using the term “non-systematic review”, or just “literature review”.Thank you for flagging this. We have now used “literature review”.9. Thank you for citing our own recent systematic review on peer review of grants in health. You mention that the review included eight studies and called for further research in this area, which is correct. However, the review focused specifically on studies aiming to improve the effectiveness and efficiency of peer review. These were drawn from a wider set of 83 studies on peer review which we systematically mapped. In the map there were some studies which focused on assessing the impact of funded research, eg. In terms of bibliometrics. Thus, there is a body of evidence on this topic, though we didn't systematically review it in detail. We are happy to provide you with a list of these studies.We have changed our language to make the goal of your review clearer. We would gladly see the study list. Perhaps this is best uploaded to our github page for this project (if it can be made public): https://github.com/agbarnett/funding.disagree then click “upload file”, you will need a github account (free) and request write access to our project.10. The sentence on page 3 beginning “A recent systematic review found “suggestive” evidence that funding peer review can have an anti-innovation bias2 and that innovation and risk may not often be sufficiently addressed in review feedback7” needs re-wording as not only is the Guthrie et al paper a non-systematic review, but the second reference cited in that sentence by Gallo et al is a survey (i.e. not a review at all). The way the sentence is phrased implies that it is a systematic review.Thank you for flagging this. We have now used “literature review” and added the fact that the second study is a survey.11. Suggest amending the sentence on page 3 “Many studies using large sample sizes found either no association or only a weak association between the mean score and the VOLUME of citations of subsequent publications”We have changed the text to read “number of citations”." } ] }, { "id": "37261", "date": "24 Sep 2018", "name": "Shahar Avin", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe notion that disagreements amongst reviewers might indicate a promising high-risk/high-reward project certainly has currency amongst those who consider how the system might be improved. In addition to the framework of Linton (2016)1, the idea is also present in the model of Brezis (2007) and is implied by several works on the anti-innovation bias of grant peer review (e.g. Greenberg, 19982; Gillies, 20083). However, this intuitive notion has yet to be put to the test, until the current study, which is a welcome contribution.\n\nThe experimental design and statistical analysis are compelling for an initial study, though as the authors note further complications could be addressed in later work.\n\nGiven the prevailing policy and academic discussions the negative result is surprising and it rules out some suggested approaches to reform of the grant allocation system; however, the dataset limits the generality of the conclusions that can be drawn (through no fault of the authors) – the key limitation being that only 11% of applications can be analysed (the funded ones), a set that is heavily skewed towards the top scoring applications.\n\nFor example, as shown in Figure 4 there is plenty of potential for SD to be associated with increased citation, but not enough data to tell. The authors note that the application with the largest SD drove this large uncertainty – it could be that this application is an outlier, or it could be that it is typical of applications with large SDs but there are very few of them in the dataset because they tended to fall below the funding line.\n\nThrough its analysis, the study forces a refinement of the question of what sorts of disagreements (in extent and driven by what underlying logics) should be considered useful indicators. We suggest the study raises the following THREE questions for future work:\n\nFirstly it would be valuable to understand the reasons and logics behind the disagreements that arise between reviewers, from the fact that \"disagreements between reviewers about an application can stem from different sources\". While the authors suggest addressing this with further information about scores, we can also ask if different score means might indicate different \"regimes\" for disagreement, e.g. disagreement along the lines of \"not sure this will work\" or \"this goes against received wisdom\" at relatively high score means, and disagreements along the lines of \"not sure what this is about\" or \"hasn't this been tried before?\" at lower score means. Qualitative research asking reviewers about their scoring behaviour could help understand the reasoning that goes with different scoring.\n\nSecondly, we could reframe the question to address the lack of information on unfunded applications by asking how can funding best be allocated if the amount of funding available is cut by 50%.\n\nTo test this, we reanalysed the data provided by the authors with the following question: assuming only half the funding was available, which of the following selection methods would perform better?\nRank proposals based on their mean scores until funding runs out. Pool proposals based on their mean score into buckets, rounding to the nearest integer. Fund all proposals with a mean of 1 (1.0-1.49, n=45), then from the bucket with a mean of 2 (1.5-2.49, n=114) rank proposals based on their standard deviation and fund until funding runs out.\n\nUnder both methods, all proposals with a mean of 1 are funded, for a total TRC of 319.9. When we look at the proposals in the second bucket, when funding according to method 1, the total TRC for this bucket is 224.2, whereas for method 2 it is 195.5. However, the highest citation-receiving proposal in the second bucket under method 1 has a TRC of 19.9, whereas the highest citation-receiving proposal from the second bucket under method 2 has a TRC of 35.9 (the third highest TRC in the entire sample). This fits with Gillies' description of peer-review as going for less risky proposals, but at the cost of throwing out the occasional exceptional proposal. This is also born out by the distribution of TRCs, with the standard deviation of TRCs for the second bucket under method 1 being 4.2, and under method 2 being 6.2.\n\nThirdly, it would be valuable to understand what sorts of disagreement exist in the scoring of all the applications (both successful and unsuccessful). It could be argued that the type of disagreement likely to indicate high risk/high return proposals would be bi-modal – some good reviews maybe with scores of 1-1.5, some very poor reviews maybe with scores of 3-4. Given the small fraction of applications funded, very few applications of this nature would be funded (due to the understandable aggregation in the shared data we could not examine the exact number, but fewer than 10 applications have a score lower than 3). Do such applications with such bi-modal review scores exist? Or is the level of disagreement seen in the successful applications similar to that of the unsuccessful applications?\n\nThe above questions only suggest that there is further complexity here that needs to be explored, with more complex, higher power studies. We thank the authors for paving the path for such studies, and for making their underlying data and analysis available in a form that is easy to explore.\nTwo small notes:\nIt might be easier to read the paper if scores were consistently referred to as best/worst and better/worse – using ‘higher' is confusing when better scores are lower. It would be more elegant to give 'e.g., Bornmann et al, 2008' rather than 'e.g., 36'.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes\n\nIs the study design appropriate and is the work technically sound? Yes\n\nAre sufficient details of methods and analysis provided to allow replication by others? Yes\n\nIf applicable, is the statistical analysis and its interpretation appropriate?