Datasets:

IbrahimAlAzhar
/

limitation-generation-dataset-bagels

+{
+    "v1_text": "how long is a piece of loop? : Loops are irregular structures which connect two secondary structure elements in proteins. They often play important roles in function, including enzyme reactions and ligand binding. Despite their importance, their structure remains difficult to predict. Most protein loop structure prediction methods sample local loop segments and score them. In particular protein loop classifications and database search methods depend heavily on local properties of loops. Here we examine the distance between a loop's end points (span). We find that the distribution of loop span appears to be independent of the number of residues in the loop, in other words the separation between the anchors of a loop does not increase with an increase in the number of loop residues. Loop span is also unaffected by the secondary structures at the end points, unless the two anchors are part of an anti-parallel beta sheet. As loop span appears to be independent of global properties of the protein we suggest that its distribution can be described by a random fluctuation model based on the Maxwell-Boltzmann distribution. It is believed that the primary difficulty in protein loop structure prediction comes from the number of residues in the loop. Following the idea that loop span is an independent local property, we investigate its effect on protein loop structure prediction and show how normalised span (loop stretch) is related to the structural complexity of loops. Highly contracted loops are more difficult to predict than stretched loops. Pre Prin ts Pre Prin ts Introduction1 Protein loops are patternless regions which connect two regular secondary2 structures. They are generally located on the protein\u2019s surface in solvent3 exposed areas and often play important roles, such as interacting with4 other biological objects.5 Despite the lack of patterns, loops are not completely random struc-6 tures. Early studies of short turns and hairpins showed that these peptide7 fragments could be clustered into structural classes (Richardson 1981;8 Sibanda & Thorton 1985). Such classifications have also been made9 across all loops (Burke, Deane & Blundell 2000; Chothia & Lesk 1987;10 Donate et al. 1996; Espadaler et al. 2004; Oliva et al. 1997; Vanhee11 et al. 2011) or within specific protein families such as antibody comple-12 mentarity determining regions (CDRs) (Al-Lazikani, Lesk & Chothia 1997;13 Chothia & Lesk 1987; Chothia et al. 1989). Loop classifications are gener-14 ally based on local properties such as sequence, the secondary structures15 from which the loop starts and finishes (anchor region), the distance be-16 tween the anchors, and the geometrical shape along the loop structure17 (Kwasigroch, Chomilier & Mornon 1996; Leszczynski & Rose 1986; Ring18 et al. 1992; Wojcik, Mornon & Chomilier 1999).19 Loops can also be classified in terms of function. There is some ev-20 idence that a loop can have local functionality. Experiments have been21 carried out which support the idea that swapping a local loop sequence for22 1 Pre Prin ts Pre Prin ts a different functional loop sequence enables the new function to be taken23 on (Pardon et al. 1995; Toma et al. 1991; Wolfson et al. 1991). One24 important example of functional loop exchange is in the development of25 humanised antibodies (Queen et al. 1989; Riechmann et al. 1988).26 Accurate protein loop structure prediction remains an open question.27 Protein loop predictors have dealt with the problem as a case of local pro-28 tein structure prediction. Protein structures are hypothesised to be in ther-29 modynamic equilibrium with their environment (Anfinsen 1973). Thus the30 primary determinant of a protein structure is considered to be its atomic31 interactions, i.e. its amino acid sequence. An analogous conjecture has32 arisen at the local scale where environment other than loop structure is33 fixed. Thus the modelling of protein loops is often considered a mini pro-34 tein folding problem (Fiser, Do & Sali 2000; Nagi & Regan 1997). Although35 most loop structure prediction methods are based on this conjecture, ap-36 parently loop sequence alone is not the complete determinant of the loop37 structure as even identical loop sequences can take multiple structural38 conformations depending on external environmental factors such as sol-39 vent and ligand binding (Fernandez-Fuentes & Fiser 2006). Quintessen-40 tial examples of such multiple loop structure conformations can be found41 in antibody CDR loops upon antigen binding (Choi & Deane 2011).42 Database search methods have been successful in the realm of loop43 structure prediction (Verschueren et al. 2011). They depend upon the44 assumption that similarity between local properties may suggest similar45 2 Pre Prin ts Pre Prin ts local structures. All database search methods work in an analogous fash-46 ion using either a complete set or a classified set of loops and selecting47 predictions using local features including sequence similarity and anchor48 geometry (Choi & Deane 2010; Fernandez-Fuentes, Oliva & Fiser 2006;49 Hildebrand et al. 2009; Peng & Yang 2007; Wojcik, Mornon & Chomi-50 lier 1999). Ab initio loop modelling methods aim to predict peptide frag-51 ments that do not exist in homology modelling templates without structure52 databases. Generally, ab initio methods generate large local structure con-53 formation sets and select predictions (de Bakker et al. 2003; Fiser, Do &54 Sali 2000; Jacobson et al. 2004; Mandell, Coutsias & Kortemme 2009;55 Soto et al. 2008). The generated loop candidates are optimised against56 scoring functions. In all loop modelling procedures anchor regions are57 often problematic and the accuracy of loop modelling depends upon the58 distance between the anchors (Xiang, 2006).59 Here, we focus on a specific local property of protein loop structure: the60 distance between the two terminal C\u03b1 atoms of the loop, which we refer to61 as its span. The nature of the span distribution is broadly similar across dif-62 ferent protein classes or anchor types, except for loops linking anti-parallel63 strands (anti-parallel \u03b2 loops). In particular, the most highly frequent span64 appears to stay the same irrespective of the number of residues. This sug-65 gests that the span is distributed independently of other local properties66 and global structures. We demonstrate that the observed span distribution67 can largely be explained by a simple model of random fluctuations with a68 3 Pre Prin ts Pre Prin ts given length scale, based on the Maxwell-Boltzmann distribution.69 It is widely believed that the accuracy of loop structure prediction de-70 pends on the number of residues, i.e. the larger the number of residues,71 the more difficult a loop is to predict (Choi & Deane 2010; Karen et al.72 2007). We introduce the normalised span which indicates how stretched73 a loop is (loop stretch \u03bb). Fully stretched loops (\u03bb ' 1) are almost always74 predicted accurately, whereas contracted loops (\u03bb 1) are harder to pre-75 dict. In fact, shorter loops tend to be more stretched whereas longer loops76 are likely to be highly contracted. We suggest that loop stretch should be77 addressed in practical modelling situations and loop structure prediction78 should be concerned with predicting highly contracted loops.79 Materials and Methods80 Loop Definition81 In each of the sets of protein structures loops, were identified using the fol-82 lowing protocol. Secondary structures were annotated using JOY (Mizuguchi83 et al. 1998). A loop structure was defined as any region between two84 regular secondary structures that was at least three residues in length85 (Donate et al. 1996). Short (less than 4 residues in length) loops were86 discarded. Redundancy was removed using sequence identity. If a pair87 of loops shares over 40% sequence identity (Fernandez-Fuentes & Fiser88 4 Pre Prin ts Pre Prin ts 2006), the loop which has a higher average B-factor was discarded.89 Membrane Protein Structures90 Membrane proteins (3, 789 chains) were extracted from PDBTM (Tusnady,91 Dosztanyi & Simon 2004). The membrane layer was defined as being92 from \u221220 to +20A\u030a (Scott et al. 2008) from the centre of the protein and93 loops whose two end C\u03b1 atom coordinates were outside the layer were94 discarded. A total of 1, 027 non-redundant membrane loops were defined.95 Soluble Protein Structures96 All protein chains determined by X-ray crystallography which share less97 than 99% sequence identity (< 3.0A\u030a resolution and < 0.3 R-factor) were98 collected using PISCES (Wang & Dunbrack Jr. 2005) and all of our 3, 78999 membrane chains were removed. In order to get rid of any potential mem-100 brane chains in the list, PSI-BLAST (Altschul et al. 1997) was then used to101 compare the 3, 789 membrane chains against the soluble set. Any chains102 found during 5 iterations with an E-value cut-off of 0.001 were removed from103 the list of soluble protein chains. A total of 25, 191 non-redundant soluble104 loops were identified from 27, 717 soluble protein chains.105 5 Pre Prin ts Pre Prin ts Loop Span and Loop Stretch106 The loop span (l) is the distance between the two terminal C\u03b1 atoms of a107 loop (Figure 1).108 The maximum span lmax is a function of the number of residues n and109 calculated as follows.110 lmax(n) = \u03b3 \u00b7 (n/2\u2212 1) + \u03b4 if n is even\u03b3 \u00b7 (n\u2212 1) /2 if n is odd where \u03b3 = 6.046A\u030a and \u03b4 = 3.46A\u030a (Flory 1998; Tastan, Klein-Seetharaman111 & Meirovitch 2009). If the distance between two terminal C\u03b1 atoms in the112 loop (i.e. the span) is l, the loop stretch (\u03bb) of the loop is defined as a113 normalised span.114 \u03bb \u2261 l lmax (1) Note that the values of \u03b3 and \u03b4 are theoretical approximations so the115 \u03bb of some loops may occasionally be larger than 1. Similar notations are116 found in (Ring et al. 1992) and (Tastan, Klein-Seetharaman & Meirovitch117 2009).118 6 Pre Prin ts Pre Prin ts Protein Structure Prediction and Loop Stretch119 Loop Modelling Test Sets120 There are two modelling test sets. The first set includes loops of 8 residues.121 The loops were binned every 0.1 loop stretch. In each bin, 40 test loops122 were randomly selected. A total of 320 test loops from 0.2 to 1 in loop123 stretch were used (A full list is given in Table S1).124 The second set consists of loops of between 6 and 10 residues in125 length. Two classes of loops were collected at each length: contracted126 loops (\u03bb < 0.4) and stretched loops (\u03bb > 0.95); an identical number of127 loops was kept in each of these classes at each length. A total of 346128 test loops were identified (58, 72, 110, 58 and 48 loops respectively, See Ta-129 ble S2 and S3). For example, there are 55 contracted test loops and 55130 stretched loops for loops of 8 residues.131 The measurement of accuracy is loop RMSD of all backbone atoms (N,132 C\u03b1, C and O) after superimposing anchor structures.133 MODELLER Setting134 The default loop refinement script was used. One hundred loop models135 were sampled under the molecular dynamics level of slow. The DOPE po-136 tential energy (Shen & Sali 2006) was used for model quality assessment.137 7 Pre Prin ts Pre Prin ts FREAD Setting138 A database was constructed using the 27, 717 soluble protein chains de-139 fined above. All the parameters were set as default (the environment sub-140 stitution score cut-off value \u2265 25). Any results from self-prediction were141 eliminated.142 Results143 Nomenclature144 In this paper, proteins are divided into two main classes: membrane and145 soluble proteins. Loops from membrane protein structures are called \u201cmem-146 brane loops\u201d and those from soluble protein structures are referred to as147 \u201csoluble loops\u201d. Loops are also described by their secondary structure148 types: for example, loops connecting anti-parallel \u03b2 sheets are termed149 \u201canti-parallel \u03b2 loops\u201d. The physical spatial distance between the two end150 C\u03b1 atoms of a loop is referred to as \u201cspan\u201d (l). Maximum loop span (lmax)151 is the furthest that a set of residues can spread. \u201cLoop stretch\u201d (\u03bb) is the152 normalised loop span: the observed span between two C\u03b1 atoms at each153 end of a loop in a protein structure over the loops maximum span (Figure154 1).155 8 Pre Prin ts Pre Prin ts Loop Span Distribution156 The number of residues in a loop is distributed in a similar fashion regard-157 less of anchor types except for the loops linking anti-parallel \u03b2 sheets due158 to the constraint of hydrogen bonds between adjacent \u03b2 strands (Figure159 2A). Figure 2B displays how loop spans are distributed for different anchor160 types. Again, apart from anti-parallel \u03b2 loops, the loop span distributions161 do not change with anchor structures.162 The loop span distribution also does not alter when considering differ-163 ent protein classes. Figures 2C\u20132G show how the loop spans of mem-164 brane loops and soluble loops are distributed in a similar manner.165 Essentially a loop span value reflects how distant the end tips of the166 two secondary structures that the loop connects are. These observations167 suggest that the loop span may be distributed independently of local an-168 chor structures and protein types, i.e. anchor distances do not depend on169 local secondary structure elements or global protein structures.170 The modes of loop span distributions are roughly constant (Figure 2B),171 even if we split the loops in terms of the number of residues (Figure 3A).172 We fit our data using the Gaussian kernel density estimation. The esti-173 mated distributions show a nearly constant mode (' 13A\u030a on average, Fig-174 ure 3B). This constant span value may be due to protein packing. Folded175 proteins tend to be tightly packed and thus secondary structures are placed176 close to one another while avoiding side chain steric clashes. This packing177 9 Pre Prin ts Pre Prin ts effect may mean that the end points of two secondary structures (i.e. span)178 will lie within a constant span value regardless of the number of residues179 in a loop.180 Maxwell-Boltzmann Distribution for Loop Span181 From the above observations, it appears that loop span is distributed in-182 dependently of local anchor structures or global protein classes. Here we183 assume that a protein loop is an independent unit of the protein structure184 and the span is determined regardless of any other effects including se-185 quence or the rest of the structure.186 Here a model for the loop span distribution is established under the187 hypothesis that the two end points of a loop fluctuate in three dimensional188 space, following the Maxwell-Boltzmann distribution. Two constraints are189 imposed in this model: the minimum span lmin and the maximum span190 as a function of the number of residues lmax(n). Within these constraints,191 the span oscillates according to a normal distribution N (\u00b5, \u03c32) with a given192 length-scale lmode in three dimensional space.193 The underlying assumptions are that the end points cannot approach194 each other too closely, and that there is a maximum span achievable for195 a loop with a given number of residues (n). Within these constraints, the196 span is allowed to fluctuate around the given length-scale lmode in three197 dimensional space. Thus, in this model, the loop span l of n residues is198 10 Pre Prin ts Pre Prin ts distributed as199 l = \u221a l2x + l 2 y + l 2 z lx, ly, lz \u223c N ( 0, l2mode 2 ) (2) subject to the constraints that l \u2265 lmin and l \u2264 lmax(n), as stated above.200 The variance of l2mode/2 corresponds to a modal span of lmode. Thus there201 are two parameters to be determined in our model: lmin and lmode. We set202 lmin to 3.8A\u030a, which is the typical distance between two neighbouring C\u03b1203 atoms in a protein chain. lmode is set to an estimate of the empirical mode204 using the Gaussian kernel density estimation (12.7A\u030a).205 As there are not many longer loops in the data set, loops longer than206 20 residues were discarded. In addition, all anti-parallel \u03b2 loops were elim-207 inated due to their physical constraints. These eliminations left 21, 597208 soluble loops (The frequency distribution for each number of residues is in209 Figure S2). Having set the two parameters lmin and lmode, loop spans were210 generated 10 times per model in accordance with the Maxwell-Boltzmann211 distribution, preserving the observed distribution of the number of residues212 (i.e. 10 simulated loop spans were generated for each real loop in the data213 set). The simulation outcome is depicted in Figure 4A. The two distri-214 butions show the same shape and the quantile comparison in Figure 4B215 indicates that they are statistically similar except for the tail region.216 There are apparent anomalies between the simulated and real span217 distributions towards the extremes. The model seems to predict more218 11 Pre Prin ts Pre Prin ts short-span loops than observed. Our model imposes a sharp lower thresh-219 old at lmin = 3.8A\u030a, whereas in reality we expect a smoother transition. In220 other words, we expect our assumption of free fluctuation to break down221 when the span gets close to the lower bound and the physical constraints222 begin to become relevant. On the other side of the distribution, we see a223 substantially higher number of long-span loops (> 20A\u030a) than predicted by224 the model. The mismatches in the long-span region tend to become more225 prominent as the number of residues is increased. When we examined226 which loops tend to have exceptionally long spans, we found that some of227 these \u201cloops\u201d are domain linkers between independent folding units and228 therefore likely to be under different constraints. Others appear to have229 been misclassified, as the loop definition used here is based only on the230 anchors containing at least three consecutive residues of secondary struc-231 tures and the loop containing none. This allows segments such as termini232 structures to be included if there happen to be very short helical segments233 at a protein structure\u2019s terminus (Figure S1).234 Protein Structure Prediction and Loop Stretch235 The number of residues in loops is known to be related to the protein236 stability (Nagi & Regan 1997) and the accuracy of most loop modelling237 techniques. Based on our observation that the loop span is independent238 of other properties, we examine its effects on protein loop structure pre-239 12 Pre Prin ts Pre Prin ts diction. Here we introduce loop stretch, the normalised loop span (Eq.240 1). Loop stretch values take on a range of 0 to 1, which indicates how241 stretched a loop is (1: fully stretched).242 Figure 5 displays how loop stretch frequencies are distributed for dif-243 ferent numbers of residues, demonstrating that the number of residues is244 negatively correlated with loop stretch, i.e. the longer a loop is, the more245 likely it is to be contracted. This may suggest that, instead of the stan-246 dard belief that loop modelling performs worse as the number of residues247 in the loop increases, it may be that the real problem is better described248 by considering how stretched the loop to be predicted is. For example, if249 a loop contains many residues but is highly stretched, it will be predicted250 relatively accurately, as it can take on only a small number of different251 conformations.252 In order to check the relationship between accuracy and loop stretch253 we used a test set containing only 8 residue loops with 40 non-redundant254 loops in every 0.1 loop stretch bin. Two loop modelling methods, which255 use two different sampling methods, were tested. MODELLER (Fiser, Do256 & Sali 2000) is a popular protein structure prediction programme which has257 a built-in ab initio loop modelling module. FREAD (Choi & Deane 2010)258 is a database search method which samples candidate loops depending259 on local properties and ranks predictions based on local loop sequence260 similarity and anchor geometry matches.261 The average accuracy of MODELLER shows a negative linear corre-262 13 Pre Prin ts Pre Prin ts lation against loop stretch for the first test set (Figure 6A). In the case of263 fully stretched loops (\u03bb > 0.95), MODELLER can produce consistently ac-264 curate predictions, but its predictions worsen as the target loops are less265 stretched. FREAD produces more accurate predictions than MODELLER266 in general. However its predictions also begin to disperse as the loops267 become more contracted (Figure 6B). FREAD generates candidate loops268 based on anchor matches and sequence similarity for a given loop target.269 This may imply that contracted loops tend to have multiple structural con-270 formations or stringent sequence identity is required to predict such highly271 contracted loops. It should be noted that FREAD is not able to predict272 all the target loops due to the incompleteness of the structure database it273 uses (Figure 6C).274 In order to further assess the effect of loop stretch in loop structure275 prediction, MODELLER was re-examined on a second set. The second276 test set consists of loops from 6 to 10 residues in length. In this set, for277 each number of residues, the same numbers of loops (See Materials and278 Methods) were selected for both contracted (\u03bb < 0.4) and fully stretched279 loops (\u03bb > 0.95). MODELLER produces consistently accurate results for280 fully stretched loops regardless of the number of residues, but fails to ac-281 curately predict contracted loops (Figure 6D).282 We calculated the partial correlations (Spearman\u2019s rank correlation)283 between accuracy, and the number of residues and loop stretch on the284 second test set. so as to investigate what affects the prediction accuracy285 14 Pre Prin ts Pre Prin ts more (the number of residues or loop stretch). The partial correlation be-286 tween loop stretch and RMSD is larger than that between the number of287 residues and RMSD (\u22120.465 and 0.367 respectively). Loop stretch, just like288 the number of residues is something that can be calculated without knowl-289 edge of loop conformation and therefore can be used in the design of loop290 structure prediction software.291 Discussion292 In this paper, we focus on a specific local property (span) and demonstrate293 that the modes of loop span distribution appear to be independent of the294 number of residues. Loop span shows a distinct frequency distribution295 which does not depend on anchor types or protein classes. From these296 observations, we hypothesised that loop span is independent of the other297 effects and showed how the loop span distribution appears to correspond298 to a truncated Maxwell-Boltzmann distribution.299 The reason behind the independence of loop span from the number300 of loop residues or secondary structure type is not known. The fact that301 the loop span distribution can be captured by a simple Maxwell-Boltzmann302 model allows one to speculate that protein loop structure prediction is in-303 deed a local mini protein folding problem.304 15 Pre Prin ts Pre Prin ts Acknowledgments305 Yoonjoo Choi was funded by the Department of Statistics, St. Cross Col-306 lege and the University of Oxford. Sumeet Agarwal was funded by a Uni-307 versity of Oxford Clarendon Fund Scholarship.308 References309 Al-Lazikani B, Lesk AM, Chothia C. 1998. Standard conformations for the310 canonical structures of immunoglobulins. J Mol Biol, 273: 927-948.311 Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, Lip-312 man DJ. 1997. Gapped BLAST and PSI-BLAST: a new generation of pro-313 tein database search programs. Nucleic Acids Res, 25: 3389-3402.314 Anfinsen CB. 1973. Principles that govern the folding of protein chains.315 Science, 181: 223-230.316 Burke DF, Deane CM, Blundell TL. 2000. Browsing the SLoop database317 of structurally classified loops connecting elements of protein secondary318 structure. Bioinformatics, 16: 513-519.319 Choi Y, Deane CM. 2010. FREAD revisited: Accurate loop structure320 prediction using a database search algorithm. Proteins, 78: 1431-1440.321 Choi Y, Deane CM. 2011. Predicting antibody complementarity deter-322 mining region structures without classification. Mol Biosyst, 7: 3327-3334.323 Chothia C, Lesk AM. 1987. Canonical Structures for the Hypervariable324 Regions of Immunoglobulins. J Mol Biol, 196: 901-917.325 Chothia C, Lesk AM, Tramontano A, Levitt M, Smith-Gill SJ, Air G, Sher-326 iff S, Padlan EA, Davies D, Tulip WR, Colman PM, Spinelli S, Alzari PM,327 Poljak RJ. 1989. Conformations of immunoglobulin hypervariable regions.328 Nature, 342: 877-883.329 de Bakker PI, DePristo MA, Burke DF, Blundell TL. 2003. Ab initio330 construction of polypeptide fragments: Accuracy of loop decoy discrimina-331 tion by an all-atom statistical potential and the AMBER force field with the332 Generalized Born solvation model. Proteins, 51: 21-40.333 16 Pre Prin ts Pre Prin ts Donate LE, Rufino SD, Canard LH, Blundell TL. 1996. Conformational334 analysis and clustering of short and medium size loops connecting regular335 secondary structures: a database for modeling and prediction. Protein336 Sci, 5: 2600-2616.337 Espadaler J, Fernandez-Fuentes N, Hermoso A, Querol E, Aviles FX,338 Sternberg MJE, Oliva B. 2004. ArchDB: automated protein loop classifica-339 tion as a tool for structural genomics. Nucleic Acids Res, 32: D185-D188.340 