\nYes\n\nAre all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes", "responses": [ { "c_id": "4055", "date": "15 Oct 2018", "name": "Adrian Barnett", "role": "Author Response", "response": "The notion that disagreements amongst reviewers might indicate a promising high-risk/high-reward project certainly has currency amongst those who consider how the system might be improved. In addition to the framework of Linton (2016)1, the idea is also present in the model of Brezis (2007) and is implied by several works on the anti-innovation bias of grant peer review (e.g. Greenberg, 1998 2; Gillies, 2008 3). However, this intuitive notion has yet to be put to the test, until the current study, which is a welcome contribution. Thanks for the interesting references, we have now mentioned Brezis in our Introduction. The experimental design and statistical analysis are compelling for an initial study, though as the authors note further complications could be addressed in later work. Given the prevailing policy and academic discussions the negative result is surprising and it rules out some suggested approaches to reform of the grant allocation system; however, the dataset limits the generality of the conclusions that can be drawn (through no fault of the authors) – the key limitation being that only 11% of applications can be analysed (the funded ones), a set that is heavily skewed towards the top scoring applications. We agree this is a key limitation, and it would be useful to repeat the study in a more generous scheme, where a greater proportion of applications were funded. For example, as shown in Figure 4 there is plenty of potential for SD to be associated with increased citation, but not enough data to tell. The authors note that the application with the largest SD drove this large uncertainty – it could be that this application is an outlier, or it could be that it is typical of applications with large SDs but there are very few of them in the dataset because they tended to fall below the funding line. This is entirely possible, and if there are a small number of extremely high pay-off projects then we would need a much larger sample to robustly test if there is a consistent pattern of more high pay-off projects associated with greater disagreement. We have now mentioned this in the limitations. Through its analysis, the study forces a refinement of the question of what sorts of disagreements (in extent and driven by what underlying logics) should be considered useful indicators. We suggest the study raises the following THREE questions for future work: Firstly it would be valuable to understand the reasons and logics behind the disagreements that arise between reviewers, from the fact that \"disagreements between reviewers about an application can stem from different sources\". While the authors suggest addressing this with further information about scores, we can also ask if different score means might indicate different \"regimes\" for disagreement, e.g. disagreement along the lines of \"not sure this will work\" or \"this goes against received wisdom\" at relatively high score means, and disagreements along the lines of \"not sure what this is about\" or \"hasn't this been tried before?\" at lower score means. Qualitative research asking reviewers about their scoring behaviour could help understand the reasoning that goes with different scoring. This is a good point. “Disagreement” is likely to be too general a word, and future studies could look in far more detail at the types of disagreement. We have previously used a qualitative researcher as an observer of funding panel dynamics, and we have now added this potential approach to our discussion. Secondly, we could reframe the question to address the lack of information on unfunded applications by asking how can funding best be allocated if the amount of funding available is cut by 50%. To test this, we reanalysed the data provided by the authors with the following question: assuming only half the funding was available, which of the following selection methods would perform better? Rank proposals based on their mean scores until funding runs out. Pool proposals based on their mean score into buckets, rounding to the nearest integer. Fund all proposals with a mean of 1 (1.0-1.49, n=45), then from the bucket with a mean of 2 (1.5-2.49, n=114) rank proposals based on their standard deviation and fund until funding runs out. Under both methods, all proposals with a mean of 1 are funded, for a total TRC of 319.9. When we look at the proposals in the second bucket, when funding according to method 1, the total TRC for this bucket is 224.2, whereas for method 2 it is 195.5. However, the highest citation-receiving proposal in the second bucket under method 1 has a TRC of 19.9, whereas the highest citation-receiving proposal from the second bucket under method 2 has a TRC of 35.9 (the third highest TRC in the entire sample). This fits with Gillies' description of peer-review as going for less risky proposals, but at the cost of throwing out the occasional exceptional proposal. This is also born out by the distribution of TRCs, with the standard deviation of TRCs for the second bucket under method 1 being 4.2, and under method 2 being 6.2. This is an interesting comment and thanks for the new analysis. As an aside, we note that only by sharing our data can we have such an in-depth discussion with our reviewers. We have added R code to run these suggested analyses to our results on github (https://github.com/agbarnett/funding.disagree). It is an interesting approach to use the mean for the “top” tier of applications and the standard deviation for the second tier. This may appeal to funders and panel members, because we believe there will always be a desire to fund those applications with the best mean. This two-tier approach may also be easier to explain than the Black–Scholes approach suggested by Linton, which combines the mean and variance in scores using an equation. We are wary of using the standard deviation in the total relative citations because of the strong positive skew in its distribution, and the inter-quartile range in citations is actually larger for the approach using the mean at 4.3, compared with 2.9 when using the standard deviation. Thirdly, it would be valuable to understand what sorts of disagreement exist in the scoring of all the applications (both successful and unsuccessful). It could be argued that the type of disagreement likely to indicate high risk/high return proposals would be bi-modal – some good reviews maybe with scores of 1-1.5, some very poor reviews maybe with scores of 3-4. Given the small fraction of applications funded, very few applications of this nature would be funded (due to the understandable aggregation in the shared data we could not examine the exact number, but fewer than 10 applications have a score lower than 3). Do such applications with such bi-modal review scores exist? Or is the level of disagreement seen in the successful applications similar to that of the unsuccessful applications? Unfortunately our data only contains the summary statistics on the applications’ scores and so we can’t examine these distributions. We have added this as a limitation. The above questions only suggest that there is further complexity here that needs to be explored, with more complex, higher power studies. We thank the authors for paving the path for such studies, and for making their underlying data and analysis available in a form that is easy to explore. Thanks, we agree that larger studies are needed. Two small notes: It might be easier to read the paper if scores were consistently referred to as best/worst and better/worse – using ‘higher' is confusing when better scores are lower. Agreed and changed. We’ve also added text labels to Figure 5 and the new Figure 4. It would be more elegant to give 'e.g., Bornmann et al, 2008' rather than 'e.g., 36'. Changed as suggested." } ] } ]	1	https://f1000research.com/articles/7-1030
https://f1000research.com/articles/7-1636/v1	12 Oct 18	{ "type": "Method Article", "title": "A unified GenomeSpace recipe to identify essential genes and associated subnetworks from Genome-Scale CRISPR-Cas9 knockout screens", "authors": [ "Daniel E. Carlin", "Forrest Kim", "Trey Ideker", "Jill P. Mesirov", "Forrest Kim", "Trey Ideker", "Jill P. Mesirov" ], "abstract": "We present a unified GenomeSpace recipe that combines the results of a high throughput CRISPR genetic screen and a biological network to return a subnetwork that suggests a mechanistic explanation of the screen’s results. The explanatory subnetwork is found by network propagation, a popular systems biology approach. We demonstrate our pipeline on an alpha toxin screen, revealing a subnetwork that is both highly interconnected and highly enriched for hits in the screen.", "keywords": [ "genetic screen", "network biology", "network propagation", "CRISPR" ], "content": "Introduction\n\nThe rise of next generation sequencing technology and CRISPR gene editing technology has opened up new opportunities for high throughput genetic screens. Increasingly systems biology and molecular networks are becoming more important in the analysis of the mechanisms that are implicated in these screens. Here we present a GenomeSpace recipe providing a standardized pipeline for combining the analysis of a screen with networks and represents a logical next step in providing user-friendly bioinformatics workflows for these types of screens.\n\nThis recipe provides a way to process the results of a CRISPR-Cas9 genome wide knockout screen. In such a screen, single guide RNAs (sgRNAs) are designed to target and knock down genes by binding to the target gene and introducing double strand DNA breaks. (Koike-Yusa et al., 2014). Those bound mRNA are subsequently digested by the Cas9 complex and thus do not yield a gene product. In a cell, if the sgRNA is introduced for a gene that is essential for the survival of the cell, that cell will die, and the sgRNA will be depleted. Thus, by sequencing the sgRNAs and looking for a depletion of the sgRNAs targeting a particular gene, we can infer the essentiality of that target gene. Since a large number of sgRNAs can be introduced in a single screen, the essentiality of many genes can be tested at once. However, there are challenges that can arise in the normalization and processing of the read counts; often more than one sgRNA corresponds to a gene but with different efficiency, and significant biases exist in sequencing different sgRNAs. For these reasons the MAGeCK (Li et al., 2014) method was developed to handle data resulting from such a screen.\n\nOn the systems biology side of the analysis, we have chosen to employ network propagation as a method of identifying subnetworks representing inferred mechanisms that are implicated by the CRISPR screen. Network propagation has become an essential tool in many network applications; it has been used to identify mechanisms of cancer (Leiserson et al., 2015), to implicate genes in GWAS studies (Qian et al., 2014), and to find functional modules (Vanunu et al., 2010). Network propagation considers genes as nodes on the graph of a biological network. It performs a random walk along the edges of the graph from a set of query nodes. We expect that genes that are implicated in a phenotype will occur in regions of the network that represent mechanisms that are relevant to the screen conditions, and so the random walk will be likely to land on relevant genes. Genes that are near query nodes are therefore implicated by association. For a review of the many flavors and applications of network propagation, see (Cowen et al., 2017).\n\n\nMethods\n\nAn overview of the pipeline appears in Figure 1. The recipe begins using the raw read counts of sgRNAs as input to the MAGeCK module in GenePattern (www.genepattern.org). After normalizing the data, MAGeCK detects differential read counts for each sgRNA using an over-dispersed Poisson model. Next, it detects statistical underrepresentation of the sgRNAs corresponding to particular genes to infer that a gene is essential for the survival of a cell. The reasoning behind this is that if a sgRNA targets an essential gene, the cells that contain that sgRNA will not replicate and the sgRNA will be underrepresented compared to other genes.\n\nThis recipe shows how using GenomeSpace seamlessly integrates multiple bioinformatics tools into a single, easily reproducible pipeline. The publicly available preprocessed knockout screen files and sgRNA library from Koike-Yusa et al. can be transferred directly from GenomeSpace to the ported MAGeCK module in GenePattern (Li et al., 2014). By exporting the resulting list of significant genes to GenomeSpace, the data can be imported directly to Cytoscape without having to download the files locally. Cytoscape’s integrated plugins for NDEx (Pratt et al., 2015), Network Diffusion (Carlin et al., 2017; Cowen et al., 2017), and GeneMANIA (Montojo et al., 2010) allow for the remainder of the recipe to completed within its user-friendly environment.\n\nAfter we determine a set of essential genes, we pass their identity via GenomeSpace to Cytoscape (Shannon et al., 2003). All analysis in Cytoscape is based on networks, thus a relevant reference molecular network must be imported; the recipe uses the NDEx database (Pratt et al., 2015) to identify such a network. In this case, we choose the National Cancer Institute’s Pathway Interaction Database (Schaefer et al., 2009). The set of essential genes, i.e., the hits from the genetic screen, are imported from GenomeSpace as a table, then used as the seed nodes for network propagation.\n\nThe propagation process starts with a single unit of “heat” on each of the nodes that represent the genes that are found to be underrepresented in the screen, and therefore essential for the growth of the cell. We use a heat diffusion process, treating the network as an unweighted, undirected graph. Heat diffusion smooths the original signal over the network, iteratively passing the signal on each node to its neighbor. It identifies regions of the network that have a high concentration of hits. Here we use a time parameter (which represents the amount of time that the heat is allowed to diffuse over the graph) of 0.1. This is a common choice for the time parameter (see Paull et al., 2013). The recipe employs the network diffusion service built into Cytoscape natively (Carlin et al., 2017).\n\nNext, applying a cutoff of the top 200 genes with the most heat after diffusion, we choose a subnetwork that has a high concentration of hits. Finally, in order to understand the composition of the subnetwork, we apply the GeneMANIA Cytoscape plugin. For any network, GeneMANIA (Montojo et al., 2010) shows what functional Gene Ontology categories are enriched in that network.\n\n\nUse case\n\nWe used a previously published CRISPR study (Koike-Yusa et al., 2014) to illustrate the use of this pipeline. In this study, the authors use mouse embryonic stem cells grown in the presence of alpha-toxin. This screen was therefore designed to expose the genes involved in the mechanism of resistance to the toxin. The largest connected network component of the top 200 genes after propagation appears in Figure 2.\n\nThe black nodes represent genes that are significantly deplete in a CRISPR screen. Grey nodes represent genes that are closely associated with the hits by the network and are scaled by the strength of their association.\n\nThe results of the GeneMania enrichment suggest that DNA repair is the single most important gene set in handling alpha-toxin. This is consistent with the findings in (Bantel et al., 2001) that alpha-toxin causes an influx of monovalent ions that can cause DNA fragmentation. In the absence of DNA repair machinery, the cells cannot recover from this stress and therefore die. The complete table of the Gene Ontology terms that were significantly enriched in the subnetwork appear in Table S1.\n\n\nVariations\n\nThere are several variations that can be used depending on preferences of the user. For example, different tools such as DESeq (Anders & Huber, 2010) and edgeR (Robinson et al., 2010) can be used to identify the hits. Also, the final biological interpretation of the subnetworks was performed with the GeneMania plugin, but the gene list can also be exported and interpreted by another annotation tool. Another approach is to export the gene list and corresponding heats using GenomeSpace and use the Molecular Signature Database gene set overlap tool (http://software.broadinstitute.org/gsea/msigdb/index.jsp), (Liberzon et al., 2011) applied to the genes in the identified subnetwork.\n\n\nData and software availability\n\nThe recipe and Koike-Yusa et al. datasets are publicly available at http://recipes.genomespace.org/view/75. GenomeSpace, an open-source bioinformatics tool, serves as the data highway allowing for seamless transfer of information between tools, and can be found at http://www.genomespace.org/. The MAGeCK algorithm has been wrapped as a GenePattern module, which can be run locally or on the public GenePattern servers at http://genepattern.org. Additionally, GenePattern has added Jupyter Notebook compatibility through GenePattern Notebook (http://genepattern-notebook.org/). Finally, Cytoscape and all associated plugins (ie. GeneMANIA and NDEx) can be found at http://www.cytoscape.org/.", "appendix": "Grant information\n\nThis work is supported by the National Institute of Health and National Human Genome Research Institute project number 5U41HG007517-05.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary material\n\nTable S1: Enriched Gene Ontology categories for the network appearing in Figure 2.\n\nClick here to access the data\n\n\nReferences\n\nAnders S, Huber W: Differential expression analysis for sequence count data. Genome Biol. 2010; 11(10): R106. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nBantel H, Sinha B, Domschke W, et al.: alpha-Toxin is a mediator of Staphylococcus aureus-induced cell death and activates caspases via the intrinsic death pathway independently of death receptor signaling. J Cell Biol. 2001; 155(4): 637–48. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCarlin DE, Demchak B, Pratt D, et al.: Network propagation in the cytoscape cyberinfrastructure. PLoS Comput Biol. 2017; 13(10): e1005598. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCowen L, Ideker T, Raphael BJ, et al.: Network propagation: a universal amplifier of genetic associations. Nat Rev Genet. 2017; 18(9): 551–562. PubMed Abstract \| Publisher Full Text\n\nKoike-Yusa H, Li Y, Tan EP, et al.: Genome-wide recessive genetic screening in mammalian cells with a lentiviral CRISPR-guide RNA library. Nat Biotechnol. 2014; 32(3): 267–73. PubMed Abstract \| Publisher Full Text\n\nLeiserson MD, Vandin F, Wu HT, et al.: Pan-cancer network analysis identifies combinations of rare somatic mutations across pathways and protein complexes. Nat Genet. 2015; 47(2): 106–14. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLi W, Xu H, Xiao T, et al.: MAGeCK enables robust identification of essential genes from genome-scale CRISPR/Cas9 knockout screens. Genome Biol. 2014; 15(12): 554. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLiberzon A, Subramanian A, Pinchback R, et al.: Molecular signatures database (MSigDB) 3.0. Bioinformatics. 2011; 27(12): 1739–40. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nMontojo J, Zuberi K, Rodriguez H, et al.: GeneMANIA Cytoscape plugin: fast gene function predictions on the desktop. Bioinformatics. 2010; 26(22): 2927–8. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPaull EO, Carlin DE, Niepel M, et al.: Discovering causal pathways linking genomic events to transcriptional states using Tied Diffusion Through Interacting Events (TieDIE). Bioinformatics. 2013; 29(21): 2757–64. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nPratt D, Chen J, Welker D, et al.: NDEx, the Network Data Exchange. Cell Syst. 2015; 1(4): 302–5. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nQian Y, Besenbacher S, Mailund T, et al.: Identifying disease associated genes by network propagation. BMC Syst Biol. BioMed Central, 2014; 8 Suppl 1: S6. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nRobinson MD, McCarthy DJ, Smyth GK: edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 2010; 26(1): 139–40. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSchaefer CF, Anthony K, Krupa S, et al.: PID: the pathway interaction database. Nucleic Acids Res. 2009; 37(Database issue): D674–9. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nShannon P, Markiel A, Ozier O, et al.: Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003; 13(11): 2498–504. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVanunu O, Magger O, Ruppin E, et al.: Associating genes and protein complexes with disease via network propagation. PLoS Comput Biol. 2010; 6(1): e1000641. PubMed Abstract \| Publisher Full Text \| Free Full Text" }	[ { "id": "49975", "date": "24 Jun 2019", "name": "Shouhong Guang", "expertise": [ "Reviewer Expertise Genetics", "epigenetics", "small RNAs", "C. elegans." ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe CRISPR gene editing technology followed by next generation sequencing technology has brought up new means for high throughput genetic screening. However, a streamlined and systematic method for downstream data analysis is required to examine these high throughput screenings. In this work, the authors provide a pipeline for combining the analysis of a screen with network illustrations that facilitate the assay of this type of screens. The work is timely and applicable for a lot of related research.\n\nI am not an expert in bioinformatics. Therefore, I will provide several comments from the biological side.\nIn a CRISPR-mediated high throughput genetic screening, people frequently apply several sgRNAs to target a single gene, which exhibit different knocking out efficiencies. In addition, these sgRNAs may target distinct isoforms of a gene, which may lead to different knocking out phenotypes. How the pipeline deals with these sophisticated cases requires elaboration. The author used one example, the alpha toxin resistance screen, to validate their method. The analysis of another independent screen is needed to test its validility. During the analysis of alpha-toxin screen, the authors conclude that the result is consistent with previous findings. This argument requires more detailed comparison with previous work. In addition, if possible, at least some of the new hits coming out of the work could be validated by wet lab experiments.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Partly\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Partly", "responses": [] }, { "id": "52690", "date": "03 Sep 2019", "name": "Xiaowei Wang", "expertise": [ "Reviewer Expertise Bioinformatics", "Genomics", "miRNA", "CRISPR" ], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors present a bioinformatic recipe for network-based functional analysis of CRISPR screening data. CRISPR screening has quickly become a mainstream tool for functional genomics studies. However, bioinformatics tools are lacking in this emerging field. This study provides a timely bioinformatic framework for functional analysis of CRISPR screening data. This new pipeline combines multiple existing bioinformatics tools and provides a streamlined data analysis process for general users, especially for those who are not experts in bioinformatics. To demonstrate the utility of the pipeline, the authors present a specific example by analyzing a public dataset. Overall, this study is well described. The website also provides very detailed step-by-step instruction for users to follow the recipe. I have a couple of suggestions for further improvement.\n\nIt would be helpful to provide additional details on individual components of the recipe, such as the rationale behind adopting the widely-used MAGeCK tool (i.e. summarizing its advantages) for robust identification of sgRNA hits. Similarly, there are multiple tools available for network propagation, and it is not clear why a heat diffusion process is adopted by this recipe (i.e. summarizing its advantages).\n\nIt would be helpful to implement version control for this presented recipe. It is likely that new bioinformatics tools, such as new versions of MAGeCK (or similar tools) or improved strategies for network propagation, will be available in the near future. Accordingly, this recipe needs to be updated to accommodate the latest progress in the bioinformatics field.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Yes\n\nAre sufficient details provided to allow replication of the method development and its use by others? Partly\n\nIf any results are presented, are all the source data underlying the results available to ensure full reproducibility? Yes\n\nAre the conclusions about the method and its performance adequately supported by the findings presented in the article? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-1636
https://f1000research.com/articles/7-75/v1	17 Jan 18	{ "type": "Software Tool Article", "title": "Explicit interaction information from WikiPathways in RDF facilitates drug discovery in the Open PHACTS Discovery Platform", "authors": [ "Ryan A. Miller", "Peter Woollard", "Egon L. Willighagen", "Daniela Digles", "Martina Kutmon", "Antonis Loizou", "Andra Waagmeester", "Stefan Senger", "Chris T. Evelo", "Peter Woollard", "Egon L. Willighagen", "Daniela Digles", "Martina Kutmon", "Antonis Loizou", "Andra Waagmeester", "Stefan Senger", "Chris T. Evelo" ], "abstract": "Open PHACTS is a pre-competitive project to answer scientific questions developed recently by the pharmaceutical industry. Having high quality biological interaction information in the Open PHACTS Discovery Platform is needed to answer multiple pathway related questions. To address this, updated WikiPathways data has been added to the platform. This data includes information about biological interactions, such as stimulation and inhibition. The platform's Application Programming Interface (API) was extended with appropriate calls to reference these interactions. These new methods of the Open PHACTS API are available now.", "keywords": [ "Open PHACTS", "drug discovery", "semantic", "bioinformatics", "WikiPathways", "pathway database", "API" ], "content": "Introduction\n\nTargeting proteins to ideally restore normal biological processes is a common starting point in drug discovery1. The Open PHACTS Discovery Platform (OPDP) was designed to help identify protein targets and information about their associations with each other2–4. The OPDP supports target identification and validation by including target-target interactions from WikiPathways5–7. Of these interaction networks, proteins sharing a downstream path allows investigation of alternative drug target combinations. Even the knowledge of which biological pathways participate in disease-related processes provides insight in the pathway topology between the targets. The importance and need of providing access to interaction information for real-world research questions was outlined in a recent Open PHACTS paper8.\n\nThe Open PHACTS project was born out of the desire to integrate pharmacological data from multiple pre-competitive sources to efficiently address scientific questions that cannot be answered with single data sources8. It integrates data using linked data approaches3 from chemical and biological sources such as ChEBI, ChEMBL, UniProt, and WikiPathways6. However, the OPDP did not previously include calls to access specific up- and downstream interaction effects. This information is needed for questions related to drug repositioning and repurposing. Up- or downstream targets may be interesting alternatives with similar therapeutic effect to targets, for which it is particularly hard to develop an drug agent. Thus, finding a target that has already been drugged or is more drug tractable will be advantageous. Here we describe how to identify alternative targets in the same cellular pathway using OPDP against the WikiPathways data.