Fernandez-Fuentes N, Fiser A. 2006. Saturating representation of341 loop conformational fragments in structure databanks. BMC Struc Biol,342 6: doi:10.1186/1472-6807-1186-1115.343 Fernandez-Fuentes N, Oliva B, Fiser A. 2006. A supersecondary struc-344 ture library and search algorithm for modeling loops in protein structures.345 Nucleic Acids Res, 34: 2085-2097.346 Fiser A, Do RK, Sali A. 2000. Modeling of loops in protein structures.347 Protein Sci, 9: 1753-1773.348 Flory P. 1998. Statistical Mechanics of Chain Molecules: Hanser.349 Hildebrand PW, Goede A, Bauer RA, Gruening B, Ismer J, Michalsky E,350 Preissner R. 2009. SuperLooper - a prediction server for the modeling of351 loops in globular and membrane proteins. Nucleic Acids Res, 37: W571-352 W574.353 Jacobson MP, Pincus DL, Rapp CS, Day TJ, Honig B, Shaw DE, Fries-354 ner RA. 2004. A hierarchical approach to all-atom protein loop prediction.355 Proteins, 55: 351-367.356 Karen AR, Weigelt CA, Nayeem A, Krystek Jr SR. 2007. Loopholes357 and missing links in protein modeling. Protein Sci, 16: 1-14.358 Kwasigroch KM, Chomilier J, Mornon JP. 1996. A global taxonomy of359 loops in globular proteins. J Mol Biol, 259: 855-872.360 Leszczynski JF, Rose GD. 1986. Loops in globular proteins: a novel361 category of secondary structure. Science, 234: 849-855.362 Mandell DJ, Coutsias EA, Kortemme T. 2009. Sub-angstrom accuracy363 in protein loop reconstruction by robotics-inspired conformational sam-364 pling. Nat Methods, 6: 551-552.365 Mizuguchi K, Deane CM, Blundell TL, Johnson MS, Overington JP.366 1998. JOY: protein sequence-structure representation and analysis. Bioin-367 formatics, 14: 617-623.368 Nagi AD, Regan L. 1997. An inverse correlation between loop length369 and stability in a four-helix-bundle protein. Fold Des, 2: 67-75.370 17 Pre Prin ts Pre Prin ts Oliva B, Bates PA, Querol E, Aviles FX, Sternberg MJE. 1997. An371 Automated Classfication of the Structure of Protein Loops. J Mol Biol,372 266: 814-830.373 Pardon E, Haezebrouck P, De Baetselier A, Hooke SD, Fancourt KT,374 Dobson JDCM, Dael HV, Joniau M. 1995. A Ca(2+)-binding chimera of375 human lysozyme and bovine alpha-lactalbumin that can form a molten376 globule. J Biol Chem, 270: 10514-10524.377 Peng H, Yang A. 2007. Modling protein loos with knoledge-based pre-378 diction of sequence-structure alignment. Bioinformatics, 23: 2836-2842.379 Queen C, Schneider WP, Selick HE, Payne PW, Landolfi NF, Duncan380 JF, Avdalovic NM, Levitt M, Junghans RP, Waldmann TA. 1989. A human-381 ized antibody that binds to the interleukin 2 receptor. PNAS, 86: 10029-382 10033.383 Richardson JS. 1981. The anatomy and taxonomy of protein structure.384 Adv Protein Chem, 34: 167-339.385 Riechmann L, Clark M, Waldmann H, Winter G. 1988. Reshaping hu-386 man antibodies for therapy. Nature, 332: 323-327.387 Ring CS, Kneller DG, Langridge R, Cohen FE. 1992. Taxonomy and388 conformational analysis of loops in proteins. J Mol Biol, 224: 685-699.389 Scott KA, Bond PJ, Ivetac A, Chetwynd AP, Khalid S, Sansom MSP.390 2008. Coarse-Grained MD simulations of membrane protein-bilayer self-391 assembly. Structure, 16: 621-630.392 Shen MY, Sali A. 2006. Statistical potential for assessment and predic-393 tion of protein structures. Protein Sci, 15: 2507-2524.394 Sibanda BL, Thorton JM. 1985. Beta-hairpin families in globular pro-395 teins. Nature, 316: 170-174.396 Soto CS, Fasnacht M, Zhu J, Forrest L, Honig B. 2008. Loop modeling:397 Sampling, filtering, and scoring. Proteins, 70: 834-843.398 Tastan O, Klein-Seetharaman J, Meirovitch H. 2009. The Effect of399 Loops on the Structural Organization of -Helical Membrane Proteins. Bio-400 phys J, 96: 2299-2312.401 Toma S, Campagnoli S, Margarit I, Gianna R, Grandi G, Bolognesi M.402 De Filippis V, Fontana A. 1991. Grafting of a calcium-binding loop of ther-403 molysin to Bacillus subtilis neutral protease. Biochemistry, 30: 97-106.404 Tusnady GE, Dosztanyi ZD, Simon I. 2004. Transmembrane proteins405 in the Protein Data Bank: identification and classification. Bioinformatics,406 20: 2964-2972.407 18 Pre Prin ts Pre Prin ts Vanhee P, Verschueren E, Baeten L, Stricher F, Serrano L, Rousseau408 F, Schymkowitz J. 2011. BriX: a database of protein building blocks for409 structural analysis, modeling and design. Nucleic Acids Res, 39: D435-410 D442.411 Verschueren E, Vanhee P, van der Sloot AM, Serrano L, Rousseau F,412 Schymkowitz J. 2011. Protein design with fragment databases. Curr Opin413 Struct Biol, 21: 452-459.414 Wang G, Dunbrack Jr. RL. 2005. PISCES: recent improvements to a415 PDB sequence culling server. Nucleic Acids Res, 33: W94-W98.416 Wojcik J, Mornon JP, Chomilier J. 1999. New efficient statistical sequence-417 dependent structure prediction of short to medium-sized protein loops based418 on an exhaustive loop classification. J Mol Biol, 289: 1469-1490.419 Wolfson AJ, Kanaoka M, Lau FT, Ringe D. 1991. Insertion of an elastase-420 binding loop into interleukin-1 beta. Protein Eng, 4: 313-317.421 Xiang Z. 2006. Advances in Homology Protein Structure Modeling.422 Curr Protein Pept Sci, 7: 217-227.423 19 Pre Prin ts Pre Prin ts Figure 1 The definition of loop span and loop stretch Loop span is the separation of the two C\u03b1s at either end of the loop. In this example, 2J9O Chain A (198-205) has a span of 13.7\u00c5 and contains 8 residues. Maximum span can be calculated from the number of residues in the loop to be 21.6\u00c5. Loop stretch is the normalised span (13.7/21.6 0.63).\u2243 Pre Prin ts Pre Prin ts Figure 2 statistics of protein loops : (A) The frequency distribution of loops containing different numbers of residues. Anti-parallel \u03b2 loops tend to have fewer residues. (B) The loop span distribution in terms of the anchor secondary structure do not show differences except for anti-parallel \u03b2 loops. The upper part of the anti-parallel \u03b2 loop span distribution is omitted in the figure. (C) The distributions of soluble loop span and membrane loop span appear to be similar. (D)-(G) Q\u2013Q plots showing that the membrane and soluble loop span distributions are from the same probability distribution. Pre Prin ts Pre Prin ts Figure 3 The span distributions for loops containing different numbers of residues (A) These appear to show a constant mode. Data here is soluble loops excluding anti-parallel beta loops. (B) The modes for the span distributions for loops containing different numbers of residues compared to the maximum span for that length. The span modes were estimated using the Gaussian kernel density estimation. Note that the estimated mode of loops of 4 residues is close to its maximum span. Pre Prin ts Pre Prin ts Pre Prin ts Pre Prin ts Figure 4 Maxwell-Boltzmann distribution and loop span distribution (A) The loop span distribution (black) from soluble loops and that of the Maxwell-Boltzmann distribution (red). (B) The Q-Q plot suggesting that they follow the same distribution. Pre Prin ts Pre Prin ts Pre Prin ts Pre Prin ts Figure 5 Loop stretch of long and short loops Loop stretch distributions for loops containing different numbers of residues Shorter loops tend to be more stretched whereas longer loops are likely to be more contracted. Only soluble loops excluding anti-parallel \u03b2 loops are plotted. Pre Prin ts Pre Prin ts Figure 6 Protein loop structure prediction and loop stretch Accuracy of protein loop structure prediction methods do not only depend on the number of residues, but also on loop stretch. MODELLER (A) and FREAD (B) both show accurate results when the target loop is stretched on the first set (including loops of 8 residues in length only). MODELLER shows worse prediction as loop stretch decreases whereas FREAD gives consistent accuracy on loop stretch. However both fail to predict very contracted loops (\u03bb &lt; 0.4) (C) The coverage of FREAD predictions in terms of loop stretch. (D) The second test set (contracted (\u03bb &lt; 0.4) and stretched (\u03bb &gt; 0.95) loops). The test loops are also split by the number of residues. For fully stretched loops (\u03bb &gt; 0.95), regardless of the number of residues, MODELLER predicts accurately.Pre Prin ts Pre Prin ts",
+    "v2_text": "how long is a piece of loop? : Loops are irregular structures which connect two secondary structure elements in proteins. They often play important roles in function, including enzyme reactions and ligand binding. Despite their importance, their structure remains difficult to predict. Most protein loop structure prediction methods sample local loop segments and score them. In particular protein loop classifications and database search methods depend heavily on local properties of loops. Here we examine the distance between a loop's end points (span). We find that the distribution of loop span appears to be independent of the number of residues in the loop, in other words the separation between the anchors of a loop does not increase with an increase in the number of loop residues. Loop span is also unaffected by the secondary structures at the end points, unless the two anchors are part of an anti-parallel beta sheet. As loop span appears to be independent of global properties of the protein we suggest that its distribution can be described by a random fluctuation model based on the Maxwell-Boltzmann distribution. It is believed that the primary difficulty in protein loop structure prediction comes from the number of residues in the loop. Following the idea that loop span is an independent local property, we investigate its effect on protein loop structure prediction and show how normalised span (loop stretch) is related to the structural complexity of loops. Highly contracted loops are more difficult to predict than stretched loops. Pre Prin ts Pre Prin ts Introduction1 Protein loops are patternless regions which connect two regular secondary2 structures. They are generally located on the protein\u2019s surface in solvent3 exposed areas and often play important roles, such as interacting with4 other biological objects.5 Despite the lack of patterns, loops are not completely random struc-6 tures. Early studies of short turns and hairpins showed that these peptide7 fragments could be clustered into structural classes (Richardson 1981;8 Sibanda & Thorton 1985). Such classifications have also been made9 across all loops (Burke, Deane & Blundell 2000; Chothia & Lesk 1987;10 Donate et al. 1996; Espadaler et al. 2004; Oliva et al. 1997; Vanhee et al.11 2011) or within specific protein families such as antibody complementarity12 determining regions (Al-Lazikani, Lesk & Chothia 1997; Chothia & Lesk13 1987; Chothia et al. 1989). Loop classifications are generally based on14 local properties such as sequence, the secondary structures from which15 the loop starts and finishes (anchor region), the distance between the an-16 chors, and the geometrical shape along the loop structure (Kwasigroch,17 Chomilier & Mornon 1996; Leszczynski & Rose 1986; Ring et al. 1992;18 Wojcik, Mornon & Chomilier 1999).19 Loops can also be classified in terms of function. There is some ev-20 idence that a loop can have local functionality. Experiments have been21 carried out which support the idea that swapping a local loop sequence for22 1 Pre Prin ts Pre Prin ts a different functional loop sequence enables the new function to be taken23 on (Pardon et al. 1995; Toma et al. 1991; Wolfson et al. 1991). One24 important example of functional loop exchange is in the development of25 humanised antibodies (Queen et al. 1989; Riechmann et al. 1988).26 Accurate protein loop structure prediction remains an open question.27 Protein loop predictors have dealt with the problem as a case of local pro-28 tein structure prediction. Protein structures are hypothesised to be in ther-29 modynamic equilibrium with their environment (Anfinsen 1973). Thus the30 primary determinant of a protein structure is considered to be its atomic31 interactions, i.e. its amino acid sequence. An analogous conjecture has32 arisen at the local scale. The modelling of protein loops is often consid-33 ered a mini protein folding problem (Fiser, Do & Sali 2000; Nagi & Regan34 1997). In fact, most loop structure prediction methods are based on this35 conjecture.36 Database search methods have been successful in the realm of loop37 structure prediction (Verschueren et al. 2011). They depend upon the38 assumption that similarity between local properties may suggest similar39 local structures. All database search methods work in an analogous fash-40 ion using either a complete set or a classified set of loops and selecting41 predictions using local features including sequence similarity and anchor42 geometry (Choi & Deane 2010; Fernandez-Fuentes, Oliva & Fiser 2006;43 Hildebrand et al. 2009; Peng & Yang 2007; Wojcik, Mornon & Chomi-44 lier 1999). Ab initio loop modelling methods aim to predict peptide frag-45 2 Pre Prin ts Pre Prin ts ments that do not exist in homology modelling templates without structure46 databases. Generally, ab initio methods generate large local structure con-47 formation sets and select predictions (de Bakker et al. 2003; Fiser, Do &48 Sali 2000; Jacobson et al. 2004; Mandell, Coutsias & Kortemme 2009;49 Soto et al. 2008). The generated loop candidates are optimised against50 scoring functions. In all loop modelling procedures anchor regions are51 often problematic and the accuracy of loop modelling depends upon the52 distance between the anchors (Xiang, 2006).53 Here, we focus on a specific local property of protein loop structure: the54 distance between the two terminal C\u03b1 atoms of the loop, which we refer to55 as its span. The nature of the span distribution is broadly similar across dif-56 ferent protein classes or anchor types, except for loops linking anti-parallel57 strands (anti-parallel \u03b2 loops). In particular, the most highly frequent span58 appears to stay the same irrespective of the number of residues. This sug-59 gests that the span is distributed independently of other local properties60 and global structures. We demonstrate that the observed span distribution61 can largely be explained by a simple model of random fluctuations with a62 given length scale, based on the Maxwell-Boltzmann distribution.63 It is widely believed that the accuracy of loop structure prediction de-64 pends on the number of residues, i.e. the larger the number of residues,65 the more difficult a loop is to predict (Choi & Deane 2010; Karen et al.66 2007). We introduce the normalised span which indicates how stretched67 a loop is (loop stretch \u03bb). Fully stretched loops (\u03bb \u2243 1) are almost always68 3 Pre Prin ts Pre Prin ts predicted accurately, whereas contracted loops (\u03bb \u226a 1) are harder to pre-69 dict. In fact, shorter loops tend to be more stretched whereas longer loops70 are likely to be highly contracted. We suggest that loop stretch should be71 addressed in practical modelling situations and loop structure prediction72 should be concerned with predicting highly contracted loops.73 Materials and Methods74 Loop Definition75 In each of the sets of protein structures loops, were identified using the fol-76 lowing protocol. Secondary structures were annotated using JOY (Mizuguchi77 et al. 1998). A loop structure was defined as any region between two78 regular secondary structures that was at least three residues in length79 (Donate et al. 1996). Short (less than 4 residues in length) loops were80 discarded. Redundancy was removed using sequence identity. If a pair81 of loops shares over 40% sequence identity (Fernandez-Fuentes & Fiser82 2006), the loop which has a higher average B-factor was discarded. Mem-83 brane Protein Structures84 Membrane proteins (3, 789 chains) were extracted from PDBTM (Tus-85 nady, Dosztanyi & Simon 2004). The membrane layer was defined as86 being from \u221220 to +20A\u030a (Scott et al. 2008) from the centre of the protein87 and loops whose two end C atom coordinates were outside the layer were88 4 Pre Prin ts Pre Prin ts discarded. A total of 1, 027 non-redundant membrane loops were defined.89 Soluble Protein Structures90 All protein chains determined by X-ray crystallography which share less91 than 99% sequence identity (< 3.0A\u030a resolution and < 0.3 R-factor) were92 collected using PISCES (Wang & Dunbrack Jr. 2005) and all of our 3, 78993 membrane chains were removed. In order to get rid of any potential mem-94 brane chains in the list, PSI-BLAST (Altschul et al. 1997) was then used to95 compare the 3, 789 membrane chains against the soluble set. Any chains96 found during 5 iterations with an E-value cut-off of 0.001 were removed97 from the soluble chains list. A total of 25, 191 non-redundant soluble loops98 were identified from 27, 717 soluble protein chains.99 Loop Span and Loop Stretch100 The loop span (l) is the distance between the two terminal C\u03b1 atoms of a101 loop (Figure 1).102 The maximum span lmax is a function of the number of residues n and103 calculated as follows.104 lmax(n) = \u03b3 \u00b7 (n/2\u2212 1) + \u03b4 if n is even\u03b3 \u00b7 (n\u2212 1) /2 if n is odd where \u03b3 = 6.046A\u030a and \u03b4 = 3.46A\u030a (Flory 1998; Tastan, Klein-Seetharaman105 5 Pre Prin ts Pre Prin ts & Meirovitch 2009). If the distance between two terminal C\u03b1 atoms in the106 loop (i.e. the span) is l, the loop stretch (\u03bb) of the loop is defined as a107 normalised span.108 \u03bb \u2261 l lmax (1) Note that the values of \u03b3 and \u03b4 are theoretical approximations so the109 \u03bb of some loops may occasionally be larger than 1. Similar notations are110 found in (Ring et al. 1992) and (Tastan, Klein-Seetharaman & Meirovitch111 2009).112 Loop Modelling Test Sets113 There are two modelling test sets. The first set includes loops of 8 residues.114 The loops were binned every 0.1 loop stretch. In each bin, 40 test loops115 were randomly selected. A total of 320 test loops from 0.2 to 1 in loop116 stretch were used (A full list is given in Table S1).117 The second set consists of loops of between 6 and 10 residues in118 length. Two classes of loops were collected at each length: contracted119 loops (\u03bb < 0.4) and stretched loops (\u03bb > 0.95); an identical number of120 loops was kept in each of these classes at each length. A total of 346121 test loops were identified (58, 72, 110, 58 and 48 loops respectively, See Ta-122 ble S2 and S3). For example, there are 55 contracted test loops and 55123 stretched loops for loops of 8 residues.124 6 Pre Prin ts Pre Prin ts The measurement of accuracy is loop RMSD of all backbone atoms (N,125 C\u03b1, C and O) after superimposing anchor structures.126 MODELLER Setting127 The default loop refinement script was used. One hundred loop models128 were sampled under the molecular dynamics level of slow. The DOPE po-129 tential energy (Shen & Sali 2006) was used for model quality assessment.130 Results131 Nomenclature132 In this paper, proteins are divided into two main classes: membrane and133 soluble proteins. Loops from membrane protein structures are called \u201cmem-134 brane loops\u201d and those from soluble protein structures are referred to as135 \u201csoluble loops\u201d. Loops are also described by their secondary structure136 types: for example, loops connecting anti-parallel \u03b2 sheets are termed137 \u201canti-parallel \u03b2 loops\u201d. The physical spatial distance between the two end138 C atoms of a loop is referred to as \u201cspan\u201d (l). Maximum loop span (lmax)139 is the furthest that a set of residues can spread. \u201cLoop stretch\u201d (\u03bb) is the140 normalised loop span: the observed span between two C\u03b1 atoms at each141 end of a loop in a protein structure over the loops maximum span (Figure142 1).143 7 Pre Prin ts Pre Prin ts Loop Span Distribution144 The number of residues in a loop is distributed in a similar fashion regard-145 less of anchor types except for the loops linking anti-parallel \u03b2 sheets due146 to the constraint of hydrogen bonds between adjacent \u03b2 strands (Figure147 2A). Figure 2B displays how loop spans are distributed for different anchor148 types. Again, apart from anti-parallel \u03b2 loops, the loop span distributions149 do not change with anchor structures.150 The loop span distribution also does not alter when considering dif-151 ferent protein classes. Figures 2C and D show how the loop spans of152 membrane loops and soluble loops are distributed in a similar manner.153 Essentially a loop span value reflects how distant the end tips of the154 two secondary structures that the loop connects are. These observations155 suggest that the loop span may be distributed independently of local an-156 chor structures and protein types, i.e. anchor distances do not depend on157 local secondary structure elements or global protein structures.158 The modes of loop span distributions are roughly constant (Figure 2B),159 even if we split the loops in terms of the number of residues (Figure 3A).160 We fit our data using the Gaussian kernel density estimation. The es-161 timated distributions show a nearly constant mode (\u2243 13A\u030a on average,162 Figure 3B). On the face of it the fact that the mode is a constant inde-163 pendent of the number of residues in the loop is surprising. However it164 might be due to protein structural features. Apart from long loops linking165 8 Pre Prin ts Pre Prin ts two remote anchors (e.g. Figure S1), the secondary structures tend to be166 packed against one another. Due to the sizes of side chains the anchors167 are not able to approach too closely, but it may be that they pack against168 one another potentially leading to a constant value.169 Maxwell-Boltzmann Distribution for Loop Span170 From the above observations, it appears that loop span is distributed in-171 dependently of local anchor structures or global protein classes. Here we172 assume that a protein loop is an independent unit of the protein structure173 and the span is determined regardless of any other effects including se-174 quence or the rest of the structure.175 Here a model for the loop span distribution is established under the176 hypothesis that the two end points of a loop fluctuate in three dimensional177 space, following the Maxwell-Boltzmann distribution. Two constraints are178 imposed in this model: the minimum span lmin and the maximum span179 as a function of the number of residues lmax(n). Within these constraints,180 the span oscillates according to a normal distribution N (\u00b5, \u03c32) with a given181 length-scale lmode in three dimensional space.182 The underlying assumptions are that the end points cannot approach183 each other too closely, and that there is a maximum span achievable for184 a loop with a given number of residues (n). Within these constraints, the185 span is allowed to fluctuate around the given length-scale lmode in three186 9 Pre Prin ts Pre Prin ts dimensional space. Thus, in this model, the loop span l of n residues is187 distributed as188 l = \u221a l2x + l 2 y + l 2 z lx, ly, lz \u223c N ( 0, l2mode 2 ) (2) subject to the constraints that l \u2265 lmin and l \u2264 lmax(n), as stated above.189 The variance of l2mode/2 corresponds to a modal span of lmode. Thus there190 are two parameters to be determined in our model: lmin and lmode. We set191 lmin to 3.8A\u030a, which is the typical distance between two neighbouring C\u03b1192 atoms in a protein chain. lmode is set to an estimate of the empirical mode193 using the Gaussian kernel density estimation (12.7A\u030a).194 As there are not many longer loops in the data set, loops longer than195 20 residues were discarded. In addition, all anti-parallel \u03b2 loops were elim-196 inated due to their physical constraints. These eliminations left 21, 597197 soluble loops (The frequency distribution for each number of residues is in198 Figure S2). Having set the two parameters lmin and lmode, loop spans were199 generated 10 times per model in accordance with the Maxwell-Boltzmann200 distribution, preserving the observed distribution of the number of residues201 (i.e. 10 simulated loop spans were generated for each real loop in the data202 set). The simulation outcome is depicted in Figure 4A. The two distri-203 butions show the same shape and the quantile comparison in Figure 4B204 indicates that they are statistically similar except for the tail region.205 There are apparent anomalies between the simulated and real span206 10 Pre Prin ts Pre Prin ts distributions towards the extremes. The model seems to predict more207 short-span loops than observed. Our model imposes a sharp lower thresh-208 old at lmin = 3.8A\u030a, whereas in reality we expect a smoother transition. In209 other words, we expect our assumption of free fluctuation to break down210 when the span gets close to the lower bound and the physical constraints211 begin to become relevant. On the other side of the distribution, we see a212 substantially higher number of long-span loops (> 20A\u030a) than predicted by213 the model. The mismatches in the long-span region tend to become more214 prominent as the number of residues is increased. When we examined215 which loops tend to have exceptionally long spans, we found that some of216 these \u201cloops\u201d are domain linkers between independent folding units and217 therefore likely to be under different constraints. Others appear to have218 been misclassified, as the loop definition used here is based only on the219 anchors containing at least three consecutive residues of secondary struc-220 tures and the loop containing none. This allows segments such as termini221 structures to be included if there happen to be very short helical segments222 