\n\n\nMethods\n\nThe WikiPathways Resource Description Framework data (WPRDF) is released as part of the monthly releases5. It includes details about directed and undirected interactions. Directed biochemical interactions capture the source and target which are depicted as an arrow in simple pathway drawings. WikiPathways adds biological meaning to interactions with Molecular Interaction Map (MIM) interaction types, like inhibitions, enzyme catalyzed reactions, and stimulations9, as well as Systems Biology Graphical Notation (SBGN) interactions10. Reactome pathways in WikiPathways use SBGN interactions11,12. However, because MIM and SBGN use different drawing styles, we normalize their inhibition types into a common inhibition type, defined by the WikiPathways ontology (https://vocabularies.wikipathways.org/wp).\n\nThe WikiPathways basic drawing tools also contain generic arrows and t-bar annotations that give the user the ability to create basic diagrams without the semantic meaning of MIM or SBGN notations. The interactions connecting these nodes are captured, but the only explicit information is that it is a directed interaction from a source to a target. To handle more complicated enzyme reaction drawings, where there is not a single line that directly connects targets in a cascade of enzymatic reactions, a query was developed that recognizes these types of reactions. However, this is not implemented in the current Open PHACTS Application Programming Interface (API).\n\nVersion 2.1 of the OPDP API contains three new calls for interactions and their pathways. The first call, /pathway/getInteractions, returns all interactions involved in a pathway. To use this feature, the user specifies a pathway URI and OPDP returns its interactions including information about direction and the connected entities. The direction information is relayed as a starting node having a wp:source annotation, while the end of the interaction has the wp:target annotation. In its simplest form, this means that if gene product A is interacting with a gene product B, then we have wp:source for product A and wp:target for product B. However, the new methods also support interactions with multiple sources and targets.\n\nThe second added call, /pathways/interactions/byEntity, returns the direction of the interactions involving this entity. An entity is specified by a URI and can be a metabolite, protein, gene product, or RNA. API options allow the user to select only upstream or only downstream interactions. The results also specify the interaction type (e.g. inhibition, stimulation, conversion). This ability to select the interaction direction is specifically what allows users to answer scientific questions around upstream and downstream effects, such as those defined by Open PHACTS. The third API call is /pathways/interactions/byEntity/count which is a helper function that returns the number of interactions for a target.\n\nThe OPDP API calls are backed by SPARQL searches against the loaded WikiPathways RDF. The query parameters that are required or optional are given in the documentation of Open PHACTS (https://dev.openphacts.org/docs/2.1). As in previous versions, the API uses HTTP GET to call methods and needs a (free) application ID and key3.\n\n\nExample queries\n\nWe are demonstrating the platform with three example calls. All the API calls require use of an application ID and an application key. This key and ID can be acquired by creating a free Open PHACTS account. The first example is an application to the PI3K/AKT pathway for cell growth regulation which contain important targets for cancer treatment13. The AKT protein has a central role and usefully shows the API call’s ability to return connected elements with the first and third calls. The API call can help aid drug discovery by taking a target, in this case AKT, and easily identify other connected proteins that could potentially be used as drug targets with a common downstream effect.\n\nFigure 1 shows the web interface of the API call that returns the connectivity of the AKT2 target to both upstream or downstream proteins or gene products. This method allows the user to identify connections to other targets in the pathway. The results of that API call (Figure 2) show the AKT2 interaction with microRNA. A helper method (Figure 3): /pathways/interactions/byEntity/count is also included. It returns the number of all interactions in which an entity is participates. This helps the user get a sense of the prevalence of the queried entity with interactions in pathways found on WikiPathways.\n\nThe GET portion tells the API to retrieve data with the associated call. It takes an entity URI, the Ensembl ID for AKT2, and returns a list interactions for AKT2. The obligatory parameters are shown in bold.\n\nThe participants of the interaction are directed from source (hsa-let7b) to target (AKT2). It also shows the type of interaction (inhibition), and the biological types of the interaction participants.\n\nIt takes a URI for an entity, in this case the Ensembl ID for AKT2 and returns a count of the interactions to which this gene product is involved. Only the entity URI, app ID, and app key are required fields. Optional parameters are pathway organism, direction, or type of interaction.\n\nThe other call implemented, /pathway/getInteractions (Figure 4), demonstrates an API call to return all interactions in the MicroRNAs in cardiomyocyte hypertrophy pathway14. This pathway has interaction details for AKT, mTOR, and PI3K, which are all important targets in cancer research15. For each interaction the participants are given and whether it is a directed or undirected interaction.\n\nIt is intended to take the pathway URI from WikiPathways and return a list of interaction involved in that particular pathway. Pathway URI, app ID, and app key are the only required values for this call.\n\n\nSummary\n\nThe addition of interactions with direction information allows OPDP to answering more of the pre-defined scientific questions16. The directional information allows the user to explore how proteins and gene products are connected with one another and easily access this information. This is illustrated in the example queries using the cancer target AKT.\n\n\nSoftware availability\n\nOnline service: https://dev.openphacts.org/docs/2.1\n\nLatest source code is available at: https://github.com/openphacts/OPS_LinkedDataApi\n\nArchived source code of discussed version: https://doi.org/10.5281/zenodo.106825217\n\nLicense: Apache License 2.0", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis work was supported by the Innovative Medicines Initiative Joint Undertaking [grant number 115191], resources of which are composed of financial contribution from the European Union’s Seventh Framework Programme (FP7/2007-2013) and EFPIA companies’ in kind contribution.\n\n\nAcknowledgments\n\nA special thanks goes to all members of the Open PHACTS project that provided the platform that was necessary.\n\n\nReferences\n\nSchreiber SL: Target-oriented and diversity-oriented organic synthesis in drug discovery. Science. 2000; 287(5460): 1964–1969. PubMed Abstract \| Publisher Full Text\n\nAzzaoui K, Jacoby E, Senger S, et al.: Scientific competency questions as the basis for semantically enriched open pharmacological space development. Drug Discov Today. 2013; 18(17–18): 843–852. PubMed Abstract \| Publisher Full Text\n\nWilliams AJ, Harland L, Groth P, et al.: Open PHACTS: semantic interoperability for drug discovery. Drug Discov Today. 2012; 17(21–22): 1188–1198. PubMed Abstract \| Publisher Full Text\n\nDigles D, Zdrazil B, Neefs JM, et al.: Open PHACTS computational protocols for in silico target validation of cellular phenotypic screens: knowing the knowns. Medchemcomm. 2016; 7(6): 1237–1244. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWaagmeester A, Kutmon M, Riutta A, et al.: Using the Semantic Web for Rapid Integration of WikiPathways with Other Biological Online Data Resources. PLoS Comput Biol. 2016; 12(6): e1004989. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGray AJG, Groth P, Loizou A, et al.: Applying linked data approaches to pharmacology: Architectural decisions and implementation. Semant Web. 2014; 5(2): 101–113. Publisher Full Text\n\nKelder T, Pico AR, Hanspers K, et al.: Mining biological pathways using WikiPathways web services. PLoS One. 2009; 4(7): e6447. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nChichester C, Digles D, Siebes R, et al.: Drug discovery FAQs: workflows for answering multidomain drug discovery questions. Drug Discov Today. 2015; 20(4): 399–405. PubMed Abstract \| Publisher Full Text\n\nLuna A, Karac EI, Sunshine M, et al.: A formal mim specification and tools for the common exchange of mim diagrams: an xml-based format, an api, and a validation method. BMC Bioinformatics. 2011; 12(1): 167. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nLe Novère NL, Hucka M, Mi H, et al.: The systems biology graphical notation. Nat Biotechnol. 2009; 27(8): 735–741. PubMed Abstract \| Publisher Full Text\n\nKutmon M, Riutta A, Nunes N, et al.: WikiPathways: capturing the full diversity of pathway knowledge. Nucleic Acids Res. 2016; 44(D1): D488–D494. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCroft D, Mundo AF, Haw R, et al.: The reactome pathway knowledgebase. Nucleic Acids Res. 2014; 42(Database issue): D472–D477. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nVanhaesebroeck B, Stephens L, Hawkins P: PI3K signalling: the path to discovery and understanding. Nat Rev Mol Cell Biol. 2012; 13(3): 195–203. PubMed Abstract \| Publisher Full Text\n\nLevels M, Hanspers K, Kutmon M, et al.: Micrornas in cardiomyocyte hypertrophy (homo sapiens). 2017. Reference Source\n\nLi H, Zeng J, Shen K: PI3K/AKT/mTOR signaling pathway as a therapeutic target for ovarian cancer. Arch Gynecol Obstet. 2014; 290(6): 1067–1078. PubMed Abstract \| Publisher Full Text\n\nAzzaoui K, Jacoby E, Senger S, et al.: Scientific competency questions as the basis for semantically enriched open pharmacological space development. Drug Discov Today. 2013; 18(17–18): 843–852. PubMed Abstract \| Publisher Full Text\n\nfundatureanu-sever, Kerber R, Soiland-Reyes S, et al.: openphacts/OPS_LinkedDataApi: Open PHACTS Linked Data API 2.1.0 (Version 2.1.0). Zenodo. 2016. Data Source" }	[ { "id": "29933", "date": "24 Jan 2018", "name": "Augustin Luna", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe tool provides additional features to the Open PHACTS platform to simplify the access of interaction information from WikiPathways. Some comments:\n\nThe authors go into some detail about the MIM/SBGN representations used by WikiPathways. To my understanding, WikiPathways makes use of the format GPML as a data storage format. Some description of the relationship of GPML and the WPRDF format would be nice for readers to understand the data processing involved.\n\nLooking at the code repository, the API interface seems to be built using Swagger. Is there the plan to provide supported client packages to access the API from various programming languages? Are users encouraged to produce their own using the swagger-codegen (https://github.com/swagger-api/swagger-codegen) project? Have the authors had any experience with these auto-generated packages in the context of their API?\n\nThe sentence \"However, the new methods also support interactions with multiple sources and targets.\", are these the methods that are elsewhere described as not implemented currently?\n\nWhat entity IDs are acceptable? The examples show ENSEMBL is this the only acceptable ID source? Does the documentation provide any examples for ID conversion from much more common gene symbols, for example?\n\nI assume that when \"up\" or \"down\" interactions are not specified for return then both are returned is this correct?\n\nAre the example interactions (i.e. inhibition, stimulation, conversion) the only ones available in the API? If not, is there a listing of the interaction types with description of the interaction type.\n\nIt would be more clear to just state the call function name rather than \"the first and third calls\" and \"the api call\" in the example queries section.\n\nAdditional examples of the result output for all described function calls would be helpful for the described API calls.\n\nAdditional examples that highlight the integrative nature of Open PHACTS would be nice. For example, to show how results of the currently described API calls can paired the other data available, such as the CHEBI data from the platform.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Partly\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4059", "date": "12 Oct 2018", "name": "Ryan Miller", "role": "Author Response", "response": "WikiPathways is represented using GPML (Graphical Pathway Markup Language), which is a modified XML format. The WikiPathways RDF is divided into two divisions, the GPML and the WPRDF. The GPML side is the portion that is responsible for the drawing and graphical elements. The WPRDF side is there to describe the semantic elements of the RDF. We have updated the Methods section and clarified this in the first “Implementation” paragraph. The API is REST-like and dedicated clients are not required. However, a few client libraries are already available for various languages and platforms: JavaScript (ops.js), Java (ops4j), ropenphacts, and KNIME nodes. Personal experience with auto-generated packages has been negative, but based on code generation in general, not the Open PHACTS API. Supported libraries are referenced in the summary sections along with the workflow environments. We intended to clarify that this simple example referred to single-source, single-target interactions, but that the framework also supports interactions with more than one source and/or target. Such complex interactions occur frequently. The last sentence of the third paragraph under Methods: Implementation has been updated to make this more clear. The Open PHACTS API includes the Identifier Mapping Service component which allows use of many identifier schemes, as long as suitable link sets are available (at http://data.openphacts.org/1.5/ims/linksets/). These sets include many of the popular gene, protein, and metabolite identifier:for genes, proteins, and RNAs Ensembl, Entrez Gene, UniProt, and others; for metabolites the ID sources can include HMDB, ChEBI, PubChem, CAS registry numbers, and ChemSpider. The caption of Figure 1 has been extended and a paragraph added to Operation section to reflect this. Yes, you are correct: if the user does not specify the directions for the interactions, all the immediate interactions are retrieved regardless of their direction. This can be found in the last paragraph of the Methods section and updated here to clarify. Vocabularies.wikipathways.org also identifies catalysis and binding events as well as a more generic directedInteraction in the case where the type of the interaction is not identified. This can be found in the last paragraph of the Methods: Implementation. The text in the under Example Queries section has been updated to use the names of the API calls used. Additional figures that show example outputs for the remaining figures can be found in the supplemental info section. Workflow tools make it possible to take advantage of the integrative nature of the OPDP to make API calls in succession. With these tools, it is possible to perform a directional query of a target and identify alternative targets that can then be queried against the chemistry calls to identify active compounds for these alternative targets. An Example Workflows section has been added describing two simple workflows and the code for the workflows can be found in the supplementary information." } ] }, { "id": "32325", "date": "03 Apr 2018", "name": "Yi-An Chen", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe authors add new features to the Open PHACTS APIs which facilitate querying reactions in WikiPathways.\nSome comments:\nIn the first example, the query for AKT2 in the figure 1 uses http://identifiers.org/ensembl/ENSG00000105221 but the result in the figure 2 shows http://identifiers.org/ncbigene/208. I guess the API should be able to accept different commonly used identifiers, however, either the article or the web interface mention about the acceptable identifiers. In addition, It would be nice if the input of the query accepts simple identifiers, instead of constructing the full URI.\n\nThe result in figure 2 is difficult to interpret. It would be nice to expand the query to get the basic information for those interacting entities. For example, the gene symbol or the compound name.\n\nThe examples in figure 3 and figure 4 didn't show the query results.\n\nSimple queries in the examples may be difficult to show the usefulness of the new implemented APIs. A more sophisticated application which contains either bulk query or pipeline query should be helpful for the readers to understand the demand of using the APIs.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Partly\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [ { "c_id": "4058", "date": "12 Oct 2018", "name": "Ryan Miller", "role": "Author Response", "response": "1. The Open PHACTS API includes the Identifier Mapping Service component which allows use of many identifier schemes, as long as suitable link sets are available (at http://data.openphacts.org/1.5/ims/linksets/). These sets include many of the popular gene, protein, and metabolite identifier:for genes, proteins, and RNAs Ensembl, Entrez Gene, UniProt, and others; for metabolites the ID sources can include HMDB, ChEBI, PubChem, CAS registry numbers, and ChemSpider. 2. Retrieving specific gene symbol or compound names aside from the identifier URI, resource, and identifier can be part of a workflow in workflow tools like KNIME or Pipeline Pilot. We did create a simple workflow to address this. In this example, we use the /pathway/getInteractions API call to return the directed interactions for a pathway and the resulting IDs for the interaction and then returned the more human readable labels for the URIs. 3. Additional example output figures were added to the supplementary materials section to reflect the addition of example query results for the remaining calls. 4. We added an Example Workflows section describing two simple workflows. More complex applications and workflows have been recently published for KNIME and Pipeline Pilot, which we added as reference 19." } ] } ]	1	https://f1000research.com/articles/7-75