at a protein structure\u2019s terminus (Figure S1).223 Protein Structure Prediction and Loop Stretch224 The number of residues in loops is known to be related to the protein225 stability (Nagi & Regan 1997) and the accuracy of most loop modelling226 techniques. Based on our observation that the loop span is independent227 11 Pre Prin ts Pre Prin ts of other properties, we examine its effects on protein loop structure pre-228 diction. Here we introduce loop stretch, the normalised loop span (Eq.229 1). Loop stretch values take on a range of 0 to 1, which indicates how230 stretched a loop is (1: fully stretched).231 Figure 5 displays how loop stretch frequencies are distributed for dif-232 ferent numbers of residues, demonstrating that the number of residues is233 negatively correlated with loop stretch, i.e. the longer a loop is, the more234 likely it is to be contracted. This may suggest that, instead of the stan-235 dard belief that loop modelling performs worse as the number of residues236 in the loop increases, it may be that the real problem is better described237 by considering how stretched the loop to be predicted is. For example, if238 a loop contains many residues but is highly stretched, it will be predicted239 relatively accurately, as it can take on only a small number of different240 conformations.241 In order to check the relationship between accuracy and loop stretch,242 we use the ab initio loop modelling programme, MODELLER (Fiser, Do243 & Sali 2000). MODELLER is a popular protein structure prediction pro-244 gramme which has a built-in ab initio loop modelling module. Two test sets245 were prepared. The first test set contains loops of only 8 residues in length246 and 40 non-redundant loops in every 0.1 loop stretch bin. The second test247 set consists of loops from 6 to 10 residues in length. In this set, for each248 number of residues, the same numbers of loops (See Materials and Meth-249 ods) were selected for both contracted (\u03bb < 0.4) and fully stretched loops250 12 Pre Prin ts Pre Prin ts (\u03bb > 0.95).251 The average accuracy of MODELLER shows a negative linear corre-252 lation against loop stretch for the first test set (Figure 6A). In the case of253 fully stretched loops (\u03bb > 0.95), MODELLER can produce consistently ac-254 curate predictions, but its predictions worsen as the target loops are less255 stretched. Fully stretched loops are predicted accurately regardless of the256 number of residues (Figure 6B).257 However MODELLER failed to accurately predict contracted loops (Fig-258 ure 6A). In order to investigate what affects the prediction accuracy more259 (the number of residues or loop stretch), we calculated the partial cor-260 relations (Spearman\u2019s rank correlation) between accuracy, and the num-261 ber of residues and loop stretch. The partial correlation between loop262 stretch and RMSD is larger than that between the number of residues and263 RMSD (\u22120.465 and 0.367 respectively). Loop stretch, just like the number264 of residues is something that can be calculated without knowledge of loop265 conformation and therefore can be used in the design of loop structure266 prediction software.267 Discussion268 In this paper, we focus on a specific local property (span) and demonstrate269 that the modes of loop span distribution appear to be independent of the270 number of residues. Loop span shows a distinct frequency distribution271 13 Pre Prin ts Pre Prin ts which does not depend on anchor types or protein classes. From these272 observations, we hypothesised that loop span is independent of the other273 effects and showed how the loop span distribution appears to correspond274 to a truncated Maxwell-Boltzmann distribution.275 The reason behind the independence of loop span from the number276 of loop residues or secondary structure type is not known. The fact that277 the loop span distribution can be captured by a simple Maxwell-Boltzmann278 model allows one to speculate that protein loop structure prediction is in-279 deed a local mini protein folding problem.280 Acknowledgments281 Yoonjoo Choi was funded by the Department of Statistics, St. Cross Col-282 lege and Oxford University. Sumeet Agarwal was funded by a University283 of Oxford Clarendon Fund Scholarship.284 References285 Al-Lazikani B, Lesk AM, Chothia C. 1998. Standard conformations for the286 canonical structures of immunoglobulins. J Mol Biol, 273: 927-948.287 Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, Lip-288 man DJ. 1997. Gapped BLAST and PSI-BLAST: a new generation of pro-289 tein database search programs. Nucleic Acids Res, 25: 3389-3402.290 Anfinsen CB. 1973. Principles that govern the folding of protein chains.291 Science, 181: 223-230.292 14 Pre Prin ts Pre Prin ts Burke DF, Deane CM, Blundell TL. 2000. Browsing the SLoop database293 of structurally classified loops connecting elements of protein secondary294 structure. Bioinformatics, 16: 513-519.295 Choi Y, Deane CM. 2010. FREAD revisited: Accurate loop structure296 prediction using a database search algorithm. Proteins, 78: 1431-1440.297 Chothia C, Lesk AM. 1987. Canonical Structures for the Hypervariable298 Regions of Immunoglobulins. J Mol Biol, 196: 901-917.299 Chothia C, Lesk AM, Tramontano A, Levitt M, Smith-Gill SJ, Air G, Sher-300 iff S, Padlan EA, Davies D, Tulip WR, Colman PM, Spinelli S, Alzari PM,301 Poljak RJ. 1989. Conformations of immunoglobulin hypervariable regions.302 Nature, 342: 877-883.303 de Bakker PI, DePristo MA, Burke DF, Blundell TL. 2003. Ab initio304 construction of polypeptide fragments: Accuracy of loop decoy discrimina-305 tion by an all-atom statistical potential and the AMBER force field with the306 Generalized Born solvation model. Proteins, 51: 21-40.307 Donate LE, Rufino SD, Canard LH, Blundell TL. 1996. Conformational308 analysis and clustering of short and medium size loops connecting regular309 secondary structures: a database for modeling and prediction. Protein310 Sci, 5: 2600-2616.311 Espadaler J, Fernandez-Fuentes N, Hermoso A, Querol E, Aviles FX,312 Sternberg MJE, Oliva B. 2004. ArchDB: automated protein loop classifica-313 tion as a tool for structural genomics. Nucleic Acids Res, 32: D185-D188.314 Fernandez-Fuentes N, Fiser A. 2006. Saturating representation of315 loop conformational fragments in structure databanks. BMC Struc Biol,316 6: doi:10.1186/1472-6807-1186-1115.317 Fernandez-Fuentes N, Oliva B, Fiser A. 2006. A supersecondary struc-318 ture library and search algorithm for modeling loops in protein structures.319 Nucleic Acids Res, 34: 2085-2097.320 Fiser A, Do RK, Sali A. 2000. Modeling of loops in protein structures.321 Protein Sci, 9: 1753-1773.322 Flory P. 1998. Statistical Mechanics of Chain Molecules: Hanser.323 Hildebrand PW, Goede A, Bauer RA, Gruening B, Ismer J, Michalsky E,324 Preissner R. 2009. SuperLooper - a prediction server for the modeling of325 loops in globular and membrane proteins. Nucleic Acids Res, 37: W571-326 W574.327 Jacobson MP, Pincus DL, Rapp CS, Day TJ, Honig B, Shaw DE, Fries-328 ner RA. 2004. A hierarchical approach to all-atom protein loop prediction.329 Proteins, 55: 351-367.330 15 Pre Prin ts Pre Prin ts Karen AR, Weigelt CA, Nayeem A, Krystek Jr SR. 2007. Loopholes331 and missing links in protein modeling. Protein Sci, 16: 1-14.332 Kwasigroch KM, Chomilier J, Mornon JP. 1996. A global taxonomy of333 loops in globular proteins. J Mol Biol, 259: 855-872.334 Leszczynski JF, Rose GD. 1986. Loops in globular proteins: a novel335 category of secondary structure. Science, 234: 849-855.336 Mandell DJ, Coutsias EA, Kortemme T. 2009. Sub-angstrom accuracy337 in protein loop reconstruction by robotics-inspired conformational sam-338 pling. Nat Methods, 6: 551-552.339 Mizuguchi K, Deane CM, Blundell TL, Johnson MS, Overington JP.340 1998. JOY: protein sequence-structure representation and analysis. Bioin-341 formatics, 14: 617-623.342 Nagi AD, Regan L. 1997. An inverse correlation between loop length343 and stability in a four-helix-bundle protein. Fold Des, 2: 67-75.344 Oliva B, Bates PA, Querol E, Aviles FX, Sternberg MJE. 1997. An345 Automated Classfication of the Structure of Protein Loops. J Mol Biol,346 266: 814-830.347 Pardon E, Haezebrouck P, De Baetselier A, Hooke SD, Fancourt KT,348 Dobson JDCM, Dael HV, Joniau M. 1995. A Ca(2+)-binding chimera of349 human lysozyme and bovine alpha-lactalbumin that can form a molten350 globule. J Biol Chem, 270: 10514-10524.351 Peng H, Yang A. 2007. Modling protein loos with knoledge-based pre-352 diction of sequence-structure alignment. Bioinformatics, 23: 2836-2842.353 Queen C, Schneider WP, Selick HE, Payne PW, Landolfi NF, Duncan354 JF, Avdalovic NM, Levitt M, Junghans RP, Waldmann TA. 1989. A human-355 ized antibody that binds to the interleukin 2 receptor. PNAS, 86: 10029-356 10033.357 Richardson JS. 1981. The anatomy and taxonomy of protein structure.358 Adv Protein Chem, 34: 167-339.359 Riechmann L, Clark M, Waldmann H, Winter G. 1988. Reshaping hu-360 man antibodies for therapy. Nature, 332: 323-327.361 Ring CS, Kneller DG, Langridge R, Cohen FE. 1992. Taxonomy and362 conformational analysis of loops in proteins. J Mol Biol, 224: 685-699.363 Scott KA, Bond PJ, Ivetac A, Chetwynd AP, Khalid S, Sansom MSP.364 2008. Coarse-Grained MD simulations of membrane protein-bilayer self-365 assembly. Structure, 16: 621-630.366 Shen MY, Sali A. 2006. Statistical potential for assessment and predic-367 tion of protein structures. Protein Sci, 15: 2507-2524.368 16 Pre Prin ts Pre Prin ts Sibanda BL, Thorton JM. 1985. Beta-hairpin families in globular pro-369 teins. Nature, 316: 170-174.370 Soto CS, Fasnacht M, Zhu J, Forrest L, Honig B. 2008. Loop modeling:371 Sampling, filtering, and scoring. Proteins, 70: 834-843.372 Tastan O, Klein-Seetharaman J, Meirovitch H. 2009. The Effect of373 Loops on the Structural Organization of -Helical Membrane Proteins. Bio-374 phys J, 96: 2299-2312.375 Toma S, Campagnoli S, Margarit I, Gianna R, Grandi G, Bolognesi M.376 De Filippis V, Fontana A. 1991. Grafting of a calcium-binding loop of ther-377 molysin to Bacillus subtilis neutral protease. Biochemistry, 30: 97-106.378 Tusnady GE, Dosztanyi ZD, Simon I. 2004. Transmembrane proteins379 in the Protein Data Bank: identification and classification. Bioinformatics,380 20: 2964-2972.381 Vanhee P, Verschueren E, Baeten L, Stricher F, Serrano L, Rousseau382 F, Schymkowitz J. 2011. BriX: a database of protein building blocks for383 structural analysis, modeling and design. Nucleic Acids Res, 39: D435-384 D442.385 Verschueren E, Vanhee P, van der Sloot AM, Serrano L, Rousseau F,386 Schymkowitz J. 2011. Protein design with fragment databases. Curr Opin387 Struct Biol, 21: 452-459.388 Wang G, Dunbrack Jr. RL. 2005. PISCES: recent improvements to a389 PDB sequence culling server. Nucleic Acids Res, 33: W94-W98.390 Wojcik J, Mornon JP, Chomilier J. 1999. New efficient statistical sequence-391 dependent structure prediction of short to medium-sized protein loops based392 on an exhaustive loop classification. J Mol Biol, 289: 1469-1490.393 Wolfson AJ, Kanaoka M, Lau FT, Ringe D. 1991. Insertion of an elastase-394 binding loop into interleukin-1 beta. Protein Eng, 4: 313-317.395 Xiang Z. 2006. Advances in Homology Protein Structure Modeling.396 Curr Protein Pept Sci, 7: 217-227.397 17 Pre Prin ts Pre Prin ts Figure 1 The definition of loop span and loop stretch Loop span is the separation of the two C\u03b1s at either end of the loop. In this example, 2J9O Chain A (198-205) has a span of 13.7\u00c5 and contains 8 residues. Maximum span can be calculated from the number of residues in the loop to be 21.6\u00c5. Loop stretch is the normalised span (13.7/21.6 0.63).\u2243 Pre Prin ts Pre Prin ts Figure 2 statistics of protein loops : (A) The frequency distribution of loops containing different numbers of residues. Anti-parallel \u03b2 loops tend to have fewer residues. (B) The loop span distribution in terms of the anchor secondary structure do not show differences except for anti-parallel \u03b2 loops. The upper part of the anti-parallel \u03b2 loop span distribution is omitted in the figure. (C) The distributions of soluble loop span and membrane loop span appear to be similar. (D) A Q\u2013Q plot showing that the membrane and soluble loop span distributions are from the same probability distribution. Pre Prin ts Pre Prin ts Figure 3 The span distributions for loops containing different numbers of residues (A) These appear to show a constant mode. Data here is soluble loops excluding anti-parallel beta loops. (B) The modes for the span distributions for loops containing different numbers of residues compared to the maximum span for that length. The span modes were estimated using the Gaussian kernel density estimation. Note that the estimated mode of loops of 4 residues is close to its maximum span. Pre Prin ts Pre Prin ts Pre Prin ts Pre Prin ts Figure 4 Maxwell-Boltzmann distribution and loop span distribution (A) The loop span distribution (black) from soluble loops and that of the Maxwell-Boltzmann distribution (red). (B) The Q-Q plot suggesting that they follow the same distribution. Pre Prin ts Pre Prin ts Pre Prin ts Pre Prin ts Figure 5 Loop stretch of long and short loops Loop stretch distributions for loops containing different numbers of residues Shorter loops tend to be more stretched whereas longer loops are likely to be more contracted. Only soluble loops excluding anti-parallel \u03b2 loops are plotted. Pre Prin ts Pre Prin ts Figure 6 Protein loop structure prediction and loop stretch Accuracy of protein loop structure prediction methods do not only depend on the number of residues, but also on loop stretch. (A) The accuracy of loop prediction by MODELLER for loops which contain 8 residues with different values of loop stretch. there are 40 loops in each 0.1 split of loop stretch. a : moving average is shown. As loop stretch decreases prediction accuracy decreases. (B) Two sets of loops one contracted (\u03bb < 0.4) and one stretched (\u03bb > 0.95). Loops are also split by number of residues. For fully stretched loops (\u03bb > 0.95), regardless of the number of residues, MODELLER predicts accurately. Pre Prin ts Pre Prin ts Pre Prin ts Pre Prin ts",
+    "url": "https://peerj.com/articles/1/reviews/",
+    "review_1": "Walter de Azevedo Jr. \u00b7 Nov 21, 2012 \u00b7 Academic Editor\nACCEPT\nPaper accepted after modifications.",
+    "review_2": "Walter de Azevedo Jr. \u00b7 Nov 15, 2012 \u00b7 Academic Editor\nMINOR REVISIONS\nYou manuscript needs minor revision.",
+    "review_3": "Andras Fiser \u00b7 Nov 14, 2012\nBasic reporting\nNo comments\nExperimental design\nNo comments\nValidity of the findings\nNo comments\nAdditional comments\nThe manuscript of Choi et al. explores a specific property of protein loops: their span. It observes that loops with very different lengths (number of residues) can be fitted roughly to the same spans. They also describe a metric to asses the loop stretch, which in turn is informative to the quality of prediction: the more compact a loop, the more difficult to model it properly. As a consequence the authors argue that loop modeling methods should be tested and should be more focused on highly contracted loops, irrespectively of loop length, because these are much more challenging to model. Overall I found this work interesting and useful, the analysis is well done. I have a few minor comments:\n\n(1) The Introduction describing the loop modeling approaches is slightly misleading. It is important to stress that loop conformations -once the modeling is treated as a mini protein problem - do not depend on the loop sequence only. As in case of full length proteins a given sequence folds in its native fold under physiological conditions. In case of full length proteins the physiological environment is the solvent (and temperature etc), while in case of loop it is the solvent AND the rest of the protein spanning it. As a consequence, as it was described by us (BMC Struct Biol. 2006 Jul 4;6:15.Saturating representation of loop conformational fragments in structure databanks. Fernandez-Fuentes N, Fiser A.) and others, loops with identical sequences can have different conformations depending on the protein they show up.\n\n(2) In Fig 2C, I would guess that probably almost all membrane loops come from helix-helix connections, and this would require a comparison to helix-helix connected soluble loops and not to a mixed class, as it is now. In Fig2B helix-helix soluble loops spans are slightly shifted right for longer spans and on Fig2C membrane ones to the left, to shorter spans and therefore it is possible that once properly compared there will be a small but possibly significant difference.\n\n(3) There might be some trivial explanation for the roughly constant span of loops: tightly packed proteins cannot afford large cavities and therefore the typical anchor distances of ends of secondary structures will be under strong thermodynamic selection. All loops , irrespectively of length will have to be accommodated within that given span. There are some references to it in the text, but I think it might be a driving effect.\n\n(4) Fig. 6 illustrates one of the main points of the work that loop prediction accuracy is negatively correlated with loop span. While I do not doubt this effect, it seems that among compacted loops a few outliers contribute disproportionally to the worsening performance. I would suggest to use mean instead of average to account for the non-gaussian distribution more accurately.\nCite this review as\nFiser A (2013) Peer Review #1 of \"How long is a piece of loop? (v0.1)\". PeerJ https://doi.org/10.7287/peerj.1v0.1/reviews/1",
+    "review_4": "Joost Schymkowitz \u00b7 Nov 13, 2012\nBasic reporting\nThe paper meets all the criteria of PeerJ, it is sound, well written and it represents an important intelectual contribution to the field of loop structure prediction.\nExperimental design\nThe authors do two things: they analyse some test sets of loops in terms of the stretch and span parameters and then evaluate loop prediction accuracy of the ab initio method Modeller to try to get an idea of difficulty of the structure prediction problem (judged by the accuracy of modeller's performance).\n\nI like the concept a lot and it makes a lot of sense, complicated loop folds are essentially harder to predict - which still allows for longer loops to be harder than shorter ones, but length in not in itself the key determinant.\n\nAs an adept of database-based methods, I would have liked to see the same analysis using eg our freely available loopbrix method (http://brix.switchlab.org/), since this would tell if DB methods somehow have a different accuracy profile.\nValidity of the findings\nI think the study is well done and the results are clearly presented, the conclusions are supported by the data.\nAdditional comments\nCool paper.\nCite this review as\nSchymkowitz J (2013) Peer Review #2 of \"How long is a piece of loop? (v0.1)\". PeerJ https://doi.org/10.7287/peerj.1v0.1/reviews/2",
+    "pdf_1": "https://peerj.com/articles/1v0.2/submission",
+    "pdf_2": "https://peerj.com/articles/1v0.1/submission",
+    "all_reviews": "Review 1: Walter de Azevedo Jr. \u00b7 Nov 21, 2012 \u00b7 Academic Editor\nACCEPT\nPaper accepted after modifications.\nReview 2: Walter de Azevedo Jr. \u00b7 Nov 15, 2012 \u00b7 Academic Editor\nMINOR REVISIONS\nYou manuscript needs minor revision.\nReview 3: Andras Fiser \u00b7 Nov 14, 2012\nBasic reporting\nNo comments\nExperimental design\nNo comments\nValidity of the findings\nNo comments\nAdditional comments\nThe manuscript of Choi et al. explores a specific property of protein loops: their span. It observes that loops with very different lengths (number of residues) can be fitted roughly to the same spans. They also describe a metric to asses the loop stretch, which in turn is informative to the quality of prediction: the more compact a loop, the more difficult to model it properly. As a consequence the authors argue that loop modeling methods should be tested and should be more focused on highly contracted loops, irrespectively of loop length, because these are much more challenging to model. Overall I found this work interesting and useful, the analysis is well done. I have a few minor comments:\n\n(1) The Introduction describing the loop modeling approaches is slightly misleading. It is important to stress that loop conformations -once the modeling is treated as a mini protein problem - do not depend on the loop sequence only. As in case of full length proteins a given sequence folds in its native fold under physiological conditions. In case of full length proteins the physiological environment is the solvent (and temperature etc), while in case of loop it is the solvent AND the rest of the protein spanning it. As a consequence, as it was described by us (BMC Struct Biol. 2006 Jul 4;6:15.Saturating representation of loop conformational fragments in structure databanks. Fernandez-Fuentes N, Fiser A.) and others, loops with identical sequences can have different conformations depending on the protein they show up.\n\n(2) In Fig 2C, I would guess that probably almost all membrane loops come from helix-helix connections, and this would require a comparison to helix-helix connected soluble loops and not to a mixed class, as it is now. In Fig2B helix-helix soluble loops spans are slightly shifted right for longer spans and on Fig2C membrane ones to the left, to shorter spans and therefore it is possible that once properly compared there will be a small but possibly significant difference.\n\n(3) There might be some trivial explanation for the roughly constant span of loops: tightly packed proteins cannot afford large cavities and therefore the typical anchor distances of ends of secondary structures will be under strong thermodynamic selection. All loops , irrespectively of length will have to be accommodated within that given span. There are some references to it in the text, but I think it might be a driving effect.\n\n(4) Fig. 6 illustrates one of the main points of the work that loop prediction accuracy is negatively correlated with loop span. While I do not doubt this effect, it seems that among compacted loops a few outliers contribute disproportionally to the worsening performance. I would suggest to use mean instead of average to account for the non-gaussian distribution more accurately.\nCite this review as\nFiser A (2013) Peer Review #1 of \"How long is a piece of loop? (v0.1)\". PeerJ https://doi.org/10.7287/peerj.1v0.1/reviews/1\nReview 4: Joost Schymkowitz \u00b7 Nov 13, 2012\nBasic reporting\nThe paper meets all the criteria of PeerJ, it is sound, well written and it represents an important intelectual contribution to the field of loop structure prediction.\nExperimental design\nThe authors do two things: they analyse some test sets of loops in terms of the stretch and span parameters and then evaluate loop prediction accuracy of the ab initio method Modeller to try to get an idea of difficulty of the structure prediction problem (judged by the accuracy of modeller's performance).\n\nI like the concept a lot and it makes a lot of sense, complicated loop folds are essentially harder to predict - which still allows for longer loops to be harder than shorter ones, but length in not in itself the key determinant.\n\nAs an adept of database-based methods, I would have liked to see the same analysis using eg our freely available loopbrix method (http://brix.switchlab.org/), since this would tell if DB methods somehow have a different accuracy profile.\nValidity of the findings\nI think the study is well done and the results are clearly presented, the conclusions are supported by the data.\nAdditional comments\nCool paper.\nCite this review as\nSchymkowitz J (2013) Peer Review #2 of \"How long is a piece of loop? (v0.1)\". PeerJ https://doi.org/10.7287/peerj.1v0.1/reviews/2\nReview 5: nan\nReview 6: \n",
+    "LLM_extracted_review": "['Review 1: Paper accepted after modifications.\\n\\nReview 2: Your manuscript needs minor revision.\\n\\nReview 3: \\n(1) The Introduction describing the loop modeling approaches is slightly misleading. It is important to stress that loop conformations do not depend on the loop sequence only; they also depend on the rest of the protein spanning it.\\n(2) In Fig 2C, almost all membrane loops likely come from helix-helix connections, requiring a comparison to helix-helix connected soluble loops rather than a mixed class.\\n(3) There might be a trivial explanation for the roughly constant span of loops: tightly packed proteins cannot afford large cavities, leading to strong thermodynamic selection on anchor distances.\\n(4) Fig. 6 illustrates that loop prediction accuracy is negatively correlated with loop span, but a few outliers among compacted loops may disproportionately worsen performance. It is suggested to use mean instead of average for a more accurate representation.\\n\\nReview 4: \\nThe paper meets all the criteria of PeerJ, is sound, well written, and represents an important intellectual contribution to the field of loop structure prediction. \\nThe authors analyze test sets of loops in terms of stretch and span parameters and evaluate loop prediction accuracy of the ab initio method Modeller. The concept makes sense, as complicated loop folds are harder to predict. \\nIt would be beneficial to see the same analysis using the loopbrix method to determine if database methods have a different accuracy profile. \\nThe study is well done, results are clearly presented, and conclusions are supported by the data. \\nCool paper.']"
+}