https://f1000research.com/articles/7-566/v1	10 May 18	{ "type": "Software Tool Article", "title": "epicontacts: Handling, visualisation and analysis of epidemiological contacts", "authors": [ "VP Nagraj", "Nistara Randhawa", "Finlay Campbell", "Thomas Crellen", "Bertrand Sudre", "Thibaut Jombart", "Nistara Randhawa", "Finlay Campbell", "Thomas Crellen", "Bertrand Sudre", "Thibaut Jombart" ], "abstract": "Epidemiological outbreak data is often captured in line list and contact format to facilitate contact tracing for outbreak control. epicontacts is an R package that provides a unique data structure for combining these data into a single object in order to facilitate more efficient visualisation and analysis. The package incorporates interactive visualisation functionality as well as network analysis techniques. Originally developed as part of the Hackout3 event, it is now developed, maintained and featured as part of the R Epidemics Consortium (RECON). The package is available for download from the Comprehensive R Archive Network (CRAN) and GitHub.", "keywords": [ "contact tracing", "outbreaks", "R" ], "content": "Introduction\n\nIn order to study, prepare for, and intervene against disease outbreaks, infectious disease modellers and public health professionals need an extensive data analysis toolbox. Disease outbreak analytics involve a wide range of tasks that need to be linked together, from data collection and curation to exploratory analyses, and more advanced modelling techniques used for incidence forecasting1,2 or to predict the impact of specific interventions3,4. Recent outbreak responses suggest that for such analyses to be as informative as possible, they need to rely on a wealth of available data, including timing of symptoms, characterisation of key delay distributions (e.g. incubation period, serial interval), and data on contacts between patients5–8.\n\nThe latter type of data is particularly important for outbreak analysis, not only because contacts between patients are useful for unravelling the drivers of an epidemic9,10 , but also because identifying new cases early can reduce ongoing transmission via contact tracing, i.e. follow-up of individuals who reported contacts with known cases11,12. However, curating contact data and linking them to existing line lists of cases is often challenging, and tools for storing, handling, and visualising contact data are often missing13,14.\n\nHere, we introduce epicontacts, an R15 package providing a suite of tools aimed at merging line lists and contact data, and providing basic functionality for handling, visualising and analysing epidemiological contact data. Maintained as part of the R Epidemics Consortium (RECON), the package is integrated into an ecosystem of tools for outbreak response using the R language.\n\n\nMethods\n\nepicontacts is released as an open-source R package. A stable release is available for Windows, Mac and Linux operating systems via the CRAN repository. The latest development version of the package is available through the RECON Github organization. At minimum users must have R installed. No other system dependencies are required.\n\n\n\n\n\nData handling. epicontacts includes a novel data structure to accommodate line list and contact list datasets in a single object. This object is constructed with the make_epiconctacts() function and includes attributes from the original datasets. Once combined, these are mapped internally in a graph paradigm as nodes and edges. The epicontacts data structure also includes a logical attribute for whether or not this resulting network is directed.\n\nThe package takes advantage of R’s generic functions, which call specific methods depending on the class of an object. This is implemented in several places, including the summary.epicontacts() and print.epicontacts() methods, both of which are respectively called when the summary() or print() functions are used on an epicontacts object. The package does not include built-in data, as exemplary contact and line list datasets are available in the outbreaks package16.\n\n\n\n\n\n\n\n\n\n\n\nData visualisation. epicontacts implements two interactive network visualisation packages: visNetwork and threejs17,18. These frameworks provide R interfaces to the vis.js and three.js JavaScript libraries respectively. Their functionality is incorporated in the generic plot() method (Figure 1) for an epicontacts object, which can be toggled between either with the “type” parameter. Alternatively, the visNetwork interactivity is accessible via vis_epicontacts() (Figure 2), and threejs through graph3D() (Figure 3). Each function has a series of arguments that can also be passed through plot(). Both share a color palette, and users can specify node, edge and background colors. However, vis_epicontacts() includes a specification for “node_shape” by a line list attribute as well as a customization of that shape with an icon from the Font Awesome icon library. The principal distinction between the two is that graph3D() is a three-dimensional visualisation, allowing users to rotate clusters of nodes to better inspect their relationships.\n\n\n\n\n\n\n\nData analysis. Subsetting is a typical preliminary step in data analysis. epicontacts leverages a customized subset method to filter line list or contacts based on values of particular attributes from nodes, edges or both. If users are interested in returning only contacts that appear in the line list (or vice versa), the thin() function implements such logic.\n\n\n\nFor analysis of pairwise contact between individuals, the get_pairwise() feature searches the line list based on the specified attribute. If the given column is a numeric or date object, the function will return a vector containing the difference of the values of the corresponding “from” and “to” contacts. This can be particularly useful, for example, if the line list includes the date of onset of each case. The subtracted value of the contacts would approximate the serial interval for the outbreak19. For factors, character vectors and other non-numeric attributes, the default behavior is to print the associated line list attribute for each pair of contacts. The function includes a further parameter to pass an arbitrary function to process the specified attributes. In the case of a character vector, this can be helpful for tabulating information about different contact pairings with table().\n\n\n\n\nUse cases\n\nThose interested in using epicontacts should have a line list of cases as well as a record of contacts between individuals. Both datasets must be enumerated in tabular format with rows and columns. At minimum the line list requires one column with a unique identifier for every case. The contact list needs two columns for the source and destination of each pair of contacts. The datasets can include arbitrary features of case or contact beyond these columns. Once loaded into R and stored as data.frame objects, these datasets can be passed to the make_epicontacts() function (see ‘Methods’ section for more detail). For an example of data prepared in this format, users can refer to the outbreaks R package.\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\nDiscussion\n\nWhile there are software packages available for epidemiological contact visualisation and analysis, none aim to accommodate line list and contact data as purposively as epicontacts20–22. Furthermore, this package strives to solve a problem of plotting dense graphs by implementing interactive network visualisation tools. A static plot of a network with many nodes and edges may be difficult to interpret. However, by rotating or hovering over an epicontacts visualisation, a user may better understand the data.\n\nThe maintainers of epicontacts anticipate new features and functionality. Future development could involve performance optimization for visualising large networks, as generating these interactive plots is resource intensive. Additionally, attention may be directed towards inclusion of alternative visualisation methods.\n\n\nConclusions\n\nepicontacts provides a unified interface for processing, visualising and analyzing disease outbreak data in the R language. The package and its source are freely available on CRAN and GitHub. By developing functionality with line list and contact list data in mind, the authors aim to enable more efficient epidemiological outbreak analyses.\n\n\nSoftware availability\n\nSoftware available from: https://CRAN.R-project.org/package=epicontacts\n\nSource code available from: https://github.com/reconhub/epicontacts\n\nArchived source code as at time of publication: https://zenodo.org/record/121099323\n\nSoftware license: GPL 2", "appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nThe authors would like to thank all of the organizers and participants of the Hackout3 event held in Berkeley, California June 20–24, 2016. In particular, the authors acknowledge the support of the following organizations: MRC Centre for Outbreak Analysis, and Modelling at Imperial College London, the NIHR’s Modelling Methodology Health Protection Research Unit at Imperial College London, and the Berkeley Institute for Data Science.\n\n\nReferences\n\nFunk S, Camacho A, Kucharski AJ, et al.: Real-time forecasting of infectious disease dynamics with a stochastic semi-mechanistic model. Epidemics. 2018; 22: 56–61. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNouvellet P, Cori A, Garske T, et al.: A simple approach to measure transmissibility and forecast incidence. Epidemics. 2018; 22: 29–35. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nNouvellet P, Garske T, Mills HL, et al.: The role of rapid diagnostics in managing Ebola epidemics. Nature. 2015; 528(7580): S109–116. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nParker EP, Molodecky NA, Pons-Salort M, et al.: Impact of inactivated poliovirus vaccine on mucosal immunity: implications for the polio eradication endgame. Expert Rev Vaccines. 2015; 14(8): 1113–1123. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCauchemez S, Fraser C, Van Kerkhove MD, et al.: Middle East respiratory syndrome coronavirus: quantification of the extent of the epidemic, surveillance biases, and transmissibility. Lancet Infect Dis. 2014; 14(1): 50–56. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWHO Ebola Response Team, Aylward B, Barboza P, et al.: Ebola virus disease in West Africa--the first 9 months of the epidemic and forward projections. N Engl J Med. 2014; 371(16): 1481–1495. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWHO Ebola Response Team, Agua-Agum J, Ariyarajah A, et al.: West African Ebola epidemic after one year--slowing but not yet under control. N Engl J Med. 2015; 372(6): 584–587. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCori A, Donnelly CA, Dorigatti I, et al.: Key data for outbreak evaluation: building on the Ebola experience. Philos Trans R Soc Lond B Biol Sci. 2017; 372(1721). PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nInternational Ebola Response Team, Agua-Agum J, Ariyarajah A, et al.: Exposure Patterns Driving Ebola Transmission in West Africa: A Retrospective Observational Study. PLoS Med. 2016; 13(11): e1002170. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCauchemez S, Nouvellet P, Cori A, et al.: Unraveling the drivers of MERS-CoV transmission. Proc Natl Acad Sci U S A. 2016; 113(32): 9081–9086. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSenga M, Koi A, Moses L, et al.: Contact tracing performance during the Ebola virus disease outbreak in Kenema district, Sierra Leone. Philos Trans R Soc Lond B Biol Sci. 2017; 372(1721). PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nSaurabh S, Prateek S: Role of contact tracing in containing the 2014 Ebola outbreak: a review. Afr Health Sci. 2017; 17(1): 225–236. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nWorld Health Organization: Response to Measles Outbreaks in Measles Mortality Reduction Settings: Immunization, Vaccines and Biologicals. 2009. PubMed Abstract\n\nRakesh P, Sherin D, Sankar H, et al.: Investigating a community-wide outbreak of hepatitis a in India. J Glob Infect Dis. 2014; 6(2): 59–64. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nR Core Team: R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. 2017. Reference Source\n\nJombart T, Frost S, Nouvellet P, et al.: outbreaks: A Collection of Disease Outbreak Data. R package version 1.3.0. 2017. Reference Source\n\nAlmende BV, Thieurmel B, Robert T: visNetwork: Network Visualization using ‘vis.js’ Library. R package version 2.0.3. 2018. Reference Source\n\nLewis BW: threejs: Interactive 3D Scatter Plots, Networks and Globes. R package version 0.3.1. 2017. Reference Source\n\nFine PE: The interval between successive cases of an infectious disease. Am J Epidemiol. 2003; 158(11): 1039–1047. PubMed Abstract \| Publisher Full Text\n\nNöremark M, Widgren S: EpiContactTrace: an R-package for contact tracing during livestock disease outbreaks and for risk-based surveillance. BMC Vet Res. 2014; 10: 71. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nCarroll LN, Au AP, Detwiler LT, et al.: Visualization and analytics tools for infectious disease epidemiology: a systematic review. J Biomed Inform. 2014; 51: 287–298. PubMed Abstract \| Publisher Full Text \| Free Full Text\n\nGuthrie JL, Alexander DC, Marchand-Austin A, et al.: Technology and tuberculosis control: the OUT-TB Web experience. J Am Med Inform Assoc. 2017; 24(e1): e136–e142. PubMed Abstract \| Publisher Full Text\n\nNagraj VP, Jombart T, Randhawa N, et al.: epicontacts (Version 1.1.1). Zenodo. 2018. Data Source" }	[ { "id": "34084", "date": "31 May 2018", "name": "Melissa A. Rolfes", "expertise": [], "suggestion": "Approved With Reservations", "report": "Approved With Reservations\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThe article describes an R-based software tool aimed to facilitate analysis of data from outbreaks that include line lists of cases and case-contact data. The R package, epicontacts, is part of a larger suite of tools housed at the R Epidemics Consortium (RECON). The epicontacts package has the ability to merge data about cases in a line list with case-contact details, which then allows the user to describe and visualize contact networks, incubation periods, and serial intervals within an outbreak.\nThe codes and methods for analysis are partly described in the article, and the authors should provide a link to the packages documentation, either at CRAN or RECON webpages, where readers could learn more about the package and its options.\nThe output of the package provided in the article was interesting and intriguing. I felt that it was only partly explained and the article could benefit from the authors annotating the output and its interpretation a bit further. I have explored the RECON website and found the RECON Learn modules to be quite helpful in providing annotation of the epicontacts output and some guidance on interpretation. I would recommend that the authors consider either expanding the annotation of the output in this article or explicitly direct readers to the RECON Learn website for further instruction.\n\nAdditional suggestions:\nConsider moving the section of the article called \"Use cases\" to before the \"Data handling\" subsection of the \"Implementation\" section. I felt that the description of the input datasets under \"Use cases\" was very informative and would have been organizational more helpful had it been placed earlier in the article.\n\nConsider describing the sample outbreak data in a bit further detail. It appears to be data describing the MERS outbreak that occurred in South Korea in 2015. I think the description should include whether the data are simulated or from a real outbreak (if from a real outbreak, then a reference to the outbreak description should be included), the scenario of the outbreak, how many cases, how many contacts, place of the outbreak, duration of the outbreak, and a brief description of the demographic details included in the dataset. This amount of detail would allow the reader to translate the details of the outbreak from your text to the output provided by epicontacts.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Partly\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [] }, { "id": "36044", "date": "02 Aug 2018", "name": "Peter Adebayo Adewuyi", "expertise": [], "suggestion": "Approved", "report": "Approved\n\ninfo_outline\nAlongside their report, reviewers assign a status to the article:\n\nApproved The paper is scientifically sound in its current form and only minor, if any, improvements are suggested\n\nApproved with reservations\nA number of small changes, sometimes more significant revisions are required to address specific details and improve the papers academic merit.\n\nNot approved Fundamental flaws in the paper seriously undermine the findings and conclusions\n\nThis is a good software developed which could help in continuous visualization of contacts and their progression in disease tracking.\n\nIt is user friendly for those who are not computer specialist and still want to visualize data.\nData visualization is pertinent to disease monitoring and what the authors have done will aid in helping epidemiologist and public health specialist involved in outbreak response to quickly visualize progression and spread of disease from primary to secondary contacts and how the disease is evolving among contacts.\nThe software will actually achieve its purpose as stated in the conclusion of the write-up. Good work done.\n\nIs the rationale for developing the new software tool clearly explained? Yes\n\nIs the description of the software tool technically sound? Yes\n\nAre sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others? Yes\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes", "responses": [] } ]	1	https://f1000research.com/articles/7-566

Subsets and Splits

No community queries yet

The top public SQL queries from the community will appear here once available.