peerj_json_files/PeerJ_Json_10.json ADDED Viewed

	@@ -0,0 +1,18 @@

+{
+    "v1_Abstract": "A prospective cross-sectional study of 1000 households was implemented in 2005 to evaluate characteristics of the owned and unowned population of dogs and cats in Santa Clara County, California. The same population was previously studied 12 years earlier. During this time period, the county instituted in 1994 and then subsequently disestablished a municipal spay/neuter voucher program for cats. Dog and cat intakes declined from 1992-2005, as they similarly did for an adjacent county (San Mateo). Time series analysis showed a greater than expected decline in the number of cats surrendered to shelters in Santa Clara County during the years the voucher program was in effect (1994-2005). The net savings to the county by reducing the number of cat shelter intakes was estimated at approximately $1.5 million.",
+    "v2_Abstract": "A prospective cross-sectional study of 1000 households was implemented in 2005 to evaluate characteristics of the owned and unowned population of dogs and cats in Santa Clara County, California. The same population was previously studied 12 years earlier. During this time period, the county instituted in 1994 and then subsequently disestablished a municipal spay/neuter voucher program for cats. Dog intakes declined from 1992-2005, as they similarly did for an adjacent county (San Mateo). However, cat intakes declined significantly more in Santa Clara County than San Mateo, with an average annual decline of approximately 700 cats for the 12 year period. Time series analysis showed a greater than expected decline in the number of cats surrendered to shelters in Santa Clara County during the years the voucher program was in effect (1994-2005). The net savings to the county by reducing the number of cat shelter intakes was estimated at approximately $1.5 million.",
+    "v1_text": "materials and methods : results : acknowledgements : We are grateful to Beth Ward and Chris Benninger at the Humane Society of Silicon Valley, Greg VanWassenhove at County of Santa Clara, and Jon Cicirelli at the City of San Jose for providing annual shelter statistics. 18 375 376 377 378 379 380 381 Pre Prin ts Pre Prin ts discussion : This study documents the positive impacts publically subsidized low-cost spay and neuter programs can have that often go unmet in communities: pet population control, leading to the prevention of the proliferation of feral dog and cat populations, slowing the flow of animals into shelters both voluntarily and through field services, and reduction in the incidence of humane destruction of animals. But they also extend to other issues of economic importance to communities; namely, reduction in capital and ongoing animal control expenditures that come under a municipality\u2019s jurisdiction. In contrast, the implication of cessation of such programs can be seen in Figures 2 and 3 when the decline in shelter admissions of dogs and cats became attenuated. The problem could be exacerbated over time as the human (and hence pet-owning) population increases. The finding that there were over 15,000 dogs (4.7% of the county\u2019s dog population) estimated to be transiently or permanently stray throughout the county is troubling from societal and public health standpoints. The absence of a domestic environment can lead stray dogs, which are by nature gregarious, to form packs that can become aggressive and endanger other animals and even humans. The origin of such a large number of dogs is worthy of further 14 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 Pre Prin ts Pre Prin ts research, as this study did not explore whether these were free-roaming dogs or those kept in temporary foster or rescue care. Specific breed information was not available for dogs in Santa Clara County animal shelters. Respondents claimed 51% of their dogs were registered and unregistered purebred dogs. This stands in contrast to a 1996 national survey that found 30% of dogs relinquished to shelters were purebred (Salman et al. 1998), and the Humane Society of the United States estimates that 25-30% of shelter dogs are purebred (The Humane Society of the United States 2011). Nationally, purebred dogs are substantially less likely to be relinquished to animal shelters than dogs of mixed breed (Salman et al. 1998). The dogs with greatest likelihood of successful adoption from county animal shelters are puppies (Lepper, Kass & Hart 2002). By the time dogs reach the age of one year, though, their risk of unsuccessful adoption following relinquishment rises considerably; again, particularly true in pit bull-like breeds (Lepper, Kass & Hart 2002). Aggressive dog behavior is a major reason dogs are euthanized at the county shelters (Kass et al. 2001). To reduce dog intakes, municipalities should consider how the establishment of free or low cost puppy training programs (potentially mandatory for shelter adoptions) might impact shelter populations. A collaborative effort among multiple community agencies, including animal control, non-profits, and local pet industry businesses should be explored. Another important finding is the enumeration of the substantial unowned cat population in Santa Clara County, two-thirds of which are feral. Also notable is that the majority of unowned cats entering the animal shelters in the study were arguably unsuitable for adoption, with over 50% being feral or unweaned kittens. Such cats are often quickly euthanized. Preventing such input defies simplistic solutions, because although 93% of cat owners were 15 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 Pre Prin ts Pre Prin ts willing to have their own pets surgically sterilized, it is unrealistic to expect the 7% of the population that feeds an average of 3 stray cats to assume the hundreds of dollars necessary to surgically alter these cats. Conversely, the cost of not altering the cats is to add 3.5 kittens per year for each stray female, which at the cost to a shelter of approximately $250 per cat would cost a shelter almost $900 in husbandry expenses for those 3.5 kittens; were they not sheltered, the kittens would be expected to have 75% mortality (Table 1). The underscores why low-cost spay and neuter programs directed to reducing the un-owned and feral cat populations continue to be integral to not only reducing cat mortality at the shelters, but also to managing the cost to the various municipalities to handle and house the stray cats. Santa Clara County\u2019s contribution of $45 to alter a stray cat under its separate feral spay/neuter program created an immediate savings of over $200 for just the first litter that permanently results in non-reproducing cats. The county program also subsidized shots, and for a time, FELV testing. The earlier such cats can be sterilized, the greater the potential savings to municipalities. The savings would be expected to grow over additional years. Moreover, under all plausible scenarios shown in Table 2 the voucher program would have resulted in a net savings in expenditure. If stray cat-feeding citizens can be convinced through public education to avail themselves of population control options by making them more affordable and they are provided with instructions and resources as to how to accomplish this activity, the savings in costs and lives will be substantial. This study shows approximately 93% of county residents did not make an effort to sterilize unowned cats. Only 5.5% of the unowned but fed cats were surgically sterilized. Efforts should be focused on removing barriers and finding ways to encourage those who feed free-roaming cats to take this important step. Because the study shows that 62% of 16 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 Pre Prin ts Pre Prin ts unowned but fed cats are fed in people\u2019s yards, efforts should be intensified to sterilize cats living in close proximity to homes, rather than less accessible colonies. This study\u2019s limitations include the assumption that the participating individuals are representative of the county\u2019s population. Interviews were conducted via telephone, with the non-telephone-owning segment of the county excluded by design, and to the extent that this subgroup differs in their pet ownership and practices the findings cannot be generalized to them. However, calls were made to each zip code in the county, and the number of respondents completing the survey in each zip code, were proportionate to their share of the county population. Although the finding that shelter intake declined in association with the inception of the voucher program, the presence of extraneous (confounding) factors associated both with time and shelter intake cannot be ruled out, including migration into and out of the county (although the human population actually increased during the study period). At the time of the study, two additional smaller shelters existed in the county: a county facility in San Martin, which served the 5% of the population not living within cities, and a city facility in Palo Alto, which only served Palo Alto residents; these shelters were not expected to have any meaningful impact on intake changes in the shelters in this study. In conclusion, this study demonstrates the financial and societal value of instituting a low-cost voucher program on a county-wide scale. Although the parameters utilized in the projections and models in this research (e.g., fecundity and mortality) will vary, perhaps substantially, from county to county, they are realistic and based on published observations. It is therefore likely that the qualitative \u2013 if not the precise quantitative -- benefits of the voucher program in Santa Clara County will be of significance if incepted elsewhere. 17 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 Pre Prin ts Pre Prin ts study population : The same private survey firm used in 1993 was commissioned to conduct a similar random telephone survey of 1,000 households throughout Santa Clara County except Palo Alto (which has its own small shelter and did not participate in 1993). An equal probability of selection method (EPSEM) phone list of residential landline telephone numbers for the survey 5 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 Pre Prin ts Pre Prin ts was purchased from a private company.1 Three attempts were made at each number over successive nights. Over 7,000 calls (including disconnected lines, no answers, refusals) were attempted to reach 1000 respondents. People who agreed to be questioned were asked whether or not they owned dogs or cats, fed stray dogs or cats, whether or not the animals had been altered, if they had reproduced, if cats had been declawed, how they obtained their pets, whether or not cats were allowed outside, purebred status, city of residence, and residence type. Data was initially recorded on written interview forms, and manually entered into a Microsoft Excel 2007 (Microsoft Corporation, Redmond, WA) spreadsheet for statistical analysis. Animal shelter entry information was provided by Santa Clara County Animal Control, Humane Society of Silicon Valley, San Jose Animal Care and Services, and for comparative purposes the Peninsula Humane Society and SPCA in neighboring San Mateo County and Los Angeles County Animal Control. Spay/neuter voucher program information, costs and statistics were obtained from the City of San Jose, and County of Santa Clara. San Mateo County was chosen for comparative purposes, as it most closely resembled Santa Clara County, as opposed to the other four more rural surrounding counties. statistical analysis : The 1982-1993 Santa Clara County shelter intake records (from before the launch of the spay/neuter voucher program) were used for projecting the expected numbers of shelter intakes from 1994 to 2005. US Census data was used to determine the number of county households, which was used with survey-derived estimates of the average number of dogs and cats per household and proportion of households that owned dogs and cats to estimate the number of owned dogs and cats in the county, and with survey-derived estimates of the percentage of 6 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 Pre Prin ts Pre Prin ts households feeding stray dogs and the average number of unowned but fed dogs and cats per household to estimate the number of stray dogs and cats in the county. Data from the survey was initially described using proportions. Pearson\u2019s chi-square test was used to compare proportions; p-values < 0.05 were considered statistically significant. The observed numbers of shelter intakes between 1994 and 2005, during which the spay/neuter program was in place, were compared with projected numbers based on varying the proportion of cats in the voucher program that were owned versus unowned to assess the program\u2019s effect (i.e., change in numbers of shelter intakes). Autoregressive integrated moving average (ARIMA (p, q, d, where p = order of autoregression, q = order of moving averages, and d = order of differencing)) models were fit to the data before the launch of the spay/neuter voucher program (1982 to 1993). Autocorrelation function (ACF) and partial autocorrelation function (PACF) plots were used to select the best-fit ARIMA models and to evaluate the model fit. The selected ARIMA models were then used to estimate and project the trend in number of shelter intakes after the implementation of spay/neuter program (1994 to 2005) with corresponding 95% confidence limits of the ARIMA projections. In addition, we constructed a stochastic model to estimate the number of additional cats that would have been born and taken into the shelters between 1994 and 2005 had the spay/neuter program never been implemented. The key parameters, their corresponding probability distributions for owned and unowned cats, and the data sources are listed in Table 1. This model was also used for benefit-cost analysis of the voucher program. The software program @Risk (version 5.0.0, Palisade Corp., Ithaca, New York) was used for the simulation, using Latin Hypercube sampling and Mersenne Twister generator with a fixed initial seed of 12345 for 10,000 iterations. Median and the 5th and 95th percentiles were reported. 7 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 Pre Prin ts Pre Prin ts dog survey results : Twenty nine percent of responding county households reported that they owned dogs (unchanged from 1993); the average household owned 1.9 dogs, representing an increase from 1.3 dogs in 1993. Using US census data led to an estimate of 332,000 owned dogs in Santa Clara County (assuming Palo Alto has the same ownership frequency). Registered and unregistered purebred dogs were 33% and 18% (total = 51%) of the dog population, respectively; the remainder (49%) of dogs was either mixed or unknown breeds. Dogs were acquired from a variety of sources; the most common were friends or relatives (30%), breeders (25%), public or private animal shelter (15%), with the remainder (less than 10% each) coming from a breed rescue group, a newspaper advertisement, found as stray, being born at home, acquired from a pet store, and rare other sources (Fig.1). Seventy five percent (75%) of owners reported surgically sterilizing their dogs. Among those that declined to alter them, 28% of owners said this was a deliberate decision, and none claimed that cost was a justification for not sterilizing. Thirty three percent (33%) of the unaltered dogs were intended for breeding purposes, and 17% were puppies too young for surgery. Of the 99 unaltered dogs, 70 (70.7%) were male and 29 (29.3%) were female. Twenty one households (2%) in the survey acknowledged feeding dogs they did not own, with an average of 2.3 dogs per feeding household. An examination of zip codes indicated that the majority of these dogs were found in the downtown and north and east sides of the city of San Jose. With some exceptions, these areas are in the lower socio-economic range of households in San Jose. Using US census data, this leads to an estimate of approximately 15,650 8 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 Pre Prin ts Pre Prin ts transiently or permanently stray dogs throughout the county, or 4.7% of the county\u2019s dog population. cat survey results : Twenty five percent of households reported owning cats, representing a decrease from 30% in 1993 (p = 0.013). With an average of 1.7 owned cats per household (a figure unchanged since 1993), the county\u2019s owned cat population was estimated at 256,000 cats. Most cats (85%) were characterized as domestic varieties; only 3% were claimed to be registered pedigree (a figure unchanged since 1993), while others were described as unregistered pedigreed or unknown breed. The percentage of cats kept strictly indoors rose from 33% in 1993 to 49% in 2005 (p < 0.001); only 8% were currently described as strictly outdoors, down from 14% in 1993 (p < 0.001). The most common source of owned cats was from a friend or relative (42% in 2005 versus 33% in 1993), followed by being found as a free-roaming homeless cat (20% in 2005 versus 32% in 1993), a public or private animal shelter (16% in 2005 versus 12% in 1993), a breed rescue group (9% in 2005 versus 2% in 1993), a breeder (4% in both years), an ad in a newspaper or adopted or purchased in a pet store (2% in 2005 versus 6% in 1993), a negligible percentage born at home (<1% in 2005 versus 6% in 1993), and the remainder coming from various minor or unknown sources. The p-value comparing the source distribution of owned cats between 2005 and 1993 was < 0.001. In 2005 most cat owners (92.8%) had their cats surgically sterilized, compared to 86% in 1993 (p < 0.001). Within the 7.2% of cats not spayed or neutered, 48% had owners that deliberately did not want their cats to be sterilized, 13% had owners who wanted to retain the cat 9 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 Pre Prin ts Pre Prin ts for breeding, 13% were kittens, 13% had owners claiming that the costs were prohibitive, and the remainder gave two or more reasons (the most common of which was lack of time to transport the cat for surgery). Thus, only approximately 6% of owned cats were sexually mature and capable of breeding, approximately half of which were female. However, less than one-half of 1% of owners of sterilized female cats allowed their cats to have a litter prior to sterilization. The rate of reproduction of owned cats in Santa Clara County was 89 cats per 1000 households, in contrast to the higher rate of 112 kittens per 1000 households in the 1996 National Council on Pet Population Study and Policy (Salman et al., 1998). This may be attributable to the high proportion of altered cats in the county (93%) relative to the comparable 1993 figure and the 2005 national average of 86%. In addition, while in 1993 16% of owned cats had a litter prior to altering, in the current study this figure was less than one-half of 1%. While it was beyond the scope of this study to determine the reasons for this change in attitude, it is likely that greater awareness prompted by considerable multimedia public education about the county-sponsored voucher program instituted in 1993 bore at least some responsibility. When owners were asked about whether their cats were declawed, 8% stated that they were, but 29% of them obtained the cat in that condition. The most common reason given by owners (84%) for declawing was to protect furniture. Owners not electing to declaw their cats protected their furniture through a variety of means, including having scratching posts and mats, using a spray bottle, clipping the claws, applying double-sided tape, and making loud deterrent noises. Many individuals fed stray cats: 7% of household respondents admitted to feeding an average of 3.2 cats, a decrease from 10% with an average of 3.4 cats in 1993. Relying on U.S. census data, the estimated fed stray cat population is therefore approximately 135,000 cats, or 10 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 Pre Prin ts Pre Prin ts approximately 35% of the total owned and fed free-roaming/unowned cat population in the county (391,000 cats, which represents a drop from 416,000 in 1993). Only 5.5% of these cats were either trapped or taken to be surgically sterilized by their people feeding them. Fifty six percent of the cats were fed daily, while the remaining cats were fed from once every other day to only occasionally. The cats were most commonly fed on the doorstep of a person\u2019s home (62%), followed by an office (15%), a park (12%), and a shopping center (<1%). Fed stray cats were either alone or belonged to colonies ranging in size from 2 to 25 cats. Two-thirds of the fed stray cats were too wild to be picked up and were defined as feral; the remaining third were classified as unowned (although some of these may have had owners unknown to the survey respondent). Forty seven percent of the female stray cats were known to have had at least one litter, which is probably a conservative estimate. Over half of the known litters were allowed to remain free and disperse into their neighborhoods. Of the remaining kittens, half were kept by the feeder, while the others were given away or taken to an animal shelter. Of the females who had litters, 58% were not trapped or taken to a veterinarian after having a litter, remaining free to potentially breed again. population changes at santa clara county animal shelters : Changes in dog shelter intakes for Santa Clara County (and the Peninsula Humane Society and SPCA shelter in adjacent San Mateo County for comparison) are shown in Fig. 2. Dog intakes declined 13,643 to 8,441 (38.1%) from 1992-2005 in Santa Clara County. An external explanation for the observed trend is supported by the findings in adjacent San Mateo County, where dog intakes declined by a similar 35.7% between 1990 and 2004. These proportions were not significantly different (p = 0.11). 11 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 Pre Prin ts Pre Prin ts A substantially different picture emerged when examining changes in cat shelter intakes in Santa Clara and San Mateo Counties (Figs. 3 and 4). Intakes in Santa Clara County dropped 22,473 to 16,369 (27.2%) from 1993 to 2004 and 22,473 to 16,807 (25.2%) from 1993 to 2005, compared to a drop of 8,252 to 6,078 (26.3%) in San Mateo County from 1993 to 2004. Although the two 1993 to 2004 proportions were similar (p = 0.16), there was an overall decline in annual intakes in Santa Clara County of 6,104 cats to 2004 (509 cats per year) for the 12 year period, compared to 2,174 cats in the same 12 year period for San Mateo County (181 cats per year). The absolute changes are economically more germane to counties with respect to shelter expenses because expenditures are based on the per diem cost of maintaining individual cats. The results of the ARIMA (1,0,1) projections indicated a higher-than-expected cat intake to shelters in Santa Clara County during the years when the voucher program was in effect, i.e., 1994-2005 (Fig.5). The ARIMA (1,0,1) projections further showed that the observed numbers of cats brought in by the field service did not substantially differ from the expected numbers during the same time period (Fig.6). Information provided by the HSSV shelter indicated that the majority of cats entering the shelter were unweaned kittens and feral cats. From 2000 to 2004, the HSSV euthanized 53,419 cats deemed unadoptable: 14,406 were too young (under four weeks of age), 7,912 were unsociable, and 7,595 were feral. Under the voucher program, 20,419 cats were surgically sterilized from 1994-2001 and an additional 6,231 cats were sterilized from 2001-2003. While the program was initiated at the end of 1994, public interest did not start until mid-1995, when a local television station and newspaper ran a story about it. 12 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 Pre Prin ts Pre Prin ts The San Jose program was initially free to the public; however, various program changes over time were instituted. Veterinarians were reimbursed at a set fee of $25 female and $15 male. Pregnancies could add to the veterinarian reimbursement up to $50, and anatomical issues adjusted the price to as high as $150. In 1996 modifications included requiring a $5 co-pay, and a requirement that cat owners obtain a $5 license and rabies inoculation. While these changes increased the veterinarian reimbursement, they also created a negative effect on the program, as voucher requests declined from 5,600 in the first 16 months of the program to only 2,800 for the year following the changes. The San Jose voucher program ended in 2003, but the county program continued. Utilizing assumptions in Table 1, if no voucher program had been initiated, the same cats enrolled in the voucher program (assuming that 65% were owned, based on the 2005 survey results) would have produced approximately 312,000 kittens between 1994 and 2005, and approximately 8,600 additional cats would have entered (6,200 surrendered and 2,500 brought in by the field service) the shelters in Santa Clara County. This would have incurred an additional cost of approximately $2.15 million, with the HSSV charge to cities for stray cat services under their contract cost of $250 per cat. If the cost per cat for spay/neuter surgery in 2001-2002 ($23.21 average for all surgeries) can be assumed to be constant from 1994-2005, then the expected cost of the HSSV voucher program was approximately $620,000. Thus, the net gain of the program from reducing the number of cat shelter intakes was approximately $1.53 million. Not counted would be the added burden of approximately 44,000 cats to the stray population in the county. The proportion of feral cats actually altered in the program considerably fluctuated: from 77% in 2006 to 82% in 2007 to 48% in 2008 (the latter data is from San Jose only). Table 2 13 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 Pre Prin ts Pre Prin ts contains projections of how county cat and shelter populations would be expected to change in the absence of the voucher program under different owned versus feral cat ratios. Under all plausible scenarios, ranging from 20% to 80% of the altered cats being feral, the costs to the shelters would have likely exceeded $2 million over the 12-year life of the program, and at the higher proportion of feral cats the costs would have likely exceeded $6 million. footnotes : 1 Scientific Telephone Samples, Foothill Ranch, California 22 439 440 441 Pre Prin ts Pre Prin ts Table 1(on next page) Information used in modeling cat population dynamics from 1994 to 2005 if no spay/neuter voucher program had been initiated in Santa Clara County, California. Pre Prin ts Pre Prin ts Parameter Owned cats Unowned cats References Kittens/litter 4.25 3.6 Pedersen, 1991; Scott et al., 2002 Kitten mortality rate (%) 30 75 Jemmett and Evans, 1977; Nutter et al., 2004; Scott et al., 1978 Life expectancy (years) 12 4.7 Levy et al., 2003; New et al. 2004 Litters per female per year 2.1 1.4 Pedersen, 1991, Nutter et al., 2004; Levy et al., 2003; Scott et al., 2002 Percent female 55 45 Levy et al., 2003 Sexually intact (%) 14 94.5 1993 and 2005 surveys Surrendered to shelter (%) 3.0 7.3 1993 survey and shelter statistics Pre Prin ts Pre Prin ts Table 2(on next page) Projected impact of hypothetical absence of the 12 year spay/neuter program on cat populations, shelter intake, and municipal cost in Santa Clara County. Median and the 5th and 95th percentiles (in parentheses) are reported (x $1,000). Pre Prin ts Pre Prin ts Percentage of surgeries performed on owned cats Additional number of owned cats Additional number of stray cats Cats voluntarily surrendered to shelter Cats brought in by field service Total additional shelter cat intake Cost to shelter for additional surrendered and stray cats 20 193 265 6 20 25 $6,333 (110, 388) (131, 593) (3, 12) (10, 44) (13, 55) (3,242, 13,817) 30 226 181 7 13 20 $5,034 (140, 412) (98, 367) (4, 13) (7, 27) (11, 40) (2,867, 9,893) 40 239 124 7 9 16 $4,078 (159, 398) (73, 226) (5, 12) (5, 17) (10, 28) (2,549, 7,089) 50 242 84 7 6 13 $3,351 (173, 368) (54, 138) (5, 11) (4, 10) (9, 21) (2,282, 5,291) 60 241 55 7 4 11 $2,823 (180, 342) (38, 85) (5, 10) (3, 6) (8, 17) (2,035, 4,135) 70 236 35 7 3 10 $2,411 (182, 319) (25, 50) (5, 10) (2, 4) (7, 13) (1,814, 3,324) 80 231 20 7 1 8 $2,099 (182, 303) (15, 27) (5, 9) (1, 2) (6, 11) (1,625, 2,792) Pre Prin ts Pre Prin ts Figure 1 Source of acquisition of dogs from Santa Clara County survey, 2005. Pre Prin ts Pre Prin ts Figure 2 Regression analysis of intake of dogs at shelters in Santa Clara County (r = 0.95) and San Mateo County (r = 0.97) over time (1990 \u2013 2005). Pre Prin ts Pre Prin ts Figure 3 Secular changes in cat intakes in Santa Clara County, 1982 \u2013 2007, indexed by historically relevant events. Pre Prin ts Pre Prin ts Figure 4 Regression analysis of intake of cats at shelters in Santa Clara (r=0.98) and San Mateo Counties (r>0.99), 1990 \u2013 2006. Field services in Santa Clara County ended in 1992; at that time 60% of cats were brought in through field services. Field services resumed in November 1993 in some cities. Pre Prin ts Pre Prin ts Figure 5 Observed numbers of cats surrendered to the shelters in Santa Clara County versus autoregressive integrated moving average (ARIMA) projected numbers of surrendered cats. Figure uses the 1982-1993 data (before the launch of the spay/neuter voucher program) shelter data. The lower (LCL) and upper (UCL) 95% confidence limits of the ARIMA projection are also presented. Pre Prin ts Pre Prin ts Figure 6 Observed numbers of cats brought to the shelters in Santa Clara County by the field service versus the autoregressive integrated moving average (ARIMA) projected numbers of cats. Figure shows cats brought in by field service using the 1982-1993 (before the launch of the spay/neuter voucher program, shelter data). The lower (LCL) and upper (UCL) 95% confidence limits of the ARIMA projection are also presented. Pre Prin ts Pre Prin ts",
+    "v2_text": "materials and methods : results : discussion : This study documents the multiple positive impacts publically subsidized low-cost spay and neuter programs can have on animal welfare and society needs. These impacts fulfill societal needs that often go unmet in communities: pet population control, prevention of the proliferation of feral dog and cat populations, reduction of communicable (including zoonotic) disease, slowing the flow of animals into shelters both voluntarily and through field services, and reduction in the incidence of humane destruction of animals. But they also extend to issues presumably of more immediate importance to the economic viability of communities; namely, reduction in capital and ongoing expenditures when the latter are proportional to the number of cats that come under a municipality\u2019s immediate jurisdiction and care. In contrast, the implication of cessation of such programs is clear, as can be seen in Figures 2 and 3 when the decline in shelter admissions of dogs and cats became attenuated. The problem can only be exacerbated over time as the human (and hence pet-owning) population increases, as in California counties historically experiencing net population growth. The finding that there were over 15,000 dogs (4.7% of the county\u2019s dog population) estimated to be transiently or permanently stray throughout the county is troubling from societal and public health standpoints. The absence of a domestic environment can lead stray dogs, which are by nature gregarious, to form packs that can become aggressive and endanger other animals and even humans. The origin of such a large number of dogs is worthy of further 14 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 Pre Prin ts Pre Prin ts research, as this study did not explore whether these were free-roaming dogs or those kept in temporary foster or rescue care. Specific breed information was not available for dogs in Santa Clara County animal shelters. Respondents claimed 51% of their dogs were registered and unregistered purebred dogs. This stands in contrast to a 1996 national survey that found 30% of dogs relinquished to shelters were purebred (Salman et al., 1998), and the Humane Society of the United States estimates that 25-30% of shelter dogs are purebred (The Humane Society of the United States, 2011). Nationally, purebred dogs are substantially less likely to be relinquished to animal shelters than dogs of mixed breed (Salman et al., 1998). The dogs with greatest likelihood of successful adoption from county animal shelters are puppies (with pit bull-like breeds an exception). By the time dogs reach the age of one year, though, their risk of unsuccessful adoption following relinquishment rises considerably; again, particularly true in pit bull-like breeds. Aggressive dog behavior is the primary reason dogs are euthanized at the county shelters. To reduce dog intakes, municipalities should consider how the establishment of free or low cost puppy training programs (potentially mandatory for shelter adoptions) might impact shelter populations. A collaborative effort among multiple community agencies, including animal control, non-profits, and local pet industry businesses should be explored. Another important finding is the enumeration of the substantial unowned cat population in Santa Clara County, two-thirds of which are feral. Also notable is that the majority of unowned cats entering the animal shelters in the study were arguably unsuitable for adoption, with over 50% being feral or unweaned kittens. Such cats are often quickly euthanized. Preventing such input defies simplistic solutions, because although 93% of cat owners were 15 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 Pre Prin ts Pre Prin ts willing to have their own pets surgically sterilized, it is unrealistic to expect the 7% of the population that feeds an average of 3 stray cats to assume the hundreds of dollars necessary to surgically alter these cats. Conversely, the cost of not altering the cats is to add 3.5 kittens per year for each stray female, which at the cost to a shelter of approximately $250 per cat would cost a shelter almost $900 in husbandry expenses for those 3.5 kittens. The underscores why low-cost spay and neuter programs directed to reducing the un-owned and feral cat populations continue to be integral to not only reducing cat mortality at the shelters, but also to managing the cost to the various municipalities to handle and house the stray cats. Santa Clara County\u2019s contribution of $45 to alter a stray cat under its separate feral spay/neuter program created , an immediate savings of over $200 for just the first litter that permanently results in non-reproducing cats. The county program also subsidized shots, and for a time, FELV testing. The earlier such cats can be sterilized, the greater the potential savings to municipalities. Moreover, if stray cat-feeding citizens can be convinced through public education to avail themselves of population control options by making them more affordable and they are provided with instructions and resources as to how to accomplish this activity, the savings in costs and lives will be substantial. This study shows approximately 93% of county residents did not make an effort to sterilize unowned cats. Only 5.5% of the unowned but fed cats were surgically sterilized. Efforts should be focused on removing barriers and finding ways to encourage those who feed free-roaming cats to take this important step. Because the study shows that 62% of stray cats are fed in people\u2019s yards, efforts should be intensified to sterilize cats living in close proximity to homes, rather than less accessible colonies. 16 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 Pre Prin ts Pre Prin ts This study\u2019s limitations include the assumption that the participating individuals are representative of the county\u2019s population. Interviews were conducted via telephone, with the non-telephone-owning segment of the county excluded by design, and to the extent that this subgroup differs in their pet ownership and practices the findings cannot be generalized to them. However, calls were made to each zip code in the county, and the number of respondents completing the survey in each zip code, were proportionate to their share of the county population. Although the finding that shelter intake declined in association with the inception of the voucher program, the presence of extraneous (confounding) factors associated both with time and shelter intake cannot be ruled out. In conclusion, this study demonstrates the financial and societal value of instituting a low-cost voucher program on a county-wide scale. Although the parameters utilized in the projections and models in this research (e.g., fecundity and mortality) will vary, perhaps substantially, from county to county, they are realistic and based on published observations. It is therefore likely that the qualitative \u2013 if not the precise quantitative -- benefits of the voucher program in Santa Clara County will be of significance if incepted elsewhere. 17 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 Pre Prin ts Pre Prin ts study population : The same private survey firm used in 1993 was commissioned to conduct a similar random telephone survey of 1,000 households throughout Santa Clara County except Palo Alto (which has its own small shelter and did not participate in 1993). An EPSEM (equal probability of selection method) phone list of residential landline telephone numbers for the survey was purchased from a private company.1 Three attempts were made at each number over successive nights. Over 7,000 calls were attempted to reach 1000 respondents. People who agreed to be questioned were asked about pet ownership, pet characteristics, method of pet acquisition, indoor versus outdoor status, neutering and cat declawing status, feeding of unowned cats and dogs, and city of residence within county. Animal shelter entry information was obtained from Santa Clara County Animal Control, Humane Society of Silicon Valley, San Jose Animal Care and Services, and for comparative purposes the Peninsula Humane Society and SPCA in neighboring San Mateo County and Los Angeles County Animal Control. San Mateo County was chosen for comparative purposes, as it most closely resembled Santa Clara County, as opposed to the other four more rural surrounding counties. 6 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 Pre Prin ts Pre Prin ts statistical analysis : An autoregressive integrated moving average (ARIMA) time series analysis was used to estimate and project the trend in number of shelter intakes. The 1982-1993 Santa Clara County shelter intake records (from before the launch of the spay/neuter voucher program) were used for projecting the expected numbers of shelter intakes from 1994 to 2005. The observed numbers of shelter intakes between 1994 and 2005, during which the spay/neuter program was in place, were then compared with the projected numbers to assess the effect (i.e., change in numbers of shelter intakes) of the spay/neuter voucher program. Autocorrelation function (ACF) and partial autocorrelation function (PACF) plots were used to select appropriate ARIMA models and to evaluate the model fit. Corresponding 95% confidence limit of the ARIMA projections were also presented. A stochastic model was constructed to estimate the number of additional cats that would have been born and taken into the shelters between 1994 and 2005 had the spay/neuter program never been implemented. The key parameters, their corresponding probability distributions for owned and unowned cats, and the data sources are listed in Table 1. This model was also used for benefit-cost analysis of the voucher program. The software program @Risk (version 5.0.0, Palisade Corp., Ithaca, New York) was used for the simulation with 10,000 iterations. Median and the 5th and 95th percentiles were reported. dogs : Twenty nine percent of responding county households resported that they owned dogs (unchanged from 1993); the average household owned 1.9 dogs, representing an increase from 7 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 Pre Prin ts Pre Prin ts 1.3 dogs in 1993. Using US census data led to an estimate of 332,000 owned dogs in Santa Clara County (assuming Palo Alto has the same ownership frequency). Registered and unregistered purebred dogs were 33% and 18% (total = 51%) of the dog population, respectively; the remainder (49%) of dogs was either mixed or unknown breeds. Dogs were acquired from a variety of sources; the most common were friends or relatives (30%), breeders (25%), public or private animal shelter (15%), with the remainder (less than 10% each) coming from a breed rescue group, a newspaper advertisement, found as stray, being born at home, acquired from a pet store, and rare other sources (Fig.1). Seventy five percent (75%) of owners reported surgically sterilizing their dogs. Among those that declined to alter them, 28% of owners said this was a deliberate decision, and none claimed that cost was a justification for not sterilizing. Thirty three percent (33%) of the unaltered dogs were intended for breeding purposes, and 17% were puppies too young for surgery. Of the 99 unaltered dogs, 70 (70.7%) were male and 29 (29.3%) were female. Twenty one households (2%) in the survey acknowledged feeding dogs they did not own, with an average of 2.3 dogs per feeding household. An examination of zip codes indicated that these dogs were predominantly found in the downtown and north and east sides of the city of San Jose. With some exceptions, these areas are in the lower socio-economic range of households in San Jose. Using US census data, this leads to an estimate of approximately 15,650 transiently or permanently stray dogs throughout the county, or 4.7% of the county\u2019s dog population. cats : Twenty five percent of households reported owning cats, representing a decrease from 30% in 1993 (p = 0.013). With an average of 1.7 owned cats per household (a figure unchanged 8 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 Pre Prin ts Pre Prin ts since 1993), the county\u2019s owned cat population was estimated at 256,000 cats. Most cats (85%) were characterized as domestic varieties; only 3% were claimed to be registered pedigree (a figure unchanged since 1993), while others were described as unregistered pedigreed or unknown breed. The percentage of cats kept strictly indoors rose from 33% in 1993 to 49% in 2005 (p < 0.001); only 8% were currently described as strictly outdoors, down from 14% in 1993 (p < 0.001). The most common source of owned cats was from a friend or relative (42% in 2005 versus 33% in 1993), followed by being found as a free-roaming homeless cat (20% in 2005 versus 32% in 1993), a public or private animal shelter (16% in 2005 versus 12% in 1993), a breed rescue group (9% in 2005 versus 2% in 1993), a breeder (4% in both years), an ad in a newspaper or adopted or purchased in a pet store (2% in 2005 versus 6% in 1993), a negligible percentage born at home (<1% in 2005 versus 6% in 1993), and the remainder coming from various minor or unknown sources. The p-value comparing the source distribution of owned cats between 2005 and 1993 was < 0.001. In 2005 most cat owners (92.8%) had their cats surgically sterilized, compared to 86% in 1993 (p < 0.001). Within the 7.2% of cats not spayed or neutered, 48% had owners that deliberately did not want their cats to be sterilized, 13% had owners who wanted to retain the cat for breeding, 13% were kittens, 13% had owners claiming that the costs were prohibitive, and the remainder gave two or more reasons (the most common of which was lack of time to transport the cat for surgery). Thus, only approximately 6% of owned cats were sexually mature and capable of breeding, approximately half of which were female. However, less than one-half of 1% of owners of sterilized female cats allowed their cats to have a litter prior to sterilization. 9 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 Pre Prin ts Pre Prin ts The rate of reproduction of owned cats in Santa Clara County was 89 cats per 1000 households, in contrast to the higher rate of 112 kittens per 1000 households in the 1996 National Council on Pet Population Study and Policy (Salman et al., 1998). This may be attributable to the high proportion of altered cats in the county (93%) relative to the comparable 1993 figure and the 2005 national average of 86%. In addition, while in 1993 16% of owned cats had a litter prior to altering, in the current study this figure was less than one-half of 1%. While it was beyond the scope of this study to determine the reasons for this change in attitude, it is likely that greater awareness prompted by considerable multimedia public education about the county-sponsored voucher program instituted in 1993 bore at least some responsibility. When owners were asked about whether their cats were declawed, 8% stated that they were, but 29% of them obtained the cat in that condition. The most common reason given by owners (84%) for declawing was to protect furniture. Owners not electing to declaw their cats protected their furniture through a variety of means, including having scratching posts and mats, using a spray bottle, clipping the claws, applying double-sided tape, and making loud deterrent noises. Many individuals fed stray cats: 7% of household respondents admitted to feeding an average of 3.2 cats, a decrease from 10% with an average of 3.4 cats in 1993. Relying on U.S. census data, the estimated fed stray cat population is therefore approximately 135,000 cats, or approximately 35% of the total owned and fed free-roaming/unowned cat population in the county (391,000 cats, which represents a drop from 416,000 in 1993). Only 5.5% of these cats were either trapped or taken to be surgically sterilized by their people feeding them. Fifty six percent of the cats were fed daily, while the remaining cats were fed from once every other day to only occasionally. The cats were most commonly fed on the doorstep of a person\u2019s home 10 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 Pre Prin ts Pre Prin ts (62%), followed by an office (15%), a park (12%), and a shopping center (<1%). Fed stray cats were either alone or belonged to colonies ranging in size from 2 to 25 cats. Two-thirds of the fed stray cats were too wild to be picked up and were defined as feral; the remaining third were classified as unowned (although some of these may have had owners unknown to the survey respondent). Forty seven percent of the female stray cats were known to have had at least one litter, which is probably a conservative estimate. Over half of the known litters were allowed to remain free and disperse into their neighborhoods. Of the remaining kittens, half were kept by the feeder, while the others were given away or taken to an animal shelter. Of the females who had litters, 58% were not trapped or taken to a veterinarian after having a litter, remaining free to potentially breed again. population changes at santa clara county animal shelters : Changes in dog shelter intakes for Santa Clara County (and the Peninsula Humane Society and SPCA shelter in adjacent San Mateo County for comparison) are shown in Fig. 2. Dog intakes declined 13,643 to 8,441 (38.1%) from 1992-2005 in Santa Clara County in the absence of a widely promoted municipally funded spay/neuter program, indicating that other external factors played an important role in the number of stray and surrendered dogs brought to shelters. An external explanation for the observed trend is supported by the findings in adjacent San Mateo County, where dog intakes declined by a similar 35.7% between 1990 and 2004. These proportions were not significantly different (p = 0.11). A substantially different picture emerged when examining changes in cat shelter intakes in Santa Clara and San Mateo Counties (Figs. 3 and 4). Intakes in Santa Clara County dropped 22,473 to 16,369 (27.2%) from 1993 to 2004 and 22,473 to 16,807 (25.2%) from 1993 to 2005, 11 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 Pre Prin ts Pre Prin ts compared to a drop of 8,252 to 6,078 (26.3%) in San Mateo County from 1993 to 2004. Although the two proportions were similar, there was an overall decline in annual intakes in Santa Clara County of 6,104 cats to 2004 (509 cats per year) for the 12 year period, compared to 2,174 cats in the same 12 year period for San Mateo County (181 cats per year). The absolute changes are economically more germane to counties with respect to shelter expenses because expenditures are based on the per diem cost of maintaining individual cats. The results of the ARIMA projections indicated a higher-than-expected decline in the number of cats being surrendered to the shelters in Santa Clara County during the years when the voucher program was in effect, i.e., 1994-2005 (Fig.5). The ARIMA projections further showed that the observed numbers of cats brought in by the field service did not substantially differ from the expected numbers during the same time period (Fig.6). Under the voucher program, 20,419 cats were surgically sterilized from 1994-2001 and an additional 6,231 cats were sterilized from 2001-2003. While the program was initiated at the end of 1994, public interest did not start until mid-1995, when a local television station and newspaper ran a story about it. The San Jose program was initially free to the public; however, various program changes over time were instituted. Veterinarians were reimbursed at a set fee of $25 female and $15 male. Pregnancies could add to the veterinarian reimbursement up to $50, and anatomical issues adjusted the price to as high as $150. In 1996 modifications included requiring a $5 co-pay, and a requirement that cat owners obtain a $5 license and rabies inoculation. While these changes increased the veterinarian reimbursement, they also created a negative effect on the program, as voucher requests declined from 5,600 in the first 16 months of the program to only 2,800 for the year following the changes. 12 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 Pre Prin ts Pre Prin ts The San Jose voucher program ended in 2003, but the county program continued. Utilizing assumptions in Table 1, if no voucher program had been initiated, these cats (assuming that 65% were owned, based on the 2005 survey results) would have produced approximately 312,000 kittens between 1994 and 2005, and approximately 8,600 additional cats would have entered (6,200 surrendered and 2,500 brought in by the field service) the shelters in Santa Clara County. This would have incurred an additional cost of approximately $2.15 million, with the HSSV charge to cities for stray cat services under their contract cost of $250 per cat. If the cost per cat for spay/neuter surgery in 2001-2002 ($23.21 average for all surgeries) can be assumed to be constant from 1994-2005, then the expected cost of the HSSV voucher program was approximately $620,000. Thus, the net gain of the program from reducing the number of cat shelter intakes was approximately $1.53 million. Not counted would be the added burden of approximately 44,000 cats to the stray population in the county. However, a different picture emerges if the proportion of owned cats surgically altered under the voucher program varied from the 65% figure utilized above. In fact, the proportion of feral cats actually altered in the program considerably fluctuated: from 77% in 2006 to 82% in 2007 to 48% in 2008 (the latter data is from San Jose only). Table 2 provides an estimate of how county cat and shelter populations would be expected to change in the absence of the voucher program under different owned versus feral cat ratios. Under all plausible scenarios, ranging from 20% to 80% of the altered cats being feral, the costs to the shelters would have likely exceeded $2 million over the 12-year life of the program, and at the higher proportion of feral cats the costs would have likely exceeded $6 million. The savings would be expected to grow over additional years. This underscores that voucher programs have not only had a restraining effect for the population control of unowned cats, but for owned cats as well. Moreover, under 13 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 Pre Prin ts Pre Prin ts all such plausible scenarios the voucher program would have resulted in a net savings in expenditure. footnotes : 1 Scientific Telephone Samples, Foothill Ranch, California 21 417 418 419 Pre Prin ts Pre Prin ts Table 1(on next page) Key parameters, their values for owned and unowned cats, and the data sources used in modeling cat population dynamics from 1994 to 2005 if no spay/neuter voucher program had been initiated in Santa C Pre Prin ts Pre Prin ts Parameter Owned cats Unowned cats References Kittens/litter 4.25 3.6 Pedersen, 1991; Scott et al., 2002 Kitten mortality rate (%) 30 75 Jemmett and Evans, 1977; Nutter et al., 2004; Scott et al., 1978 Life expectancy (years) 12 4.7 Levy et al., 2003; New et al. 2004 Litters per female per year 2.1 1.4 Pedersen, 1991, Nutter et al., 2004; Levy et al., 2003; Scott et al., 2002 Percent female 55 45 Levy et al., 2003 Sexually intact (%) 14 94.5 1993 and 2005 surveys Surrendered to shelter (%) 3.0 7.3 1993 survey and shelter statistics Pre Prin ts Pre Prin ts Table 2(on next page) Projected impact of hypothetical absence of the 12 year spay/neuter program on cat populations, shelter intake, and municipal cost in Santa Clara County. Median and the 5th and 95th percentiles (in parentheses) are reported (x $1,000). Pre Prin ts Pre Prin ts Percentage of surgeries performed on owned cats Additional number of owned cats Additional number of stray cats Cats voluntarily surrendered to shelter Cats brought in by field service Total additional shelter cat intake Cost to shelter for additional surrendered and stray cats 20 193 265 6 20 25 $6,333 (110, 388) (131, 593) (3, 12) (10, 44) (13, 55) (3,242, 13,817) 30 226 181 7 13 20 $5,034 (140, 412) (98, 367) (4, 13) (7, 27) (11, 40) (2,867, 9,893) 40 239 124 7 9 16 $4,078 (159, 398) (73, 226) (5, 12) (5, 17) (10, 28) (2,549, 7,089) 50 242 84 7 6 13 $3,351 (173, 368) (54, 138) (5, 11) (4, 10) (9, 21) (2,282, 5,291) 60 241 55 7 4 11 $2,823 (180, 342) (38, 85) (5, 10) (3, 6) (8, 17) (2,035, 4,135) 70 236 35 7 3 10 $2,411 (182, 319) (25, 50) (5, 10) (2, 4) (7, 13) (1,814, 3,324) 80 231 20 7 1 8 $2,099 (182, 303) (15, 27) (5, 9) (1, 2) (6, 11) (1,625, 2,792) Pre Prin ts Pre Prin ts Figure 1 Source of acquisition of dogs from Santa Clara County survey, 2005. Pre Prin ts Pre Prin ts Figure 2 Regression analysis of intake of dogs at shelters in Santa Clara County (r = 0.95) and San Mateo County (r = 0.97) over time (1990 \u2013 2005). Pre Prin ts Pre Prin ts Figure 3 Secular changes in cat intakes in Santa Clara County, 1982 \u2013 2007, indexed by historically relevant events. SC = Santa Clara; SJ = San Jose; S/N = spay/neuter voucher program. Pre Prin ts Pre Prin ts Figure 4 Regression analysis of intake of cats at shelters in Santa Clara (r=0.98) and San Mateo Counties (r>0.99), 1990 \u2013 2006. Field services in Santa Clara County ended in 1992; at that time 60% of cats were brought in through field services. Field services resumed in November 1993 in some cities. Pre Prin ts Pre Prin ts Figure 5 Observed numbers of cats surrendered to the shelters in Santa Clara County versus autoregressive integrated moving average (ARIMA) projected numbers of surrendered cats. Figure uses the 1982-1993 data (before the launch of the spay/neuter voucher program) shelter data. The lower (LCL) and upper (UCL) 95% confidence limits of the ARIMA projection are also presented. Pre Prin ts Pre Prin ts Figure 6 Observed numbers of cats brought to the shelters in Santa Clara County by the field service versus the autoregressive integrated moving average (ARIMA) projected numbers of cats. Figure shows cats brought in by field service using the 1982-1993 (before the launch of the spay/neuter voucher program, shelter data). The lower (LCL) and upper (UCL) 95% confidence limits of the ARIMA projection are also presented. Pre Prin ts Pre Prin ts",
+    "url": "https://peerj.com/articles/19/reviews/",
+    "review_1": "David Reser \u00b7 Jan 5, 2013 \u00b7 Academic Editor\nACCEPT\nCongratulations, and I appreciate the quick turnaround on the most recent version. I look forward to seeing the published document.",
+    "review_2": "David Reser \u00b7 Jan 4, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nThank you for your thorough response to the points raised in the previous reviews. I believe this manuscript will be acceptable, pending amendment/explanation of the point raised by Reviewer 1 regarding the statistical values in the Results section of the amended manuscript.",
+    "review_3": "Reviewer 1 \u00b7 Jan 4, 2013\nBasic reporting\nI am satisfied that the author's of this manuscript have addressed my earlier concerns.\nExperimental design\nI am satisfied that the author's of this manuscript have addressed my earlier concerns.\nValidity of the findings\nI am satisfied that the author's of this manuscript have addressed my earlier concerns.\n\nI want to point out that in the Results-> Perceptual Reports section, all of the chi-square and AUC values are identical despite different significance levels. I am not sure if this is intended, if not the author's should consider correcting the values prior to publication.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Perceptual elements in Penn & Teller\u2019s \u201cCups and Balls\u201d magic trick (v0.2)\". PeerJ https://doi.org/10.7287/peerj.19v0.2/reviews/1",
+    "review_4": "David Reser \u00b7 Dec 18, 2012 \u00b7 Academic Editor\nMAJOR REVISIONS\nBoth reviewers requested substantive changes to the figures and methods description, and Reviewer 1 identifies several potential problem areas in the statistical analysis. In your reply letter, please explicitly identify the changes in the manuscript which address these concerns.\nReviewer 2 points out that the magician's own expectation of what elements were important for misdirection of the viewer was inconsistent with the data, which is briefly addressed in the Discussion. In my opinion, it would be worthwhile to expand this point in the discussion beyond the single paragraph in the original submitted manuscript, and clarify which perceptual or cognitive elements contribute to the expectations of the viewer, especially the naive viewer. Please note that I leave this suggestion to the discretion of the authors, and this specific point will not affect my final decision if the other issues called out by the reviewers are adequately addressed.\nPlease ensure that the revised manuscript conforms to the PeerJ policy regarding data and materials sharing:\nData and Materials Sharing\n1 PeerJ is committed to improving scholarly communications and as part of this commitment, all authors are responsible for making materials, data and associated protocols available to readers without delay. The preferred way to meet this requirement is to publicly deposit as noted below. Cases of non-compliance will be investigated by PeerJ, which reserves the right to act on the results of the investigation.\n...\n- Where suitable domain-specific repositories do not exist, authors may deposit in either Dryad or an institutional repository and provide the access information with the manuscript. Alternately, authors may choose to deposit non-standard data (including figures, posters, rich media) on Figshare or PeerJ PrePrints, for example. In all cases, the DOI reference (where applicable) should be provided in the article.\n- Any supporting data sets for which there are no suitable repositories must be made available as publishable Supplemental Information files by PeerJ.\n...\n\nThe full policy is viewable in Policy and Procedures page at: https://peerj.com/about/policies-and-procedures/#data-materials-sharing.",
+    "pdf_1": "https://peerj.com/articles/19v0.3/submission",
+    "pdf_2": "https://peerj.com/articles/19v0.2/submission",
+    "review_5": "Reviewer 1 \u00b7 Dec 18, 2012\nBasic reporting\nNo comments.\nExperimental design\nA lot of the points I make with regard to \"Experimental Design\" have flow on effects to the \"Validity of the Findings\". I keep them together in one section for clarity.\n\nMethods & Results:\n\u2022 Please provide more detail as to the experimental equipment setup. What was the resolution of the stimulus (in pixels and in degrees of visual angle), What was the luminance, and contrast of the stimulus? Was the experiment conducted in a controlled environment (dark, quiet?).\n\u2022 Eye movements: You note that blink periods were removed. Were the subjects told to fixate on a particular part of the stimulus? If so then, how can you generalise your findings to a natural viewing of the magic trick. If not, then given that saccadic suppression is a very well known and studied phenomenon did you attempt to assess whether the key frames of the video that contained the manipulation occurred during a saccadic eye movement?\n\u2022 Given that one of your measures of interest was reaction time, were removals (button 1) and placings (button 2) equally easy to press?\n\u2022 How was the order of stimuli determined? If it was random was the same order used for each subject? This has profound implications for the validity of your learning findings.\n\u2022 Data analysis. Please provide more details about the results of your data analysis. For regression analyses please provide measures of goodness of fit.\n\u2022 Given that you are using the Newman-Kuels method to do pair-wise comparisons, please explain how you are controlling for the increasing experiment-wise alpha level given the number of independent variables you are testing.\n\u2022 You mention that your \u201cstatistical models determined main effects and first order interactions\u201d but you don\u2019t report these anywhere.\n\u2022 In several places throughout the paper you mention performance without qualifying which of the several measures of performance you tested you are referring to.\n\u2022 Given that you suggest that there was a learning effect how did you deal with this in the main analysis. Did you average all trials? Na\u00efve trials (such as would be most relevant to a person seeing a trick for the first time)?, learned trials?\nFigures\n\u2022 Please be consistent with your error-bars, you switch from 95% confidence intervals (figure 2) to SEM (Figures 3-4) and this can lead to a cursory reader assuming that later figures have more significance than they actually do.\n\u2022 Figure 2ABC) Please indicate where chance performance would be on these figures. Are the subjects worse than chance when tested with opaque cups?\n\u2022 Figure 3. Please indicate in the figure caption that dva refers to degrees of visual angle? or change the axis label to reflect this.\n\u2022 Figure4a. Please indicate what number the # of late findings is in relation to.\n\u2022 Please explain why you don't present the graph for opaque cups in Figure 4B&C.\n\u2022 I am concerned that given the findings concerning your other variables, comparisons of trial number to your dependent variables are potentially confounded by order effects. Without knowing how the conditions were ordered it is not possible to evaluate the data being presented in Figure 4.\nValidity of the findings\nSee my comments in the \"Experimental Design\" section.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Perceptual elements in Penn & Teller\u2019s \u201cCups and Balls\u201d magic trick (v0.1)\". PeerJ https://doi.org/10.7287/peerj.19v0.1/reviews/1",
+    "pdf_3": "https://peerj.com/articles/19v0.1/submission",
+    "review 6": "Hsin-Hao Yu \u00b7 Dec 11, 2012\nBasic reporting\nThe scientific part of the paper is well-written, but I found the non-scientific part, namely the description of the the performance of the magic, difficult to comprehend. Maybe I am simply not familiar with stage magic but it took me a long time to get a sense of what the magician actually did. The inclusion of a video clip will greatly improve the readability of the paper. If that is not possible due to copyright or technical issues, I think the introduction needs to be revised to give more detailed and precise description of the performance. Since I was not familiar with Cups and Balls, I turned to youtube, hoping to see how Penn and Teller perform the Cups and Balls magic. This probably was not such a good idea because Cups and Balls, as performed in Penn and Teller\u2019s various stage shows and television programs, appeared to be not exactly the same Cups and Balls magic studied in the paper (in the youtube clips, the magician does not tilt the transparent cup to make the ball fall into his hand). This is a source of confusion that can be avoided if the introduction provides more background information. The second paragraph of Introduction describes how the magic was invented. According to the description, the magician was surprised by how effective the illusion was. The falling of the ball distracted the attention of the magician such that he himself did not see the loading of the cup. However, Figure 2 clearly shows that the subjects could reliably detect the loading of the cup, thus showing that the magic was not magical, after all! How can a successful magician be so wrong about the effectiveness of his performance? I\u2019m puzzled. The description of the magician\u2019s invention of the magic, rather than helping to motivate the experiment, only confused this reviewer.\nExperimental design\nThe design of the experiment is appropriate for the research question. I have a few suggestions on the presentation: 1. The description of the experiment lacks some information: a) the running length of the video; b) randomization of the conditions, if any; and c) The exact instruction that was given to the subject. 2. In \u201cData analysis\u201d section, \u201cplaced or removed\u201d was written as \u201cplaced OUR removed\u201d. 3. The Results section begins with a summary figure (Figure 2). This is inadequate. I\u2019d like to see how the subject actually performed in the detection task. A few examples of \u201craw\u201d data (time points of button press in relationship to the time points where the magician placed or removed the balls) should be provided. 4. The subjects were asked to detect two events (removal and placing of the ball) but Figure 2A collapsed them into one single measure (probability of correct report). The analysis should separate the two, because the two events have different significance. The stated goal of the experiment was to determine \u201cwhether the falling ball in Penn & Teller's Cups and Balls generated stronger misdirection, as hypothesized by Teller, than alternative manipulations.\u201d In this context, the \u201cloading\u201d of the cup (ie. placement of the ball) more directly reflects the effectiveness of the illusion than the removal of the ball. Recall that in the introduction, the magician was said to be surprised that he did not see the load, thus making the load the more important. Either that, or I have misunderstood the research question. If the separating Figure 2A into two turns out to be less than insightful, the authors should at least briefly report the results or make a comment on it.\nValidity of the findings\nThe research question is: how effective was the illusion? Figure 2 answers the question: the falling of the ball was not as effective as the magician suspected. However, the authors found a learning effect. The more the subjects viewed the magic, the less effective the illusion was. Given that, the interpretation of Figure 2A becomes complicated. It is still possible that the magician's manipulation was effective, but only for the first time. This question is not addressed in the manuscript. The manuscript therefore fails to give a compelling answer to the research question expressed in Introduction. This unsatisfactory situation is difficult to avoid given the subject matter. I don't think it makes the findings invalid but the authors should either try to see if it can be addressed in some way from the experimental data, express the research question differently, or acknowledge the problem in discussion. A second issue is with Figure 2C. The effect is acknowledged in the text but its significance is never commented upon. Since it is difficult to detect the load in the opaque cup condition, how can the no load condition affect the results? What does this mean?\nCite this review as\nYu H (2013) Peer Review #2 of \"Perceptual elements in Penn & Teller\u2019s \u201cCups and Balls\u201d magic trick (v0.1)\". PeerJ https://doi.org/10.7287/peerj.19v0.1/reviews/2",
+    "all_reviews": "Review 1: David Reser \u00b7 Jan 5, 2013 \u00b7 Academic Editor\nACCEPT\nCongratulations, and I appreciate the quick turnaround on the most recent version. I look forward to seeing the published document.\nReview 2: David Reser \u00b7 Jan 4, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nThank you for your thorough response to the points raised in the previous reviews. I believe this manuscript will be acceptable, pending amendment/explanation of the point raised by Reviewer 1 regarding the statistical values in the Results section of the amended manuscript.\nReview 3: Reviewer 1 \u00b7 Jan 4, 2013\nBasic reporting\nI am satisfied that the author's of this manuscript have addressed my earlier concerns.\nExperimental design\nI am satisfied that the author's of this manuscript have addressed my earlier concerns.\nValidity of the findings\nI am satisfied that the author's of this manuscript have addressed my earlier concerns.\n\nI want to point out that in the Results-> Perceptual Reports section, all of the chi-square and AUC values are identical despite different significance levels. I am not sure if this is intended, if not the author's should consider correcting the values prior to publication.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Perceptual elements in Penn & Teller\u2019s \u201cCups and Balls\u201d magic trick (v0.2)\". PeerJ https://doi.org/10.7287/peerj.19v0.2/reviews/1\nReview 4: David Reser \u00b7 Dec 18, 2012 \u00b7 Academic Editor\nMAJOR REVISIONS\nBoth reviewers requested substantive changes to the figures and methods description, and Reviewer 1 identifies several potential problem areas in the statistical analysis. In your reply letter, please explicitly identify the changes in the manuscript which address these concerns.\nReviewer 2 points out that the magician's own expectation of what elements were important for misdirection of the viewer was inconsistent with the data, which is briefly addressed in the Discussion. In my opinion, it would be worthwhile to expand this point in the discussion beyond the single paragraph in the original submitted manuscript, and clarify which perceptual or cognitive elements contribute to the expectations of the viewer, especially the naive viewer. Please note that I leave this suggestion to the discretion of the authors, and this specific point will not affect my final decision if the other issues called out by the reviewers are adequately addressed.\nPlease ensure that the revised manuscript conforms to the PeerJ policy regarding data and materials sharing:\nData and Materials Sharing\n1 PeerJ is committed to improving scholarly communications and as part of this commitment, all authors are responsible for making materials, data and associated protocols available to readers without delay. The preferred way to meet this requirement is to publicly deposit as noted below. Cases of non-compliance will be investigated by PeerJ, which reserves the right to act on the results of the investigation.\n...\n- Where suitable domain-specific repositories do not exist, authors may deposit in either Dryad or an institutional repository and provide the access information with the manuscript. Alternately, authors may choose to deposit non-standard data (including figures, posters, rich media) on Figshare or PeerJ PrePrints, for example. In all cases, the DOI reference (where applicable) should be provided in the article.\n- Any supporting data sets for which there are no suitable repositories must be made available as publishable Supplemental Information files by PeerJ.\n...\n\nThe full policy is viewable in Policy and Procedures page at: https://peerj.com/about/policies-and-procedures/#data-materials-sharing.\nReview 5: Reviewer 1 \u00b7 Dec 18, 2012\nBasic reporting\nNo comments.\nExperimental design\nA lot of the points I make with regard to \"Experimental Design\" have flow on effects to the \"Validity of the Findings\". I keep them together in one section for clarity.\n\nMethods & Results:\n\u2022 Please provide more detail as to the experimental equipment setup. What was the resolution of the stimulus (in pixels and in degrees of visual angle), What was the luminance, and contrast of the stimulus? Was the experiment conducted in a controlled environment (dark, quiet?).\n\u2022 Eye movements: You note that blink periods were removed. Were the subjects told to fixate on a particular part of the stimulus? If so then, how can you generalise your findings to a natural viewing of the magic trick. If not, then given that saccadic suppression is a very well known and studied phenomenon did you attempt to assess whether the key frames of the video that contained the manipulation occurred during a saccadic eye movement?\n\u2022 Given that one of your measures of interest was reaction time, were removals (button 1) and placings (button 2) equally easy to press?\n\u2022 How was the order of stimuli determined? If it was random was the same order used for each subject? This has profound implications for the validity of your learning findings.\n\u2022 Data analysis. Please provide more details about the results of your data analysis. For regression analyses please provide measures of goodness of fit.\n\u2022 Given that you are using the Newman-Kuels method to do pair-wise comparisons, please explain how you are controlling for the increasing experiment-wise alpha level given the number of independent variables you are testing.\n\u2022 You mention that your \u201cstatistical models determined main effects and first order interactions\u201d but you don\u2019t report these anywhere.\n\u2022 In several places throughout the paper you mention performance without qualifying which of the several measures of performance you tested you are referring to.\n\u2022 Given that you suggest that there was a learning effect how did you deal with this in the main analysis. Did you average all trials? Na\u00efve trials (such as would be most relevant to a person seeing a trick for the first time)?, learned trials?\nFigures\n\u2022 Please be consistent with your error-bars, you switch from 95% confidence intervals (figure 2) to SEM (Figures 3-4) and this can lead to a cursory reader assuming that later figures have more significance than they actually do.\n\u2022 Figure 2ABC) Please indicate where chance performance would be on these figures. Are the subjects worse than chance when tested with opaque cups?\n\u2022 Figure 3. Please indicate in the figure caption that dva refers to degrees of visual angle? or change the axis label to reflect this.\n\u2022 Figure4a. Please indicate what number the # of late findings is in relation to.\n\u2022 Please explain why you don't present the graph for opaque cups in Figure 4B&C.\n\u2022 I am concerned that given the findings concerning your other variables, comparisons of trial number to your dependent variables are potentially confounded by order effects. Without knowing how the conditions were ordered it is not possible to evaluate the data being presented in Figure 4.\nValidity of the findings\nSee my comments in the \"Experimental Design\" section.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Perceptual elements in Penn & Teller\u2019s \u201cCups and Balls\u201d magic trick (v0.1)\". PeerJ https://doi.org/10.7287/peerj.19v0.1/reviews/1\nReview 6: Hsin-Hao Yu \u00b7 Dec 11, 2012\nBasic reporting\nThe scientific part of the paper is well-written, but I found the non-scientific part, namely the description of the the performance of the magic, difficult to comprehend. Maybe I am simply not familiar with stage magic but it took me a long time to get a sense of what the magician actually did. The inclusion of a video clip will greatly improve the readability of the paper. If that is not possible due to copyright or technical issues, I think the introduction needs to be revised to give more detailed and precise description of the performance. Since I was not familiar with Cups and Balls, I turned to youtube, hoping to see how Penn and Teller perform the Cups and Balls magic. This probably was not such a good idea because Cups and Balls, as performed in Penn and Teller\u2019s various stage shows and television programs, appeared to be not exactly the same Cups and Balls magic studied in the paper (in the youtube clips, the magician does not tilt the transparent cup to make the ball fall into his hand). This is a source of confusion that can be avoided if the introduction provides more background information. The second paragraph of Introduction describes how the magic was invented. According to the description, the magician was surprised by how effective the illusion was. The falling of the ball distracted the attention of the magician such that he himself did not see the loading of the cup. However, Figure 2 clearly shows that the subjects could reliably detect the loading of the cup, thus showing that the magic was not magical, after all! How can a successful magician be so wrong about the effectiveness of his performance? I\u2019m puzzled. The description of the magician\u2019s invention of the magic, rather than helping to motivate the experiment, only confused this reviewer.\nExperimental design\nThe design of the experiment is appropriate for the research question. I have a few suggestions on the presentation: 1. The description of the experiment lacks some information: a) the running length of the video; b) randomization of the conditions, if any; and c) The exact instruction that was given to the subject. 2. In \u201cData analysis\u201d section, \u201cplaced or removed\u201d was written as \u201cplaced OUR removed\u201d. 3. The Results section begins with a summary figure (Figure 2). This is inadequate. I\u2019d like to see how the subject actually performed in the detection task. A few examples of \u201craw\u201d data (time points of button press in relationship to the time points where the magician placed or removed the balls) should be provided. 4. The subjects were asked to detect two events (removal and placing of the ball) but Figure 2A collapsed them into one single measure (probability of correct report). The analysis should separate the two, because the two events have different significance. The stated goal of the experiment was to determine \u201cwhether the falling ball in Penn & Teller's Cups and Balls generated stronger misdirection, as hypothesized by Teller, than alternative manipulations.\u201d In this context, the \u201cloading\u201d of the cup (ie. placement of the ball) more directly reflects the effectiveness of the illusion than the removal of the ball. Recall that in the introduction, the magician was said to be surprised that he did not see the load, thus making the load the more important. Either that, or I have misunderstood the research question. If the separating Figure 2A into two turns out to be less than insightful, the authors should at least briefly report the results or make a comment on it.\nValidity of the findings\nThe research question is: how effective was the illusion? Figure 2 answers the question: the falling of the ball was not as effective as the magician suspected. However, the authors found a learning effect. The more the subjects viewed the magic, the less effective the illusion was. Given that, the interpretation of Figure 2A becomes complicated. It is still possible that the magician's manipulation was effective, but only for the first time. This question is not addressed in the manuscript. The manuscript therefore fails to give a compelling answer to the research question expressed in Introduction. This unsatisfactory situation is difficult to avoid given the subject matter. I don't think it makes the findings invalid but the authors should either try to see if it can be addressed in some way from the experimental data, express the research question differently, or acknowledge the problem in discussion. A second issue is with Figure 2C. The effect is acknowledged in the text but its significance is never commented upon. Since it is difficult to detect the load in the opaque cup condition, how can the no load condition affect the results? What does this mean?\nCite this review as\nYu H (2013) Peer Review #2 of \"Perceptual elements in Penn & Teller\u2019s \u201cCups and Balls\u201d magic trick (v0.1)\". PeerJ https://doi.org/10.7287/peerj.19v0.1/reviews/2\n",
+    "LLM_extracted_review": "[\"Review 1: Congratulations, and I appreciate the quick turnaround on the most recent version. I look forward to seeing the published document.\\n\\nReview 2: Thank you for your thorough response to the points raised in the previous reviews. I believe this manuscript will be acceptable, pending amendment/explanation of the point raised by Reviewer 1 regarding the statistical values in the Results section of the amended manuscript.\\n\\nReview 3: I want to point out that in the Results-> Perceptual Reports section, all of the chi-square and AUC values are identical despite different significance levels. I am not sure if this is intended; if not, the authors should consider correcting the values prior to publication.\\n\\nReview 4: Both reviewers requested substantive changes to the figures and methods description, and Reviewer 1 identifies several potential problem areas in the statistical analysis. Reviewer 2 points out that the magician's own expectation of what elements were important for misdirection of the viewer was inconsistent with the data, which is briefly addressed in the Discussion. It would be worthwhile to expand this point in the discussion beyond the single paragraph in the original submitted manuscript. Please ensure that the revised manuscript conforms to the PeerJ policy regarding data and materials sharing.\\n\\nReview 5: Please provide more detail as to the experimental equipment setup. What was the resolution of the stimulus (in pixels and in degrees of visual angle), what was the luminance, and contrast of the stimulus? Was the experiment conducted in a controlled environment (dark, quiet)? Were the subjects told to fixate on a particular part of the stimulus? If so, how can you generalize your findings to a natural viewing of the magic trick? If not, did you attempt to assess whether the key frames of the video that contained the manipulation occurred during a saccadic eye movement? Given that one of your measures of interest was reaction time, were removals (button 1) and placings (button 2) equally easy to press? How was the order of stimuli determined? If it was random, was the same order used for each subject? This has profound implications for the validity of your learning findings. Please provide more details about the results of your data analysis. For regression analyses, please provide measures of goodness of fit. Please explain how you are controlling for the increasing experiment-wise alpha level given the number of independent variables you are testing. You mention that your \u201cstatistical models determined main effects and first order interactions\u201d but you don\u2019t report these anywhere. In several places throughout the paper, you mention performance without qualifying which of the several measures of performance you tested you are referring to. Given that you suggest that there was a learning effect, how did you deal with this in the main analysis? Did you average all trials? Na\u00efve trials? Please be consistent with your error-bars; you switch from 95% confidence intervals to SEM, which can lead to a cursory reader assuming that later figures have more significance than they actually do. Please indicate where chance performance would be on these figures. Are the subjects worse than chance when tested with opaque cups? Please indicate in the figure caption that dva refers to degrees of visual angle or change the axis label to reflect this. Please explain why you don't present the graph for opaque cups in Figure 4B&C. I am concerned that given the findings concerning your other variables, comparisons of trial number to your dependent variables are potentially confounded by order effects.\\n\\nReview 6: The scientific part of the paper is well-written, but I found the non-scientific part, namely the description of the performance of the magic, difficult to comprehend. The inclusion of a video clip will greatly improve the readability of the paper. If that is not possible, I think the introduction needs to be revised to give more detailed and precise description of the performance. The description of the magician\u2019s invention of the magic, rather than helping to motivate the experiment, only confused this reviewer. The description of the experiment lacks some information: a) the running length of the video; b) randomization of the conditions, if any; and c) the exact instruction that was given to the subject. The Results section begins with a summary figure (Figure 2). This is inadequate. I\u2019d like to see how the subject actually performed in the detection task. A few examples of \u201craw\u201d data should be provided. The analysis should separate the two events (removal and placing of the ball), because the two events have different significance. The manuscript fails to give a compelling answer to the research question expressed in Introduction. The interpretation of Figure 2A becomes complicated. The authors should either try to see if it can be addressed in some way from the experimental data, express the research question differently, or acknowledge the problem in discussion. Since it is difficult to detect the load in the opaque cup condition, how can the no load condition affect the results? What does this mean?\"]"
+}

peerj_json_files/PeerJ_Json_100.json ADDED Viewed

The diff for this file is too large to render. See raw diff

peerj_json_files/PeerJ_Json_101.json ADDED Viewed

The diff for this file is too large to render. See raw diff

peerj_json_files/PeerJ_Json_102.json ADDED Viewed

	@@ -0,0 +1,17 @@

+{
+    "v1_Abstract": "Background /Aim: Significant physiological changes occur during pregnancy and lactation. Intrahepatic cholestasis of pregnancy (ICP) is a liver disease closely related to disruption of bile acid homeostasis. The objective of this study was to examine the regulation of bile acid synthesis and transport in normal pregnant and lactating rats. Materials and Methods: Livers from timed pregnant SD rats were collected on gestational days (GD) 10, 14 and 19, and postnatal days (PND) 1,7,14 and 21. Total bile acids were determined by the enzymatic method, total RNA was isolated and subjected to real time RT-PCR analysis. Liver protein was extracted for western-blot analysis. Results: Under physiological conditions hepatic bile acids were not elevated during pregnancy but increased during lactation in rats. Bile acid synthesis rate-limiting enzyme Cyp7a1 was unchanged in gestations days, but increased on PND14 and21 at mRNA and protein levels. Expression of Cyp8b1, Cyp27a1 and Cyp7b1 was also higher during lactation. The mRNA levels of small heterodimer partner (SHP) and protein levels of farnesoid X receptor (FXR) were increased during pregnancy and lactation. Bile acid transporters Ntcp, Bsep, Mrp3 and Mrp4 were lower at gestation, but increased during lactation. Hepatic Oatp transporters were decreased during pregnancy and lactation. Conclusion: Hepatic bile acid homeostasis is maintained during normal pregnancy in rats, probably through the FXR-SHP regulation. The expression of bile acid synthesis genes and liver bile acid accumulation were increased during lactation, together with increased expression of bile acid efflux transporter Bsep, Mrp3 and Mrp4.",
+    "v2_Abstract": "Background /Aim: Significant physiological changes occur during pregnancy and lactation. Intrahepatic cholestasis of pregnancy (ICP) is a liver disease closely related to disruption of bile acid homeostasis. The objective of this study was to examine the regulation of bile acid synthesis and transport in normal pregnant and lactation rats. Materials and Methods: Livers from timely pregnant SD rats were collected on gestational days (GD) 10, 14 and 19, and postnatal days (PND) 1,7,14 and 21. Total bile acids were determined by the enzymatic method, total RNA was isolated and subjected to real time RT-PCR analysis. Liver protein was extracted for western-blot analysis. Results: Under physiological conditions hepatic bile acids were not elevated during pregnancy but increased during lactation in rats. Bile acid synthesis rate-limiting enzyme Cyp7a1 was unchanged in gestations days, but increased on PND14 and21 at mRNA and protein levels. Expression of Cyp8b1, Cyp27a1 and Cyp7b1 was also higher during lactation. The mRNA levels of small heterodimer partner (SHP) and protein levels of farnesoid X receptor (FXR) were increased during pregnancy and lactation. Bile acid transporters Ntcp, Bsep, Mrp3 and Mrp4 were lower at gestation, but increased during lactation. Hepatic Oatp transporters were decreased during pregnancy and lactation. Conclusion: Hepatic bile acid homeostasis is maintained during normal pregnancy in rats, probably through the FXR-SHP regulation. The expression of bile acid synthesis genes and liver bile acid accumulation were increased during lactation, together with increased expression of bile acid efflux transporter Bsep, Mrp3 and Mrp4.",
+    "v3_Abstract": "Background /Aim: Intrahepatic cholestasis of pregnancy (ICP) is a liver disease which may occur in the third trimester of pregnancy. The etiology and pathogenesis of ICP is thought to be closely related to bile acid metabolism. The objective of this study was to examine the regulation of bile acid metabolism during normal pregnant and lactation in rats. Materials and Methods: Livers from timely pregnant SD rats were collected on gestational days (GD) 10, 14 and 19, and postnatal days (PND) 1,7,14 and 21. Total bile acids were determined by the enzymatic method, total RNA was isolated and subjected to real time RT-PCR analysis. Liver protein was extracted for western-blot analysis. Results: Under physiological conditions hepatic bile acids were not elevated during pregnancy but increased during lactation in rats. Bile acid synthesis rate-limiting enzyme Cyp7a1 was unchanged in gestations days, but increased on PND14 and21 at mRNA and protein levels. Expression of Cyp8b1, Cyp27a1 and Cyp7b1 was also higher during lactation. The expression of small heterodimer partner (SHP) was increased at GD19 and lactation days, and farnesoid X receptor (FXR) increased on postpartum. Bile acid transporters Ntcp, Bsep, Mrp3 and Mrp4 were lower at gestation, but increased during lactation. Hepatic Oatp transporters were decreased during pregnancy and lactation. Conclusion: Hepatic bile acid homeostasis maintained during normal pregnancy in rats, probably through the regulation of SHP. The expression of bile acid synthesis genes and liver bile acids were increased during lactation, together with increased expression of bile acid transporters.",
+    "v1_text": "expression of nuclear receptors : PeerJ reviewing PDF | (v2013:06:607:2:0:NEW 3 Aug 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:607:2:0:NEW 3 Aug 2013) R ev ie w in g M an us cr ip t Figure 5 Fig. 5 fig : Liver bile acid concentrations PeerJ reviewing PDF | (v2013:06:607:2:0:NEW 3 Aug 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:607:2:0:NEW 3 Aug 2013) R ev ie w in g M an us cr ip t Figure 2 Fig. 2 Bile acid synthesis gene expression PeerJ reviewing PDF | (v2013:06:607:2:0:NEW 3 Aug 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:607:2:0:NEW 3 Aug 2013) R ev ie w in g M an us cr ip t Figure 3 Hepatic expression of bile acid synthesis rate-limiting protein CYP7A1 in pregnant and lactating rat. Western bolts were performed using liver homogenates from control, pregnant rats in GD 10, 14, 19 and PND 1, 7, 14 and 21. The expression of CYP7A1 was semi-quantified by band intensity.Values are mean \u00b1 SEM. Dark gray bars represent pregnant rat, and black bars represent lactating rat. Significant difference was confirmed by two-tailed independent Samples test method (P<0.05). PeerJ reviewing PDF | (v2013:06:607:2:0:NEW 3 Aug 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:607:2:0:NEW 3 Aug 2013) R ev ie w in g M an us cr ip t Figure 4 Fig. 4 fxr protein expression : PeerJ reviewing PDF | (v2013:06:607:2:0:NEW 3 Aug 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:607:2:0:NEW 3 Aug 2013) R ev ie w in g M an us cr ip t Figure 6 Fig. 6 efflux transporters : PeerJ reviewing PDF | (v2013:06:607:2:0:NEW 3 Aug 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:607:2:0:NEW 3 Aug 2013) R ev ie w in g M an us cr ip t Figure 7 Fig. 7 uptake transporter expression : PeerJ reviewing PDF | (v2013:06:607:2:0:NEW 3 Aug 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:607:2:0:NEW 3 Aug 2013) R ev ie w in g M an us cr ip t",
+    "v2_text": "expression of nuclear receptors : PeerJ reviewing PDF | (v2013:06:607:1:1:REVIEW 27 Jul 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:607:1:1:REVIEW 27 Jul 2013) R ev ie w in g M an us cr ip t Figure 5 Fig. 5 fig : Liver bile acid concentrations PeerJ reviewing PDF | (v2013:06:607:1:1:REVIEW 27 Jul 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:607:1:1:REVIEW 27 Jul 2013) R ev ie w in g M an us cr ip t Figure 2 Fig. 2 Bile acid synthesis gene expression PeerJ reviewing PDF | (v2013:06:607:1:1:REVIEW 27 Jul 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:607:1:1:REVIEW 27 Jul 2013) R ev ie w in g M an us cr ip t Figure 3 Hepatic expression of bile acid synthesis rate-limiting protein CYP7A1 in pregnant and lactating rat. Western bolts were performed using liver homogenates from control, pregnant rats in GD 10, 14, 19 and PND 1, 7, 14 and 21. The expression of CYP7A1 was semi-quantified by band intensity.Values are mean \u00b1 SEM. Dark gray bars represent pregnant rat, and black bars represent lactating rat. Significant difference was confirmed by two-tailed independent Samples test method (P<0.05). PeerJ reviewing PDF | (v2013:06:607:1:1:REVIEW 27 Jul 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:607:1:1:REVIEW 27 Jul 2013) R ev ie w in g M an us cr ip t Figure 4 Fig. 4 fxr protein expression : PeerJ reviewing PDF | (v2013:06:607:1:1:REVIEW 27 Jul 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:607:1:1:REVIEW 27 Jul 2013) R ev ie w in g M an us cr ip t Figure 6 Fig. 6 efflux transporters : PeerJ reviewing PDF | (v2013:06:607:1:1:REVIEW 27 Jul 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:607:1:1:REVIEW 27 Jul 2013) R ev ie w in g M an us cr ip t Figure 7 Fig. 7 uptake transporter expression : PeerJ reviewing PDF | (v2013:06:607:1:1:REVIEW 27 Jul 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:607:1:1:REVIEW 27 Jul 2013) R ev ie w in g M an us cr ip t",
+    "url": "https://peerj.com/articles/144/reviews/",
+    "review_1": "Santosh Patnaik \u00b7 Aug 6, 2013 \u00b7 Academic Editor\nACCEPT\nThe issues raised in the last round of review have bee adequately addressed.",
+    "review_2": "Santosh Patnaik \u00b7 Jul 19, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nThe manuscript that you submitted is prepared very well, with a clear, succinct and consistently styled presentation, appropriate discourse, details and references, and suitably composed figures.\n\nOf the three reviewers, only one has suggested a few changes, all of which are minor. The comment about mentioning the phenomenon of caspase-independent cytotoxic effect of docetaxel on prostate cancer cells is relatively important and should be addressed. Regarding the comments asking whether Coomassie stain was used on protein blots or gels, and whether the LNCaP and PC3 cells that were used in the study were parental or derived, please clarify appropriately in the manuscript as other readers of the publication may have the same questions. You can choose to address the remaining comments of the reviewer at your discretion.",
+    "review_3": "Reviewer 1 \u00b7 Jul 18, 2013\nBasic reporting\nThe majority of the manuscript is coherent, simplistic and intriguing.\nHowever, the 4th paragraph of the introduction could use a little more clarification and purpose. Lines 317-320 would fit well into this paragraph.\n\nIt needs to be mentioned in the introduction or Discussion section that docetaxel induced prostate cancer death involves concomintant activation of both caspase- and lysosome-dependent pathways (Mediavilla-Varela M. et al., Mol. Cancer. 2009)\nExperimental design\nIt is stated that ABT-737 can sensitize androgen-dependent LNCAP and PC-3 cells to docetaxel and 1198. Are both these cell lines docetaxel-resistant versions of the parental cells or are they docetaxel sensitive LNCAP and PC-3 cells?\n\nMaterials and Methods- ATCC recommends growing PC-3 cells in medium with 10% FBS.\n\nIf selection of a loading control is difficult, it is better to use Ponceaus staining of the membrane rather than Coomassie blue staining of the gel because it is the same membrane and is ideal for monitoring transfer. However, some researchers are using a new technique of protein staining with Coomassie that can be used on the membrane after immunodetection (Welinder and Ekblad, J. Proteome Res. 2011). Please clarify if you stained the gel or membrane.\nValidity of the findings\nNo comments\nAdditional comments\nI would suggest uniformity in the labeling of figures such as Figures 1A and 1B. I would choose between LN or L for LNCaP and DU or D for DU145. Also, I would not use PC in labeling for PC3 cells as PC has been consistently used in the rest of the paper as an acronym for prostate cancer. I would suggest identifying prostate cancer as PCa to avoid this confusion.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"ABT-737, a small molecule Bcl-2/Bcl-xL antagonist, increases antimitotic-mediated apoptosis in human prostate cancer cells (v0.1)\". PeerJ https://doi.org/10.7287/peerj.144v0.1/reviews/1",
+    "review_4": "Reviewer 2 \u00b7 Jul 16, 2013\nBasic reporting\nNo comments.\nExperimental design\nThis was actually very well done; the results of this manuscript appear to be the collaborative work of several different projects, which, as described in the methods section of this paper, appear to be technically sound.\nValidity of the findings\nThe results of the author's experiments is clearly conveyed in the context of what is known in the general literature. Their final conclusions appear to be supported by their findings in the results.\nAdditional comments\nThis was an interesting investigation into some of the biologic mechanisms for chemotherapeutic resistance in the setting of Castration resistant prostate cancer with well-designed experiments to quantify the relative importance of roles played by certain pro/anti-apoptotic proteins. It appears that the scientific methods discussed and conclusions derived from this set of studies were well-designed and appropriate, respectively.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"ABT-737, a small molecule Bcl-2/Bcl-xL antagonist, increases antimitotic-mediated apoptosis in human prostate cancer cells (v0.1)\". PeerJ https://doi.org/10.7287/peerj.144v0.1/reviews/2",
+    "pdf_1": "https://peerj.com/articles/144v0.2/submission",
+    "pdf_2": "https://peerj.com/articles/144v0.1/submission",
+    "review_5": "Reviewer 3 \u00b7 Jun 29, 2013\nBasic reporting\nThe article is well written and conforms to the PeerJ policies.\nExperimental design\nThe experimental design is appropriate for the authors to make their conclusions.\nValidity of the findings\nThe data is sound albeit not incredibly novel or surprising. I reviewed this manuscript for a different journal and that was the main conclusion of the reviews. Thus if impact is not a factor in the decision, I believe this is an acceptable paper.\nAdditional comments\nThis is an interesting finding. The experiments are well-performed, however the findings and conclusions are not terribly surprising. So the study is not incredibly novel, but it is solid and appropriate.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #3 of \"ABT-737, a small molecule Bcl-2/Bcl-xL antagonist, increases antimitotic-mediated apoptosis in human prostate cancer cells (v0.1)\". PeerJ https://doi.org/10.7287/peerj.144v0.1/reviews/3",
+    "all_reviews": "Review 1: Santosh Patnaik \u00b7 Aug 6, 2013 \u00b7 Academic Editor\nACCEPT\nThe issues raised in the last round of review have bee adequately addressed.\nReview 2: Santosh Patnaik \u00b7 Jul 19, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nThe manuscript that you submitted is prepared very well, with a clear, succinct and consistently styled presentation, appropriate discourse, details and references, and suitably composed figures.\n\nOf the three reviewers, only one has suggested a few changes, all of which are minor. The comment about mentioning the phenomenon of caspase-independent cytotoxic effect of docetaxel on prostate cancer cells is relatively important and should be addressed. Regarding the comments asking whether Coomassie stain was used on protein blots or gels, and whether the LNCaP and PC3 cells that were used in the study were parental or derived, please clarify appropriately in the manuscript as other readers of the publication may have the same questions. You can choose to address the remaining comments of the reviewer at your discretion.\nReview 3: Reviewer 1 \u00b7 Jul 18, 2013\nBasic reporting\nThe majority of the manuscript is coherent, simplistic and intriguing.\nHowever, the 4th paragraph of the introduction could use a little more clarification and purpose. Lines 317-320 would fit well into this paragraph.\n\nIt needs to be mentioned in the introduction or Discussion section that docetaxel induced prostate cancer death involves concomintant activation of both caspase- and lysosome-dependent pathways (Mediavilla-Varela M. et al., Mol. Cancer. 2009)\nExperimental design\nIt is stated that ABT-737 can sensitize androgen-dependent LNCAP and PC-3 cells to docetaxel and 1198. Are both these cell lines docetaxel-resistant versions of the parental cells or are they docetaxel sensitive LNCAP and PC-3 cells?\n\nMaterials and Methods- ATCC recommends growing PC-3 cells in medium with 10% FBS.\n\nIf selection of a loading control is difficult, it is better to use Ponceaus staining of the membrane rather than Coomassie blue staining of the gel because it is the same membrane and is ideal for monitoring transfer. However, some researchers are using a new technique of protein staining with Coomassie that can be used on the membrane after immunodetection (Welinder and Ekblad, J. Proteome Res. 2011). Please clarify if you stained the gel or membrane.\nValidity of the findings\nNo comments\nAdditional comments\nI would suggest uniformity in the labeling of figures such as Figures 1A and 1B. I would choose between LN or L for LNCaP and DU or D for DU145. Also, I would not use PC in labeling for PC3 cells as PC has been consistently used in the rest of the paper as an acronym for prostate cancer. I would suggest identifying prostate cancer as PCa to avoid this confusion.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"ABT-737, a small molecule Bcl-2/Bcl-xL antagonist, increases antimitotic-mediated apoptosis in human prostate cancer cells (v0.1)\". PeerJ https://doi.org/10.7287/peerj.144v0.1/reviews/1\nReview 4: Reviewer 2 \u00b7 Jul 16, 2013\nBasic reporting\nNo comments.\nExperimental design\nThis was actually very well done; the results of this manuscript appear to be the collaborative work of several different projects, which, as described in the methods section of this paper, appear to be technically sound.\nValidity of the findings\nThe results of the author's experiments is clearly conveyed in the context of what is known in the general literature. Their final conclusions appear to be supported by their findings in the results.\nAdditional comments\nThis was an interesting investigation into some of the biologic mechanisms for chemotherapeutic resistance in the setting of Castration resistant prostate cancer with well-designed experiments to quantify the relative importance of roles played by certain pro/anti-apoptotic proteins. It appears that the scientific methods discussed and conclusions derived from this set of studies were well-designed and appropriate, respectively.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"ABT-737, a small molecule Bcl-2/Bcl-xL antagonist, increases antimitotic-mediated apoptosis in human prostate cancer cells (v0.1)\". PeerJ https://doi.org/10.7287/peerj.144v0.1/reviews/2\nReview 5: Reviewer 3 \u00b7 Jun 29, 2013\nBasic reporting\nThe article is well written and conforms to the PeerJ policies.\nExperimental design\nThe experimental design is appropriate for the authors to make their conclusions.\nValidity of the findings\nThe data is sound albeit not incredibly novel or surprising. I reviewed this manuscript for a different journal and that was the main conclusion of the reviews. Thus if impact is not a factor in the decision, I believe this is an acceptable paper.\nAdditional comments\nThis is an interesting finding. The experiments are well-performed, however the findings and conclusions are not terribly surprising. So the study is not incredibly novel, but it is solid and appropriate.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #3 of \"ABT-737, a small molecule Bcl-2/Bcl-xL antagonist, increases antimitotic-mediated apoptosis in human prostate cancer cells (v0.1)\". PeerJ https://doi.org/10.7287/peerj.144v0.1/reviews/3\nReview 6: \n",
+    "LLM_extracted_review": "['Review 1: The issues raised in the last round of review have been adequately addressed.\\n\\nReview 2: The manuscript is prepared very well, with a clear, succinct and consistently styled presentation, appropriate discourse, details and references, and suitably composed figures. Only one reviewer suggested a few minor changes. The comment about mentioning the phenomenon of caspase-independent cytotoxic effect of docetaxel on prostate cancer cells is relatively important and should be addressed. Clarification is needed regarding whether Coomassie stain was used on protein blots or gels, and whether the LNCaP and PC3 cells used in the study were parental or derived.\\n\\nReview 3: The majority of the manuscript is coherent, simplistic and intriguing. However, the 4th paragraph of the introduction could use a little more clarification and purpose. It needs to be mentioned that docetaxel induced prostate cancer death involves concomitant activation of both caspase- and lysosome-dependent pathways. Clarification is needed on whether the LNCAP and PC-3 cells are docetaxel-resistant or sensitive. It is better to use Ponceaus staining of the membrane rather than Coomassie blue staining of the gel. Uniformity in the labeling of figures is suggested to avoid confusion.\\n\\nReview 4: The experimental design is very well done; the results appear to be the collaborative work of several different projects and are technically sound. The results are clearly conveyed in the context of existing literature, and the conclusions appear to be supported by the findings. The investigation into the biologic mechanisms for chemotherapeutic resistance is interesting, with well-designed experiments.\\n\\nReview 5: The article is well written and conforms to PeerJ policies. The experimental design is appropriate for the conclusions made. The data is sound but not incredibly novel or surprising. The study is solid and appropriate, though the findings and conclusions are not terribly surprising.']"
+}

peerj_json_files/PeerJ_Json_103.json ADDED Viewed

The diff for this file is too large to render. See raw diff

peerj_json_files/PeerJ_Json_104.json ADDED Viewed

	@@ -0,0 +1,19 @@

+{
+    "v1_Abstract": "Our previously presented method for high throughput computational screening of mutant activity (Hediger et al., [1]) is benchmarked against experimentally measured amidase activity for 22 mutants of Candida antarctica lipase B (CalB). Using an appropriate cutoff criterion for the computed barriers, the qualitative activity of 15 out of 22 mutants is correctly predicted. The method identifies four of the six most active mutants with \u22653-fold wild type activity and seven out of the eight least active mutants with \u22640.5-fold wild type activity. The method is further used to screen all sterically possible (386) double-, tripleand quadruple-mutants constructed from the most active single mutants. Based on the benchmark test at least 20 new promising mutants are identified.",
+    "v1_col_introduction": "introduction  : In industry, one frequently tries to modify an enzyme in order to enhance its functionality in a certain way [2\u20136]. From an application point of view, one of the most interesting questions is how to modify an enzyme such that its activity is enhanced compared to wild type or such that a new kind of activity is introduced into the enzyme [7,8]. It can therefore be of considerable relevance to have a method available which efficiently allows to a priori discriminate between promising candidates for experimental study and mutants which can be excluded from the study.\nPeerJ reviewing PDF | (v2013:07:643:1:0:NEW 6 Aug 2013)\nR ev ie w in g M an\n2 Numerous methods are currently being proposed and developed for the description of enzyme activities, the theoretical background of which ranges from phenomenological and bioinformatics based approaches [9\u201313] to quantum mechanics based ab initio descriptions [14\u201322]. However one can expect that methods which are highly demanding in terms of set-up efforts and computational time are less likely to be employed in industrial contexts where qualitative or semi-quantitative conclusions can be of sufficient use in the beginning and planning phase of a wet-lab study. Few approaches, while taking into account a number of approximations and limitations in accuracy, aim at being used in parallel or prior to experimental work [23,24] and are not designed to be used for high throughput fashion. Hediger et al. have recently published a computational method for high throughput computational screening of mutant activity [1] and in this paper we benchmark the method against experimentally measured amidase activity for mutants of Candida antarctica lipase B (CalB) and apply the method to identify additional promising mutants.",
+    "v2_Abstract": "Our previously presented method for high throughput computational screening of mutant activity (Hediger et al., [1]) is benchmarked against experimentally measured amidase activity for 22 mutants of Candida antarctica lipase B (CalB). Using an appropriate cutoff criterion for the computed barriers, the qualitative activity of 15 out of 22 mutants is correctly predicted. The method identifies four of the six most active mutants with \u22653-fold wild type activity and seven out of the eight least active mutants with \u22640.5-fold wild type activity. The method is further used to screen all sterically possible (386) double-, tripleand quadruple-mutants constructed from the most active single mutants. Based on the benchmark test at least 20 new promising mutants are identified.",
+    "v2_col_introduction": "introduction  : In industry, one frequently tries to modify an enzyme in order to enhance its functionality in a certain way [2\u20136]. From an application point of view, one of the most interesting questions is how to modify an enzyme such that its activity is enhanced compared to wild type or such that a new kind of activity is introduced into the enzyme [7,8]. It can therefore be of considerable relevance to have a method available which efficiently allows to a priori discriminate between promising candidates for experimental study and mutants which can be excluded from the study.\nPeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013)\nR ev ie w in g M an\n2 Numerous methods are currently being proposed and developed for the description of enzyme activities, the theoretical background of which ranges from phenomenological and bioinformatics based approaches [9\u201313] to quantum mechanics based ab initio descriptions [14\u201322]. However one can expect that methods which are highly demanding in terms of set-up efforts and computational time are less likely to be employed in industrial contexts where qualitative or semi-quantitative conclusions can be of sufficient use in the beginning and planning phase of a wet-lab study. Few approaches, while taking into account a number of approximations and limitations in accuracy, aim at being used in parallel or prior to experimental work [23,24] and are not designed to be used for high throughput fashion. Hediger et al. have recently published a computational method for high throughput computational screening of mutant activity [1] and in this paper we benchmark the method against experimentally measured amidase activity for mutants of Candida antarctica lipase B (CalB) and apply the method to identify additional promising mutants.",
+    "v1_text": "results and discussion : methods : We introduce the experimental set-up and the methodology for comparing experimental and computational data. We describe a benchmarking and a combinatorial study of CalB mutant activity. Experimentally, variants of Candida Antarctica lipase B (CalB) were either produced in Pichia pastoris with C-terminal His6-tag for subsequent affinity purification or expressed in Aspergillus oryzae without terminal tag followed by a three-step purification procedure. It is generally accepted that in serine protease like enzymes, the formation of the tetrahedral intermediate (TI, Fig. 1) is rate determining [25\u201328] and throughout this work we assume that a lower barrier for this reaction correlates to increased overall activity of the enzyme. The substrate used throughout this study is N-benzyl-2-chloroacetamide. The organisms used for expression of the individual variants are indicated in Table 1. conclusions : Our previously presented method for high throughput computational screening of mutant activity [1] is benchmarked against experimentally measured amidase activity for 22 mutants of Candida antarctica lipase B (CalB). PeerJ reviewing PDF | (v2013:07:643:1:0:NEW 6 Aug 2013) R ev ie w in g M an us cr ip t 10 Experimentally, amidase activity is successfully introduced in 12 mutants, the highest activity is determined to be 11.2 -fold over the wild type activity. Using an appropriate cutoff criterion for the computed barriers, the qualitative activity of 15 out of 22 mutants is correctly predicted. It is noted that the correlation is best for mutants with largest activity difference compared to wild type (both positive and negative). For example the method identifies four of the six most active mutants with \u22653-fold wild type activity. Similarly, the method identifies seven out of the eight least active mutants with \u22640.5-fold wild type activity. Thus validated, the computational method is used to screen all sterically possible (386) double-, tripleand quadrupole-mutants constructed from the most active single mutants. Based on the benchmark test at least 20 new promising mutants are identified. These mutants have so far not been tested experimentally and are thus offered as scientifically testable predictions. Interestingly, we observe that single mutants that are predicted to have low activity appear to have high activity in combination with other mutants. This is illustrated in specific analysis of effects of mutations of two different positions (104 and 189). tables : Table 1. Experimental overall activities and calculated reaction barriers of Set S. Activity factors +1/-1 indicate increased/decreased overall activity. Ao and Pp indicating expression in organisms (Org.) Aspergillus oryzae or Pichia pastoris, respectively. The cutoff to distinguish higher and lower activity mutants is defined as 12.5 kcal/mol, see text. Species Experimental Calculated Activity [*WT] ActivityFactor Barriers [kcal/mol] ActivityFactor Org. G39A-T103G-W104F-L278A 11.2 1 13.9 -1 Ao G39A-L278A 7.0 1 11.3 1 Pp G39A-W104F 4.2 1 10.6 1 Ao G39A-T103G-L278A 3.8 1 7.3 1 Ao G39A-W104F-L278A 3.6 1 11.8 1 Pp T103G 3.0 1 13.6 -1 Ao G39A-W104F-I189Y-L278A 2.9 1 10.9 1 Pp G39A 2.8 1 11.2 1 Ao L278A 2.5 1 12.8 -1 Ao W104F 2.0 1 12.0 1 Ao G39A-T103G-W104Q-L278A 1.9 1 12.8 -1 Ao G39A-T103G-W104F-D223G-L278A 1.5 1 11.3 1 Pp G39A-T103G 0.8 -1 7.5 1 Ao G39A-T42A-T103G-W104F-L278A 0.7 -1 10.4 1 Pp I189H 0.5 -1 12.9 -1 Pp G39A-I189G-L278A 0.4 -1 10.7 1 Pp G41S 0.3 -1 13.4 -1 Pp I189G 0.2 -1 18.9 -1 Pp G39A-T103G-W104F-I189H-D223G-L278A 0.1 -1 13.7 -1 Pp G39A-T103G-W104F-I189H-L278A-A282G-I285A-V286A 0.1 -1 12.9 -1 Pp A132N 0.0 -1 12.5 -1 Pp P38H 0.0 -1 12.5 -1 Pp WT 1.0 -- 7.5 -- PeerJ reviewing PDF | (v2013:07:643:1:0:NEW 6 Aug 2013) R ev ie w in g M an us cr ip t 20 Table 2. Point mutations. The term active site refers to residues with potential direct Van der Waals contact to the substrate. The term first shell/second shell refers to residues which are adjecent to an active site/first shell residue. Target Mutations Type Description P38 H Second shell (H neutral) G39 A First shell G41 S First shell T42 A Second shell T103 G First shell W104 F, Q, Y Active site A132 N First shell A141 N, Q Active site I189 A, G, H, N, Y Active site (G including additional water, H neutral) D223 G First shell (Increase of charge by +1) L278 A Active site A282 G Active site I285 A Active site V286 A First shell PeerJ reviewing PDF | (v2013:07:643:1:0:NEW 6 Aug 2013) R ev ie w in g M an us cr ip t 21 Table 3. Side chains used for generation of combinatorial set L. i and gi indicate the position in the back bone and the number of mutations at that position, respectively. Mutation i gi G39A 39 1 T103G 103 1 W104{F, Q, Y} 104 3 A141{N, Q} 141 2 I189{A, G, H, N, Y} 189 5 L278A 278 1 PeerJ reviewing PDF | (v2013:07:643:1:0:NEW 6 Aug 2013) R ev ie w in g M an us cr ip t 22 Table 4. Combinatorial study details. From the possible mutants, the combinations containing the pair A141N/Q-I189Y, the mutants with inconclusive barriers and the mutants with barriers >19.0 kcal/mol are subtracted to give the number of mutants in set L. \u201cOnly Set L\u201d indicates the number of mutants uniquely present in set L and not in set S. Order Possible Containing Inconclusive Barrier >19.0 Set L Only A141N/Q-I189Y barrier [kcal/mol] Set L Single 13 0 0 0 13 7 Double 64 2 4 8 50 47 Triple 154 12 21 20 101 98 Four-fold 193 24 36 19 114 111 Total 424 38 61 47 278 263 PeerJ reviewing PDF | (v2013:07:643:1:0:NEW 6 Aug 2013) R ev ie w in g M an us cr ip t 23 Table 5. Selection of mutants from set L with lowest barriers. Mutation Barrier[kcal/mol] G39A-T103G-I189Y 5.7 G39A-I189Y 6.2 G39A-A141Q-I189G-L278A 6.3 G39A-A141N-L278A 7.6 G39A-A141N 7.7 G39A-A141N-I189H-L278A 8.3 G39A-W104F-A141Q-I189A 8.3 G39A-A141Q-I189N 9.1 G39A-A141N-I189N 9.3 G39A-T103G-W104Y-A141N 9.8 G39A-W104Y-I189Y 9.8 G39A-A141N-I189N-L278A 10.1 G39A-W104F-A141N 10.1 G39A-I189H-L278A 10.2 G39A-A141N-I189A-L278A 10.2 W104Y-I189H 10.4 G39A-T103G-W104F-I189Y 10.4 G39A-A141Q-I189A-L278A 10.4 G39A-T103G-A141Q-I189H 10.4 G39A-T103G-I189A-L278A 10.5 PeerJ reviewing PDF | (v2013:07:643:1:0:NEW 6 Aug 2013) R ev ie w in g M an us cr ip t Figure 1 Reaction scheme for the formation of TI. Nucleophilic attack by O\u03b3 of S105 on carbonyl carbon C20 of substrate. R1: -CH2-Cl, R2: -CH2-C5H6. PeerJ reviewing PDF | (v2013:07:643:1:0:NEW 6 Aug 2013) R ev ie w in g M an us cr ip t Figure 2 Position of point mutations. A. Overlay of mutations W104[F, Q, Y]. B: Overlay of mutations A141[N, Q]. C: Overlay of mutations of I189[A, G, H, N, Y]. D: Mutations P38H, G39A, G41S, T42A, T103G, A132N, L278A, A282G, I285A, V286A. Substrate shown in magenta. PeerJ reviewing PDF | (v2013:07:643:1:0:NEW 6 Aug 2013) R ev ie w in g M an us cr ip t Figure 3 Comparison of experimental and computed activities. 1/-1 correspond to increased/decreased overall activity, respectively. Prediction rate is 15/22 (68%). PeerJ reviewing PDF | (v2013:07:643:1:0:NEW 6 Aug 2013) R ev ie w in g M an us cr ip t Figure 4 Barrier scatter plot of set S. 22 mutants; The cutoff value cS is discussed in the text. PeerJ reviewing PDF | (v2013:07:643:1:0:NEW 6 Aug 2013) R ev ie w in g M an us cr ip t Figure 5 Barrier scatter plots of set L. In both panels, the labels indicate mutants containing the labeled and possibly additional mutations up to the indicated order. \"OTHER\" indicates a mutant not containing any of the labeled mutations or of higher than 4. order. A: Mutations of W104. B: Mutations of I189. PeerJ reviewing PDF | (v2013:07:643:1:0:NEW 6 Aug 2013) R ev ie w in g M an us cr ip t computational details : The computational method used to estimate the reaction barriers of the CalB mutants has been described in detail earlier [1] and is only summarized here. As described previously [1], in order to make the method computationally feasible, relatively approximate treatments of the wave function, structural model, dynamics and reaction path are used. Given this and the automated setup of calculations, some inaccurate results will be unavoidable. However, the intent of the method is similar to experimental high throughput screens of enzyme activity where, for example, negative results may result from issues unrelated to intrinsic activity of the enzyme such as imperfections in the activity assay, low expression yield, protein aggregation, etc. Just like its experimental counterpart our technique is intended to identify potentially interesting mutants for further study. The reaction barriers are estimated computationally by preparing molecular model structures [1] (consisting of around 840 atoms) of the enzyme substrate complex (ES) and the tetrahedral intermediate (TI) inbetween which linear interpolation is carried out to generate structures of the enzyme on the reaction path. Such adiabatic mapping is the most common way to estimate barriers in QM/MM studies of enzymatic reaction mechanisms. The resulting barriers tend to be in good agreement with experiment, which indicates that this is a reasonable approximation (see for example [30] and [16]). The geometry of each interpolation frame is optimized while keeping the distance between the nucleophilic carbon C20 of the substrate and O\u03b3 of serine 105 (Fig. 1) fixed at a specific value di = dini \u2212 i(dini \u2212 dfin)/10, where dini and dfin are the distances between C 20 and O\u03b3 in the ES complex and TI, respectively (in A\u030a, 10 PeerJ reviewing PDF | (v2013:07:643:1:0:NEW 6 Aug 2013) R ev ie w in g M an us cr ip t 5 being the number of interpolation frames and i the interpolation frame index). In geometry optimization calculations, the gradient convergence criteria is set to 0.5 kcal/(molA\u030a) and a linear scaling implementation of the PM6 method (MOZYME [31]) together with a NDDO cutoff of 15 A\u030a is applied. The energy profile of the reaction barrier at the PM6 level of theory [32] is subsequently mapped out by carrying out conventional SCF calculations of each optimized interpolation frame. All calculations are carried out using the MOPAC suite of programs [33, 34]. The molecular models are based on the crystal structure of the CalB enzyme with PDB identifier 1LBS [35]. In order to prevent significant rearrangement of hydrogen bonding network of surface residues during the optimization, a number of additional structural constraints are applied in the geometry optimizations, i.e. the residues S50, P133, Q156, L277 and P280 are kept fixed. These (surface) residues are observed to rearrange and form new hydrogen bonds in optimizations when no constraints are applied. Omitting the constraints leads to unconclusive barrier shapes containing many irregular minima along the reaction coordinate which do not permit to readily define a reaction barrier. For the analysis, the reaction barrier is defined by the difference between the highest energy point on the reaction profile and the energy corresponding to the enzyme substrate complex. From our calculations (PM6//MOZYME in vacuum), we estimate the wild type (WT) barrier to be 7.5 kcal/mol. Experimentally, specific activity of hydrolysis is determined. Given first order kinetics, saturation of the enzyme with substrate (usual for industrial application) and fast binding and product release, the catalytic rate constant kcat is directly proportional to the specific activity under the assumption that the amount of active enzyme remains constant. This therefore allows the catalytic rate constant kcat and, hence, the barrier height to be compared to the improvement factors reported in the results section. The approximations used here in relating the barrier height on the potential energy surface to kcat have been discussed previously [1]. It is noted that using one CPU per interpolation frame on the reaction barrier, the complete barrier of one mutant can be computed with 10 CPUs usually within less than 12 hours of wall clock time (for a molecular model of the size used in this study). Given a set of molecular models of the enzyme, and 100 available CPUs, it is possible to screen around 1000 mutants within one week. generation of calb variants without his-tags : Variants of CalB carrying the CalB signal peptide were generated at the DNA level using QuickChange mutagenesis on the corresponding gene residing in a dual E.coli/Aspergillus Pichia pastoris expression vector. The PCR was performed with proofreading DNA polymerase (New England Biolabs, NEB). To PeerJ reviewing PDF | (v2013:07:643:1:0:NEW 6 Aug 2013) R ev ie w in g M an us cr ip t 3 remove parent templates, they were methylated in vitro prior to PCR with CpG methyltransferase (from NEB) and digested in vivo after transformation of competent E.coli DH5 \u03b1 cells (TaKaRa) according to the instructions from the manufacturer. Plasmid DNA was isolated from transformed E.coli strains, and sequenced to verify the presence of the desired substitutions. Confirmed plasmid variants were used to transform an Aspergillus oryzae strain that is negative in pyrG (orotidine-5\u2019-phosphate decarboxylase), proteases pepC (aserine protease homologous to yscB), alp (an alkaline protease) NpI (a neutral metalloprotease I) to avoid degradation of the lipase variants during and after fermentation. The transformed Aspergillus strains were fermented as submerged culture in shake flasks and the lipase variants secreted into the fermentation medium. After the fermentation, the lipase variants were purified from the sterile filtered fermentation medium in a 3 step procedure with 1) hydrophobic interaction chromatography on decylamine-agarose, 2) buffer exchange by gel filtration and 3) ion exchange chromatography with cation exchange on SP-sepharose at pH 4.5. The lipase variant solutions were stored frozen. Variants of CalB carrying the CalB signal peptide and C-terminal His-tags were generated at the DNA level using SOE-PCR and inserted into a dual E.coli/Pichia pastoris expression vector using In-fusion cloning (ClonTech). The SOE-PCR was performed with Phusion DNA polymerase (NEB) and template DNA of the CalB gene. The cloned plasmids were transformed in competent E.coli DH5 \u03b1 cells (TaKaRa). Plasmid DNA was isolated from transformed E.coli strains, and sequenced to verify the presence of the desired substitutions. Confirmed plasmid variants were used to transform a Pichia pastoris strain that is Mut(s), Suc(+), His(-). The transformed Pichia strains were fermented as submerged culture in deep well plates and secretion of the lipase variants into the fermentation medium was induced by addition of methanol. After the fermentation, the lipase variants were purified from the cleared supernants using a standard His-tag purification protocol (Qiagen) and buffer-exhanged into 50 mM phosphate buffer, pH 7.0, using Amicon Ultra centrifugal filter devices with a 10 kDa cutoff (Merck Millipore). activity measurement : Amidase activity of CalB variants was determined in a two-step fluorimetric assay previously described by Henke et al [29]. First, enzymatic hydrolysis of N-benzyl-2-chloroacetamide was performed in 96-well microtiter plates in 200 \u00b5L phosphate-buffered aqueous solution pH 7.0 including 10% organic co-solvent PeerJ reviewing PDF | (v2013:07:643:1:0:NEW 6 Aug 2013) R ev ie w in g M an us cr ip t 4 (THF or DMSO). Reactions containing 5 mM amide substrate, 0.3-3 \u00b5M enzyme, and 12 \u00b5g/mL BSA were incubated for 18-20 h at 37\u25e6C in a shaker incubator. In a second step, 50 \u00b5L of a 20 mM 4- nitro-7-chloro-benzo-2-oxa-1,3-diazole (NBD-Cl) solution in 1-hexanol was added and the reaction of NBD-Cl with benzylamine formed during amide hydrolysis proceeded under identical reaction conditions for another hour. Fluorescence of the final reaction product was determined with excitation at 485 nm and measured emission at 538 nm. Calibration of the amide hydrolysis reaction was performed on each assay plate with benzylamine covering a concentration range between 0.05 and 5 mM. All enzymatic activities were corrected for non-enzymatic background reaction determined under identical conditions without enzyme present. combination mutants : The molecular model of the enzyme and the positions of the point mutations in the enzyme are illustrated in Fig. 2. The point mutations are listed in Table 2. Two sets of mutants are introduced in this section. PeerJ reviewing PDF | (v2013:07:643:1:0:NEW 6 Aug 2013) R ev ie w in g M an us cr ip t 6 A benchmarking set S and a combinatorial set L, the definitions of which are provided in the following. The point mutations are selected based on different design principles. These are either introduction of structural rearrangements in the active site to change the binding site properties of the active site (residues P38, G39, G41, T42, T103) [2], introduction of space to accomodate the substrate (W104, L278, A282, I285, V286), introduction of dipolar interactions between the enzyme and the substrate (A132, A141, I189) [36] or reduction of polarity in the active site (D223). Of course different heuristic considerations will apply for other enzymes when selecting the single mutations for combinatorial study. The mutants of the benchmarking study are collected in a small set S (22 mutants, Table 1). For the combinatorial study, out of the above we select six residues (G39, T103, W104, A141, I189, L278) which, it is assumed, contribute strongest to increased activity and define the mutations at each position as listed in Table 3. Given the position i and the number of mutations at each position gi, in general the upper limit for the number of mutants M in a combinatorial study can be calculated by writing a sum term for each type (i.e. \u201corder\u201d) of combination mutant, i.e. single, double, . . . , such that M = \u2211 i gi \ufe38 \ufe37\ufe37 \ufe38 Single (o=1) + \u2211 i,j j>i gi \u00b7 gj \ufe38 \ufe37\ufe37 \ufe38 Double (o=2) + \u2211 i,j,k k>j>i gi \u00b7 gj \u00b7 gk \ufe38 \ufe37\ufe37 \ufe38 Triple (o=3) + . . . (1) where each sum term consists of ( N o ) individual terms (N and o being the number of positions which can be mutated and the order of the mutant, respectively). By this scheme, considering the mutations listed in Table 3, hypothetically 424 (= 13 + 64 + 154 + 193) single to four-fold mutants can be constructed. This number is reduced by applying the restriction that out of the 424 hypothetically possible mutants, 0 single, 2 double, 12 triple and 24 four-fold combination mutants including the pair A141N/Q-I189Y are discarded because in the molecular modeling, these side chains could not be allocated spatially in the same mutant. We further note that 15 out of these remaining 386 mutants (Table 3) are present also in the benchmarking set S and thus the combinatorial study consists of 371 unique mutants. A detailed documentation of the number of screened residues in the combinatorial study is provided in Table 4. Prior to analysis, the reaction barriers of the combination mutants are inspected visually and mutants with irregularly shaped barriers, i.e. consisting of multiple peaks of similar height along the reaction coordinate, are discarded. This step is done simply because the calculations yield inconclusive results, PeerJ reviewing PDF | (v2013:07:643:1:0:NEW 6 Aug 2013) R ev ie w in g M an us cr ip t 7 so the most conservative choice is to consider it a non-promising candidate for a more active variant. Generating plots of the profiles is completely automated and visual inspection can easily be done for hundreds of mutants. Furthermore, out of the mutants with regular reaction barrier shapes, we discard those mutants with barriers >19.0 kcal/mol (i.e. the largest calculated barrier from set S). Following these selection criteria, 61 mutants are discarded because of inconclusive barrier shapes and 47 mutants because the barrier is higher than 19 kcal/mol (a distribution of reaction barriers is shown in Fig. S1). After these filtering steps, 278 mutants remain in the combinatorial study which we collect in the large set L (out of which 15 are in set S). An overview on the distribution of reaction barriers for the mutants from set L is provided Fig. S2 of the supporting information. We note that in set S, all barriers appear regular in shape and no mutant contains the A141N/Q and I189Y pair. set s: calibration of the accuracy : The correspondance of the computed barriers from set S with the experimental assay is shown in Fig. 3. The exact data is reported in Table 1. A scatterplot of calculated reaction barriers is presented in Fig. 4. We note that in set S, the highest experimentally observed activity is around 11 times the wild type activity (G39A-T103G-W104F-L278A, Table 1), while roughly ten mutants show no increased activity. In total, six mutants show 3-fold or higher wild type activity. In the calculations, only one mutant is observed to have a lower barrier than the wild type (7.3 kcal/mol, G39A-T103G-L278A) and the highest observed barrier is 18.9 kcal/mol (I189G). Given the approximations introduced to make the method sufficiently efficient, it is noted that the intent of the method is not a quantitative ranking of the reaction barriers, but to identify promising mutants for, and to eliminate non-promising mutants from, experimental consideration. Therefore only qualitative changes in overall activity are considered, which are represented by the activity factors (+1/-1). We categorize the experimentally observed activities and the predicted reaction barriers as follows. From experiment, a mutant with activity of 1.2 (0.8) times the wild type activity or higher (lower) is considered PeerJ reviewing PDF | (v2013:07:643:1:0:NEW 6 Aug 2013) R ev ie w in g M an us cr ip t 8 as improving (degrading). Correspondingly, the computed difference in reaction barrier height between a mutant and the wild type is expressed in qualitative terms. For the comparison with the experimental activity assay, we define a barrier cutoff cS = 12.5 kcal/mol to distinguish between potentially improving and degrading mutants in set S. The value of 12.5 kcal/mol is chosen such as to maximize the agreement with experiment, which is 68%, i.e. using a smaller or larger value for the cutoff will decrease this value. A mutant with a predicted barrier \u2265cS (12.5 kcal/mol) is considered to likely have decreased activity compared to the wild type while mutants with reaction barriers <cS are considered likely having increased activity. We note that defining the cutoff is done purely for a post hoc comparison of experimental and computed data. When using the computed barriers to identify promising experimental mutants, one simply chooses the N mutants with the lowest barriers, where N is the number of mutants affordable to do experimentally (e.g. 20 in the discussion of set L). Based on this approach, qualitative activity of 15 out of 22 mutants is correctly predicted. It is noted that the correlation is best for mutants with largest activity difference compared to wild type (both positive or negative). For example the method identifies four of the six most active mutants with \u22653-fold wild type activity. Similarly, the method identifies seven out of the eight least active mutants with \u22640.5-fold wild type activity. For mutants with only small differences in activity compared to wild type, the predictions are less accurate. Set L: Large scale screening study Set L is screened to identify new mutants for which increased activity is predicted. The 20 mutants with the lowest barriers are suggested as candidates for further experimental study in Table 5. The distributions of reaction barriers, resolved by mutations at positions 104 and 189, are shown in Figs. 5A and B. In set L, three new mutants are identified with barriers lower than the predicted wild type barrier. Out of the 20 mutants suggested in Table 5, three are double mutants, seven are three-fold and ten are four-fold mutants. No single mutants where found for which increased activity compared to wild type is predicted. All mutants except one contain the G39A mutation, five contain the T103G mutation, six contain a mutation of W104, 13 contain a mutation of A141, 16 contain a mutation of I189 and eight contain the L278A mutation. From this observation it is likely that mutations of G39, A141 and I189 will likely contribute to an increased activity of the mutant and should thus be included in future experimental activity assays. PeerJ reviewing PDF | (v2013:07:643:1:0:NEW 6 Aug 2013) R ev ie w in g M an us cr ip t 9 Set L is further analysed in terms of the effect of the mutations at the positions 104 and 189. For the mutations of W104, we note that single mutations which give rise to relatively high barriers (W104Q, W104Y, Fig. 5A) can have significantly lower barriers in combination with other mutations. For example, out of the sixty mutants with lowest barriers (Fig. S3), 33 contain a mutation of W104 out of which 17 are suggested to be W104F, while 14 are suggested to be W104Y (two contain W104Q). The mutation of I189 is analysed in a similar way. In set L, five different mutations of this residue are screened (Table 3). The single mutant with the lowest barrier is I189Y and the two mutants with the lowest predicted barrier contain this mutation as well (Table 5). Similarly to above, higher order mutants containing I189A, I189G, I189H or I189N are predicted to have considerably lower barriers than the corresponding single mutants, Fig. 5B. Particularly, out of the mutants listed in Table 5, three contain the I189A, one contains I189G mutation, four contain the I189H mutation and three contain the I189N mutation. As a special case we highlight that the single mutant I189G has one of the highest calculated barriers (18.9 kcal/mol, Table 1), however the four-fold mutant G39A-A141Q-I189G-L278A has one of the lowest barriers (6.3 kcal/mol, Table 5). Interestingly, the mutant G39A-A141Q-L278A has an intermediate barrier (10.9 kcal/mol). It would appear that I189G as a single mutant is counterproductive (high computed barrier) but lowers the barrier of G39A-A141Q-L278A. This observation is further supported by the observation that the I189G mutation is in spatial proximity to A141Q. While it is difficult to quantify the interaction, it is likely that in the mutant, the rather large side chain of A141Q is better accommodated in the active site and can better interact with the substrate. Observations as these should be kept in mind when selecting the single mutants to be considered when preparing higher order mutants.",
+    "v2_text": "results and discussion : figure legends : ES TS TI PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t 15 ! \" # $ !\"#$% &'(% )*+'& ,*$(-%./&./0./1./23 %*4*-1./53 6*+4-7./5./23 8'$0 9\"\"'& &4*: )4\"% %\"$\"& ,\"$;% <\"$=% %*'\"1 Figure 2. Positions of point mutations. A: Overlay of mutations W104[F, Q, Y]. B: Overlay of mutations of A141[N, Q]. C: Overlay of mutations of I189[A, G, H, N, Y]. D: Mutations P38H, G39A, G41S, T42A, T103G, A132N, L278A, A282G, I285A and V286A. Substrate shown in magenta. PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t 16 PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t 17 1 2 3 4 5 6 7 8 5 10 15 20 M ut at io n O rd er (1 -8 ) Reaction Barrier [kcal/mol] (calculated) PM6//MOZYME cS = 12.5 kcal/mol Set S Figure 4. Barrier scatter plot of set S. 22 mutants; The cutoff value cS is discussed in the text. PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t 18 PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t 19 methods : We introduce the experimental set-up and the methodology for comparing experimental and computational data. We describe a benchmarking and a combinatorial study of CalB mutant activity. Experimentally, variants of Candida Antarctica lipase B (CalB) were either produced in Pichia pastoris with C-terminal His6-tag for subsequent affinity purification or expressed in Aspergillus oryzae without terminal tag followed by a three-step purification procedure. It is generally accepted that in serine protease like enzymes, the formation of the tetrahedral intermediate (TI, Fig. 1) is rate determining [25\u201328] and throughout this work we assume that a lower barrier for this reaction correlates to increased overall activity of the enzyme. The substrate used throughout this study is N-benzyl-2-chloroacetamide. The organisms used for expression of the individual variants are indicated in Table 1. 2 novozymes a/s, krogshoejvej 36, dk-2880 bagsv\u00e6rd, denmark : \u2217 Corresponding Author, Email: jhjensen@chem.ku.dk conclusions : Our previously presented method for high throughput computational screening of mutant activity [1] is benchmarked against experimentally measured amidase activity for 22 mutants of Candida antarctica lipase B (CalB). Experimentally, amidase activity is successfully introduced in 12 mutants, the highest activity is determined to be 11.2 -fold over the wild type activity. Using an appropriate cutoff criterion for the computed barriers, the qualitative activity of 15 out of 22 mutants is correctly predicted. It is noted that the correlation is best for mutants with largest activity PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t 10 difference compared to wild type (both positive and negative). For example the method identifies four of the six most active mutants with \u22653-fold wild type activity. Similarly, the method identifies seven out of the eight least active mutants with \u22640.5-fold wild type activity. Thus validated, the computational method is used to screen all sterically possible (386) double-, tripleand quadrupole-mutants constructed from the most active single mutants. Based on the benchmark test at least 20 new promising mutants are identified. These mutants have so far not been tested experimentally and are thus offered as scientifically testable predictions. Interestingly, we observe that single mutants that are predicted to have low activity appear to have high activity in combination with other mutants. This is illustrated in specific analysis of effects of mutations of two different positions (104 and 189). 1 department of chemistry, university of copenhagen, universitetsparken 5, dk-2100 : tables : Table 1. Experimental overall activities and calculated reaction barriers of Set S. Category (Cat.) +1/-1 indicates increased/decreased overall activity. Category definition is discussed in text. Ao and Pp indicating expression in organisms (Org.) Aspergillus oryzae or Pichia pastoris, respectively. Species Exp. Activity [*WT] Cat. Calc. Barriers [kcal/mol] Cat. Org. G39A-T103G-W104F-L278A 11.2 1 13.9 -1 Ao G39A-L278A 7.0 1 11.3 1 Pp G39A-W104F 4.2 1 10.6 1 Ao G39A-T103G-L278A 3.8 1 7.3 1 Ao G39A-W104F-L278A 3.6 1 11.8 1 Pp T103G 3.0 1 13.6 -1 Ao G39A-W104F-I189Y-L278A 2.9 1 10.9 1 Pp G39A 2.8 1 11.2 1 Ao L278A 2.5 1 12.8 -1 Ao W104F 2.0 1 12.0 1 Ao G39A-T103G-W104Q-L278A 1.9 1 12.8 -1 Ao G39A-T103G-W104F-D223G-L278A 1.5 1 11.3 1 Pp G39A-T103G 0.8 -1 7.5 1 Ao G39A-T42A-T103G-W104F-L278A 0.7 -1 10.4 1 Pp I189H 0.5 -1 12.9 -1 Pp G39A-I189G-L278A 0.4 -1 10.7 1 Pp G41S 0.3 -1 13.4 -1 Pp I189G 0.2 -1 18.9 -1 Pp G39A-T103G-W104F-I189H-D223G-L278A 0.1 -1 13.7 -1 Pp G39A-T103G-W104F-I189H-L278A-A282G-I285A-V286A 0.1 -1 12.9 -1 Pp A132N 0.0 -1 12.5 -1 Pp P38H 0.0 -1 12.5 -1 Pp WT 1.0 -- 7.5 -- PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t 20 Table 2. Point mutations. The term active site refers to residues with potential direct Van der Waals contact to the substrate. The term first shell/second shell refers to residues which are adjecent to an active site/first shell residue. Target Mutations Type Description P38 H Second shell (H neutral) G39 A First shell G41 S First shell T42 A Second shell T103 G First shell W104 F, Q, Y Active site A132 N First shell A141 N, Q Active site I189 A, G, H, N, Y Active site (G including additional water, H neutral) D223 G First shell (Increase of charge by +1) L278 A Active site A282 G Active site I285 A Active site V286 A First shell PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t 21 Table 3. Side chains used for generation of combinatorial set L. i and gi indicate the position in the back bone and the number of mutations at that position, respectively. Mutation i gi G39A 39 1 T103G 103 1 W104{F, Q, Y} 104 3 A141{N, Q} 141 2 I189{A, G, H, N, Y} 189 5 L278A 278 1 PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t 22 Table 4. Combinatorial study details. From the possible mutants, the combinations containing the pair A141N/Q-I189Y, the mutants with inconclusive barriers and the mutants with barriers >19.0 kcal/mol are subtracted to give the number of mutants in set L. \u201cOnly Set L\u201d indicates the number of mutants uniquely present in set L and not in set S. Order Possible Containing Inconclusive Barrier >19.0 Set L Only A141N/Q-I189Y barrier [kcal/mol] Set L Single 13 0 0 0 13 7 Double 64 2 4 8 50 47 Triple 154 12 21 20 101 98 Four-fold 193 24 36 19 114 111 Total 424 38 61 47 278 263 PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t 23 Table 5. Selection of mutants from set L with lowest barriers. Mutation Barrier[kcal/mol] G39A-T103G-I189Y 5.7 G39A-I189Y 6.2 G39A-A141Q-I189G-L278A 6.3 G39A-A141N-L278A 7.6 G39A-A141N 7.7 G39A-A141N-I189H-L278A 8.3 G39A-W104F-A141Q-I189A 8.3 G39A-A141Q-I189N 9.1 G39A-A141N-I189N 9.3 G39A-T103G-W104Y-A141N 9.8 G39A-W104Y-I189Y 9.8 G39A-A141N-I189N-L278A 10.1 G39A-W104F-A141N 10.1 G39A-I189H-L278A 10.2 G39A-A141N-I189A-L278A 10.2 W104Y-I189H 10.4 G39A-T103G-W104F-I189Y 10.4 G39A-A141Q-I189A-L278A 10.4 G39A-T103G-A141Q-I189H 10.4 G39A-T103G-I189A-L278A 10.5 PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t ESI-1 computational details : The computational method used to estimate the reaction barriers of the CalB mutants has been described in detail earlier [1] and is only summarized here. As described previously [1], in order to make the method computationally feasible, relatively approximate treatments of the wave function, structural model, dynamics and reaction path are used. Given this and the automated setup of calculations, some inaccurate results will be unavoidable. However, the intent of the method is similar to experimental high throughput screens of enzyme activity where, for example, negative results may result from issues unrelated to intrinsic activity of the enzyme such as imperfections in the activity assay, low expression yield, protein aggregation, etc. Just like its experimental counterpart our technique is intended to identify potentially interesting mutants for further study. The reaction barriers are estimated computationally by preparing molecular model structures [1] (consisting of around 840 atoms) of the enzyme substrate complex (ES) and the tetrahedral intermediate (TI) inbetween which linear interpolation is carried out to generate structures of the enzyme on the reaction path. Such adiabatic mapping is the most common way to estimate barriers in QM/MM studies of enzymatic reaction mechanisms. The resulting barriers tend to be in good agreement with experiment, which indicates that this is a reasonable approximation (see for example [30] and [16]). The geometry of each interpolation frame is optimized while keeping the distance between the nucleophilic carbon C20 of the substrate and O\u03b3 of serine 105 (Fig. 1) fixed at a specific value di = dini \u2212 i(dini \u2212 dfin)/10, where dini and dfin are the distances between C 20 and O\u03b3 in the ES complex and TI, respectively (in A\u030a, 10 PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t 5 being the number of interpolation frames and i the interpolation frame index). In geometry optimization calculations, the gradient convergence criteria is set to 0.5 kcal/(molA\u030a) and a linear scaling implementation of the PM6 method (MOZYME [31]) together with a NDDO cutoff of 15 A\u030a is applied. The energy profile of the reaction barrier at the PM6 level of theory [32] is subsequently mapped out by carrying out conventional SCF calculations of each optimized interpolation frame. All calculations are carried out using the MOPAC suite of programs [33, 34]. The molecular models are based on the crystal structure of the CalB enzyme with PDB identifier 1LBS [35]. In order to prevent significant rearrangement of hydrogen bonding network of surface residues during the optimization, a number of additional structural constraints are applied in the geometry optimizations, i.e. the residues S50, P133, Q156, L277 and P280 are kept fixed. These (surface) residues are observed to rearrange and form new hydrogen bonds in optimizations when no constraints are applied. Omitting the constraints leads to unconclusive barrier shapes containing many irregular minima along the reaction coordinate which do not permit to readily define a reaction barrier. For the analysis, the reaction barrier is defined by the difference between the highest energy point on the reaction profile and the energy corresponding to the enzyme substrate complex. From our calculations (PM6//MOZYME in vacuum), we estimate the wild type (WT) barrier to be 7.5 kcal/mol. Experimentally, specific activity of hydrolysis is determined. Given first order kinetics, saturation of the enzyme with substrate (usual for industrial application) and fast binding and product release, the catalytic rate constant kcat is directly proportional to the specific activity under the assumption that the amount of active enzyme remains constant. This therefore allows the catalytic rate constant kcat and, hence, the barrier height to be compared to the improvement factors reported in the results section. The approximations used here in relating the barrier height on the potential energy surface to kcat have been discussed previously [1]. It is noted that using one CPU per interpolation frame on the reaction barrier, the complete barrier of one mutant can be computed with 10 CPUs usually within less than 12 hours of wall clock time (for a molecular model of the size used in this study). Given a set of molecular models of the enzyme, and 100 available CPUs, it is possible to screen around 1000 mutants within one week. generation of calb variants without his-tags : Variants of CalB carrying the CalB signal peptide were generated at the DNA level using QuickChange mutagenesis on the corresponding gene residing in a dual E.coli/Aspergillus Pichia pastoris expression vector. The PCR was performed with proofreading DNA polymerase (New England Biolabs, NEB). To PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t 3 remove parent templates, they were methylated in vitro prior to PCR with CpG methyltransferase (from NEB) and digested in vivo after transformation of competent E.coli DH5 \u03b1 cells (TaKaRa) according to the instructions from the manufacturer. Plasmid DNA was isolated from transformed E.coli strains, and sequenced to verify the presence of the desired substitutions. Confirmed plasmid variants were used to transform an Aspergillus oryzae strain that is negative in pyrG (orotidine-5\u2019-phosphate decarboxylase), proteases pepC (aserine protease homologous to yscB), alp (an alkaline protease) NpI (a neutral metalloprotease I) to avoid degradation of the lipase variants during and after fermentation. The transformed Aspergillus strains were fermented as submerged culture in shake flasks and the lipase variants secreted into the fermentation medium. After the fermentation, the lipase variants were purified from the sterile filtered fermentation medium in a 3 step procedure with 1) hydrophobic interaction chromatography on decylamine-agarose, 2) buffer exchange by gel filtration and 3) ion exchange chromatography with cation exchange on SP-sepharose at pH 4.5. The lipase variant solutions were stored frozen. Variants of CalB carrying the CalB signal peptide and C-terminal His-tags were generated at the DNA level using SOE-PCR and inserted into a dual E.coli/Pichia pastoris expression vector using In-fusion cloning (ClonTech). The SOE-PCR was performed with Phusion DNA polymerase (NEB) and template DNA of the CalB gene. The cloned plasmids were transformed in competent E.coli DH5 \u03b1 cells (TaKaRa). Plasmid DNA was isolated from transformed E.coli strains, and sequenced to verify the presence of the desired substitutions. Confirmed plasmid variants were used to transform a Pichia pastoris strain that is Mut(s), Suc(+), His(-). The transformed Pichia strains were fermented as submerged culture in deep well plates and secretion of the lipase variants into the fermentation medium was induced by addition of methanol. After the fermentation, the lipase variants were purified from the cleared supernants using a standard His-tag purification protocol (Qiagen) and buffer-exhanged into 50 mM phosphate buffer, pH 7.0, using Amicon Ultra centrifugal filter devices with a 10 kDa cutoff (Merck Millipore). activity measurement : Amidase activity of CalB variants was determined in a two-step fluorimetric assay previously described by Henke et al [29]. First, enzymatic hydrolysis of N-benzyl-2-chloroacetamide was performed in 96-well microtiter plates in 200 \u00b5L phosphate-buffered aqueous solution pH 7.0 including 10% organic co-solvent PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t 4 (THF or DMSO). Reactions containing 5 mM amide substrate, 0.3-3 \u00b5M enzyme, and 12 \u00b5g/mL BSA were incubated for 18-20 h at 37\u25e6C in a shaker incubator. In a second step, 50 \u00b5L of a 20 mM 4- nitro-7-chloro-benzo-2-oxa-1,3-diazole (NBD-Cl) solution in 1-hexanol was added and the reaction of NBD-Cl with benzylamine formed during amide hydrolysis proceeded under identical reaction conditions for another hour. Fluorescence of the final reaction product was determined with excitation at 485 nm and measured emission at 538 nm. Calibration of the amide hydrolysis reaction was performed on each assay plate with benzylamine covering a concentration range between 0.05 and 5 mM. All enzymatic activities were corrected for non-enzymatic background reaction determined under identical conditions without enzyme present. combination mutants : The molecular model of the enzyme and the positions of the point mutations in the enzyme are illustrated in Fig. 2. The point mutations are listed in Table 2. Two sets of mutants are introduced in this section. PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t 6 A benchmarking set S and a combinatorial set L, the definitions of which are provided in the following. The point mutations are selected based on different design principles. These are either introduction of structural rearrangements in the active site to change the binding site properties of the active site (residues P38, G39, G41, T42, T103) [2], introduction of space to accomodate the substrate (W104, L278, A282, I285, V286), introduction of dipolar interactions between the enzyme and the substrate (A132, A141, I189) [36] or reduction of polarity in the active site (D223). Of course different heuristic considerations will apply for other enzymes when selecting the single mutations for combinatorial study. The mutants of the benchmarking study are collected in a small set S (22 mutants, Table 1). For the combinatorial study, out of the above we select six residues (G39, T103, W104, A141, I189, L278) which, it is assumed, contribute strongest to increased activity and define the mutations at each position as listed in Table 3. Given the position i and the number of mutations at each position gi, in general the upper limit for the number of mutants M in a combinatorial study can be calculated by writing a sum term for each type (i.e. \u201corder\u201d) of combination mutant, i.e. single, double, . . . , such that M = \u2211 i gi \ufe38 \ufe37\ufe37 \ufe38 Single (o=1) + \u2211 i,j j>i gi \u00b7 gj \ufe38 \ufe37\ufe37 \ufe38 Double (o=2) + \u2211 i,j,k k>j>i gi \u00b7 gj \u00b7 gk \ufe38 \ufe37\ufe37 \ufe38 Triple (o=3) + . . . (1) where each sum term consists of ( N o ) individual terms (N and o being the number of positions which can be mutated and the order of the mutant, respectively). By this scheme, considering the mutations listed in Table 3, hypothetically 424 (= 13 + 64 + 154 + 193) single to four-fold mutants can be constructed. This number is reduced by applying the restriction that out of the 424 hypothetically possible mutants, 0 single, 2 double, 12 triple and 24 four-fold combination mutants including the pair A141N/Q-I189Y are discarded because in the molecular modeling, these side chains could not be allocated spatially in the same mutant. We further note that 15 out of these remaining 386 mutants (Table 3) are present also in the benchmarking set S and thus the combinatorial study consists of 371 unique mutants. A detailed documentation of the number of screened residues in the combinatorial study is provided in Table 4. Prior to analysis, the reaction barriers of the combination mutants are inspected visually and mutants with irregularly shaped barriers, i.e. consisting of multiple peaks of similar height along the reaction coordinate, are discarded. This step is done simply because the calculations yield inconclusive results, PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t 7 so the most conservative choice is to consider it a non-promising candidate for a more active variant. Generating plots of the profiles is completely automated and visual inspection can easily be done for hundreds of mutants. Furthermore, out of the mutants with regular reaction barrier shapes, we discard those mutants with barriers >19.0 kcal/mol (i.e. the largest calculated barrier from set S). Following these selection criteria, 61 mutants are discarded because of inconclusive barrier shapes and 47 mutants because the barrier is higher than 19 kcal/mol (a distribution of reaction barriers is shown in Fig. S1). After these filtering steps, 278 mutants remain in the combinatorial study which we collect in the large set L (out of which 15 are in set S). An overview on the distribution of reaction barriers for the mutants from set L is provided Fig. S2 of the supporting information. We note that in set S, all barriers appear regular in shape and no mutant contains the A141N/Q and I189Y pair. set s : The correspondance of the computed barriers from set S with the experimental assay is shown in Fig. 3. The exact data is reported in Table 1. A scatterplot of calculated reaction barriers is presented in Fig. 4. We note that in set S, the highest experimentally observed activity is around 11 times the wild type activity (G39A-T103G-W104F-L278A, Table 1), while roughly ten mutants show no increased activity. In total, six mutants show 3-fold or higher wild type activity. In the calculations, only one mutant is observed to have a lower barrier than the wild type (7.3 kcal/mol, G39A-T103G-L278A) and the highest observed barrier is 18.9 kcal/mol (I189G). Given the approximations introduced to make the method sufficiently efficient, it is noted that the intent of the method is not a quantitative ranking of the reaction barriers, but to identify promising mutants for, and to eliminate non-promising mutants from, experimental consideration. Therefore only qualitative changes in overall activity are considered. We categorize the experimentally observed activities and the predicted reaction barriers as follows. From experiment, a mutant with activity of 1.2 (0.8) times the wild type activity or higher (lower) is considered PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t 8 as improving (degrading). Correspondingly, the computed difference in reaction barrier height between a mutant and the wild type is expressed in qualitative terms. For the comparison with the experimental activity assay, we define a barrier cutoff cS = 12.5 kcal/mol to distinguish between potentially improving and degrading mutants in set S. A mutant with a predicted barrier \u2265cS (12.5 kcal/mol) is considered to likely have decreased activity compared to the wild type while mutants with reaction barriers <cS are considered likely having increased activity. We note that defining the cutoff is done purely for a post hoc comparison of experimental and computed data. When using the computed barriers to identify promising experimental mutants, one simply chooses the N mutants with the lowest barriers, where N is the number of mutants affordable to do experimentally (e.g. 20 in the discussion of set L). Based on this approach, qualitative activity of 15 out of 22 mutants is correctly predicted. It is noted that the correlation is best for mutants with largest activity difference compared to wild type (both positive or negative). For example the method identifies four of the six most active mutants with \u22653-fold wild type activity. Similarly, the method identifies seven out of the eight least active mutants with \u22640.5-fold wild type activity. For mutants with only small differences in activity compared to wild type, the predictions are less accurate. Set L is screened to identify new mutants for which increased activity is predicted. The 20 mutants with the lowest barriers are suggested as candidates for further experimental study in Table 5. The distributions of reaction barriers, resolved by mutations at positions 104 and 189, are shown in Figs. 5A and B. In set L, three new mutants are identified with barriers lower than the predicted wild type barrier. Out of the 20 mutants suggested in Table 5, three are double mutants, seven are three-fold and ten are four-fold mutants. No single mutants where found for which increased activity compared to wild type is predicted. All mutants except one contain the G39A mutation, five contain the T103G mutation, six contain a mutation of W104, 13 contain a mutation of A141, 16 contain a mutation of I189 and eight contain the L278A mutation. From this observation it is likely that mutations of G39, A141 and I189 will likely contribute to an increased activity of the mutant and should thus be included in future experimental activity assays. Set L is further analysed in terms of the effect of the mutations at the positions 104 and 189. For the PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t 9 mutations of W104, we note that single mutations which give rise to relatively high barriers (W104Q, W104Y, Fig. 5A) can have significantly lower barriers in combination with other mutations. For example, out of the sixty mutants with lowest barriers (Fig. S3), 33 contain a mutation of W104 out of which 17 are suggested to be W104F, while 14 are suggested to be W104Y (two contain W104Q). The mutation of I189 is analysed in a similar way. In set L, five different mutations of this residue are screened (Table 3). The single mutant with the lowest barrier is I189Y and the two mutants with the lowest predicted barrier contain this mutation as well (Table 5). Similarly to above, higher order mutants containing I189A, I189G, I189H or I189N are predicted to have considerably lower barriers than the corresponding single mutants, Fig. 5B. Particularly, out of the mutants listed in Table 5, three contain the I189A, one contains I189G mutation, four contain the I189H mutation and three contain the I189N mutation. As a special case we highlight that the single mutant I189G has one of the highest calculated barriers (18.9 kcal/mol, Table 1), however the four-fold mutant G39A-A141Q-I189G-L278A has one of the lowest barriers (6.3 kcal/mol, Table 5). Interestingly, the mutant G39A-A141Q-L278A has an intermediate barrier (10.9 kcal/mol). It would appear that I189G as a single mutant is counterproductive (high computed barrier) but lowers the barrier of G39A-A141Q-L278A. Observations as these should be kept in mind when selecting the single mutants to be considered when preparing higher order mutants. electronic supplementary information : In silico screening of 393 mutants facilitates enzyme engineering of amidase activity in CalB Martin R. Hediger1, Luca De Vico1, Julie B. Rannes2, Christian Ja\u0308ckel2, Werner Besenmatter2, Allan Svendsen2, Jan H. Jensen1\u2217 copenhagen, denmark : s1 barrier scatter plots : ! \" # $ % ! ! ! \" ! # ! $ ! % & ! & \" & # & $ & % \" ! \" \" \" # \" $ \" % ' ! ' \" ' # ' $ ( # !( !# &( &# \"( )*+,-./0/1234-54+ 234-5416) 7 - .. *0 .1 89 , - :; < = :> Figure S1. Discarded mutants with barriers > 19.0 kcal/mol. Single mutants: 0; Double mutants: 8; Triple mutants: 20; Four-fold mutants: 19; Total: 47. PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t ESI-2 ! \" ! # $ $ $ % & ' ( & # ) # * ' ( * ! * \" \" # % $ % % ! ) ' ! ! & ! $ ) ! $ * ! & ( ! ( ! ! ( \" ! # # ! ' $ ! ' % ! * ' ! \" & ! % ) ! % * $ ) ( $ ! ! $ ! \" $ $ # $ & $ $ & % $ ( ' $ # & $ ' ) $ ' * $ * ( ) $ ( ' \" !) !$ !( !' !\" +,--./-012310/4151674,840 974,841:; < 8 / -= > 1? @ A , BC 6 2 BD Figure S2. Combination mutants from set L. Single: 13; Double mutants: 50; Triple mutants: 101; Four-fold mutants: 114: Total: 278. PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t ESI-3 ! \" # $ % & ' ( ) !* !! !\" !# !$ !% !& !' !( !) \"* \"! \"\" \"# \"$ \"% \"& \"' \"( \") #* #! #\" ## #$ #% #& #' #( #) $* $! $\" $# $$ $% $& $' $( $) %* %! %\" %# %$ %% %& %' %( %) &* * \" $ & ( !* !\" !$ +,-./-0123-4156270-18.993790 :!*$; :!*$< :!*$= >-479 +,-./-1?@ A . 99 37 91 BC D . 5E F 6 5G Figure S3. Low-barrier mutants. 33 out of 60 contain a mutation of W104 (W104F: 17, W104Y: 14, W104Q: 2). PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t Figure 1 Reaction scheme for the formation of TI. Nucleophilic attack by O\u03b3 of S105 on carbonyl carbon C20 of substrate. R1: -CH2-Cl, R2: -CH2-C5H6. PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t Figure 2 Position of point mutations. A. Overlay of mutations W104[F, Q, Y]. B: Overlay of mutations A141[N, Q]. C: Overlay of mutations of I189[A, G, H, N, Y]. D: Mutations P38H, G39A, G41S, T42A, T103G, A132N, L278A, A282G, I285A, V286A. Substrate shown in magenta. PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t Figure 3 Comparison of experimental and computed activities. 1/-1 correspond to increased/decreased overall activity, respectively. Prediction rate is 15/22 (68%). PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t Figure 4 Barrier scatter plot of set S. 22 mutants; The cutoff value cS is discussed in the text. PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t Figure 5 Barrier scatter plots of set L. In both panels, the labels indicate mutants containing the labeled and possibly additional mutations up to the indicated order. \"OTHER\" indicates a mutant not containing any of the labeled mutations or of higher than 4. order. A: Mutations of W104. B: Mutations of I189. PeerJ reviewing PDF | (v2013:07:643:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t",
+    "url": "https://peerj.com/articles/147/reviews/",
+    "review_1": "Fred Muench \u00b7 Aug 6, 2013 \u00b7 Academic Editor\nACCEPT\nThank you for making the changes to the introduction - the paper will be useful to help understand this research area. The paper can be accepted but please make the following changes:\n1. Please copy edit the entire manuscript. There were several typos\n2. Line 24: At present little is known about.... I would start the paragraph with - line 27 - We wished to conduct a systematic review -- first. Then At present... (Switch order). That will help transition the reader - then leading to why you did this preliminary paper.\n\nCongratulations.",
+    "review_2": "Fred Muench \u00b7 Mar 19, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nThank you for making the edits to the document. It enhances the manuscript. My main concern at this point is the introduction. It is now extremely short and the paragraphs do not necessarily flow after items were deleted without modifying the text - and new text was not included. I apologize if this was not clear in the previous edit but is important for the best possible manuscript. I believe the introduction should at minimum:\n\n1- have an additional paragraph on how health videos are being used presently across health disciplines and the information/intervention they are attempting to convey to consumers. For example, it is post-op care instructions, general health advice, etc. Please give examples.Even though this is a description of reviews this will help orient the reader who may not be experts.\n\n2- Your main question is the following: \"In reviews and systematic reviews of health-related YouTube videos, what are common reporting practices, exemplary methods and methodological challenges.\"\n\nAfter the above inserted paragraph, please write an additional paragraph on the problem or challenge you are trying to address with your paper -- what discrepancies exist, why conduct this review aside from the prevalence of these videos for example. More information on the importance of this review is also important to orient the reader.\n\nIn general the background should be clearer and set the reader up for your methods.\n\nBased on all your research, this should be a minor revision. Please contact me with questions. Fred",
+    "review_3": "Fred Muench \u00b7 Jan 17, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nAs with the two reviewers, I believe this article has merit and should be accepted with revisions. Understanding how to evaluate consumer health videos is an important area of study which will only grow . Please be sure to copy edit the manuscript as there are several errors. I have three broad based suggestions (some corresponding to reviewers) that may improve the manuscript for the reader and several smaller ones. First, the introduction appears to focus more of the background of how you came to the decision to study health videos by focusing on infant pain which is never mentioned again. It would be helpful to discuss in more detail about who is creating these videos, more on why this is an important area of study and most importantly to bring it back to your conclusion and goals \u2013 what are the current accepted systematic review methods (lines 316) and why we need to modify these current methods for videos and create a new method. In the method section, please merge how you searched into one paragraph (e.g. add 61-63 into the paragraph). Please define \u201csnowball techniques\u201d in more detail. Line 138 appears to be missing some text. The review provided so much important information as presented in the table and was a well done descriptive study. As with reviewer two, it would be nice to see you expand the discussion and conclusions based on your knowledge and findings to include \u2013 what are the most important areas that researchers cover when conducting reviews of Web-based health videos beyond the descriptive table of what you covered, why it is important and include a sample rubric, and with your knowledge, it would be very interesting to include a brief paragraph on what individuals developing videos should take into account - though this is clearly a separate paper and just a suggestion to enhance the paper for a larger audience. Looking forward to seeing the revised manuscript.",
+    "review_4": "Reviewer 1 \u00b7 Jan 16, 2013\nBasic reporting\nWhile I am unfamiliar with the journal, the research question seems well-identified, contained, and sized.\nExperimental design\nThe research question is clear. This study is strictly descriptive, which is fine for addressing the authors' research question. Methods were well-described.\nValidity of the findings\nIt seems likely that these findings are valid, inasmuch as the methods for the present review are well-described.\nAdditional comments\nThis paper is a useful description of \u201creviews of reviews\u201d of YouTube videos related to health. It specifies many dimensions of review reporting, and in some cases makes recommendations (e.g., for stopping rules). As information sources grow in channel types, attempts to summarize it will need to develop standards for reporting, and papers such as this one will help to make this happen.\nIt is interesting that some institutions required IRB review for the studies and some did not; were there differences between the two categories that account for this?\nOne issue that puzzles me a bit has to do with \u201csnowballing\u201d to include videos that appear as recommendations alongside the \u201cindex\u201d video. It seems that using this procedure would result in non-replicable results (if replicability is even a relevant issue here), and would interfere with systematicity in the search. A description of how this strategy could be made more systematic would be helpful.\nI would request that the authors attend carefully to their use of the terms \u201creview\u201d and \u201cstudy\u201d. Sometimes reviews are called studies, and this is confusing: for example, in line 260 where the study that is referred to is the pain-during-vaccination one that initiated this project.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"A systematic review of methods for studying consumer health YouTube videos, with implications for systematic reviews (v0.1)\". PeerJ https://doi.org/10.7287/peerj.147v0.1/reviews/1",
+    "pdf_1": "https://peerj.com/articles/147v0.3/submission",
+    "pdf_2": "https://peerj.com/articles/147v0.2/submission",
+    "review_5": "Reviewer 2 \u00b7 Dec 29, 2012\nBasic reporting\nNo comment\nExperimental design\nNo comment\nValidity of the findings\nNo comment\nAdditional comments\nThis article addresses an important topic. Although the authors write in their conclusion that formal reporting guidelines might be premature, it might be helpful to provide a flexible sample rubric or an adaptable list of elements to include and/or report. This sample list could be designed in the spirit of the Dublin Core which enables archivists to classify items according to a flexible list of accepted search terms, which can be adapted to fit the specifics of various projects by content and context.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"A systematic review of methods for studying consumer health YouTube videos, with implications for systematic reviews (v0.1)\". PeerJ https://doi.org/10.7287/peerj.147v0.1/reviews/2",
+    "pdf_3": "https://peerj.com/articles/147v0.1/submission",
+    "all_reviews": "Review 1: Fred Muench \u00b7 Aug 6, 2013 \u00b7 Academic Editor\nACCEPT\nThank you for making the changes to the introduction - the paper will be useful to help understand this research area. The paper can be accepted but please make the following changes:\n1. Please copy edit the entire manuscript. There were several typos\n2. Line 24: At present little is known about.... I would start the paragraph with - line 27 - We wished to conduct a systematic review -- first. Then At present... (Switch order). That will help transition the reader - then leading to why you did this preliminary paper.\n\nCongratulations.\nReview 2: Fred Muench \u00b7 Mar 19, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nThank you for making the edits to the document. It enhances the manuscript. My main concern at this point is the introduction. It is now extremely short and the paragraphs do not necessarily flow after items were deleted without modifying the text - and new text was not included. I apologize if this was not clear in the previous edit but is important for the best possible manuscript. I believe the introduction should at minimum:\n\n1- have an additional paragraph on how health videos are being used presently across health disciplines and the information/intervention they are attempting to convey to consumers. For example, it is post-op care instructions, general health advice, etc. Please give examples.Even though this is a description of reviews this will help orient the reader who may not be experts.\n\n2- Your main question is the following: \"In reviews and systematic reviews of health-related YouTube videos, what are common reporting practices, exemplary methods and methodological challenges.\"\n\nAfter the above inserted paragraph, please write an additional paragraph on the problem or challenge you are trying to address with your paper -- what discrepancies exist, why conduct this review aside from the prevalence of these videos for example. More information on the importance of this review is also important to orient the reader.\n\nIn general the background should be clearer and set the reader up for your methods.\n\nBased on all your research, this should be a minor revision. Please contact me with questions. Fred\nReview 3: Fred Muench \u00b7 Jan 17, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nAs with the two reviewers, I believe this article has merit and should be accepted with revisions. Understanding how to evaluate consumer health videos is an important area of study which will only grow . Please be sure to copy edit the manuscript as there are several errors. I have three broad based suggestions (some corresponding to reviewers) that may improve the manuscript for the reader and several smaller ones. First, the introduction appears to focus more of the background of how you came to the decision to study health videos by focusing on infant pain which is never mentioned again. It would be helpful to discuss in more detail about who is creating these videos, more on why this is an important area of study and most importantly to bring it back to your conclusion and goals \u2013 what are the current accepted systematic review methods (lines 316) and why we need to modify these current methods for videos and create a new method. In the method section, please merge how you searched into one paragraph (e.g. add 61-63 into the paragraph). Please define \u201csnowball techniques\u201d in more detail. Line 138 appears to be missing some text. The review provided so much important information as presented in the table and was a well done descriptive study. As with reviewer two, it would be nice to see you expand the discussion and conclusions based on your knowledge and findings to include \u2013 what are the most important areas that researchers cover when conducting reviews of Web-based health videos beyond the descriptive table of what you covered, why it is important and include a sample rubric, and with your knowledge, it would be very interesting to include a brief paragraph on what individuals developing videos should take into account - though this is clearly a separate paper and just a suggestion to enhance the paper for a larger audience. Looking forward to seeing the revised manuscript.\nReview 4: Reviewer 1 \u00b7 Jan 16, 2013\nBasic reporting\nWhile I am unfamiliar with the journal, the research question seems well-identified, contained, and sized.\nExperimental design\nThe research question is clear. This study is strictly descriptive, which is fine for addressing the authors' research question. Methods were well-described.\nValidity of the findings\nIt seems likely that these findings are valid, inasmuch as the methods for the present review are well-described.\nAdditional comments\nThis paper is a useful description of \u201creviews of reviews\u201d of YouTube videos related to health. It specifies many dimensions of review reporting, and in some cases makes recommendations (e.g., for stopping rules). As information sources grow in channel types, attempts to summarize it will need to develop standards for reporting, and papers such as this one will help to make this happen.\nIt is interesting that some institutions required IRB review for the studies and some did not; were there differences between the two categories that account for this?\nOne issue that puzzles me a bit has to do with \u201csnowballing\u201d to include videos that appear as recommendations alongside the \u201cindex\u201d video. It seems that using this procedure would result in non-replicable results (if replicability is even a relevant issue here), and would interfere with systematicity in the search. A description of how this strategy could be made more systematic would be helpful.\nI would request that the authors attend carefully to their use of the terms \u201creview\u201d and \u201cstudy\u201d. Sometimes reviews are called studies, and this is confusing: for example, in line 260 where the study that is referred to is the pain-during-vaccination one that initiated this project.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"A systematic review of methods for studying consumer health YouTube videos, with implications for systematic reviews (v0.1)\". PeerJ https://doi.org/10.7287/peerj.147v0.1/reviews/1\nReview 5: Reviewer 2 \u00b7 Dec 29, 2012\nBasic reporting\nNo comment\nExperimental design\nNo comment\nValidity of the findings\nNo comment\nAdditional comments\nThis article addresses an important topic. Although the authors write in their conclusion that formal reporting guidelines might be premature, it might be helpful to provide a flexible sample rubric or an adaptable list of elements to include and/or report. This sample list could be designed in the spirit of the Dublin Core which enables archivists to classify items according to a flexible list of accepted search terms, which can be adapted to fit the specifics of various projects by content and context.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"A systematic review of methods for studying consumer health YouTube videos, with implications for systematic reviews (v0.1)\". PeerJ https://doi.org/10.7287/peerj.147v0.1/reviews/2\nReview 6: \n",
+    "LLM_extracted_review": "['Review 1: The paper will be useful to help understand this research area. Please copy edit the entire manuscript. There were several typos. The introduction should start with \"We wished to conduct a systematic review\" to help transition the reader.\\n\\nReview 2: The introduction is now extremely short and lacks flow after deletions. An additional paragraph on how health videos are currently used across health disciplines is needed. The main question should be clearly stated, and a paragraph on the problem or challenge being addressed should be included to orient the reader.\\n\\nReview 3: The introduction focuses too much on infant pain, which is not mentioned again. More detail on who creates these videos and why this area is important is needed. The methods section should merge how the search was conducted into one paragraph, and \"snowball techniques\" should be defined in more detail. The discussion should expand on important areas for researchers and include suggestions for video developers.\\n\\nReview 4: The research question is well-identified and the methods are well-described. The paper is a useful description of reviews of YouTube videos related to health. There are concerns about the \"snowballing\" method used for including videos, as it may lead to non-replicable results. The use of the terms \"review\" and \"study\" should be clarified.\\n\\nReview 5: Although formal reporting guidelines might be premature, providing a flexible sample rubric or adaptable list of elements to include could be helpful. This could be designed similarly to the Dublin Core for classification.']"
+}

peerj_json_files/PeerJ_Json_105.json ADDED Viewed

The diff for this file is too large to render. See raw diff

peerj_json_files/PeerJ_Json_106.json ADDED Viewed

The diff for this file is too large to render. See raw diff

peerj_json_files/PeerJ_Json_107.json ADDED Viewed

	@@ -0,0 +1,17 @@

+{
+    "v1_Abstract": "NA-glycine is an endogenous lipid molecule with analgesic properties, which is structurally similar to the endocannabinoids 2-AG and anandamide but does not interact with cannabinoid receptors. NA-glycine has been suggested to act at the G-protein coupled receptors GPR18 and GPR92. Recently, we have described that NA-glycine can also modulate recombinant \u03b11\u03b22\u03b32 GABAA receptors. Here we characterize in more detail this modulation and investigate the relationship of its binding site with that of the endocannabinoid 2-AG.",
+    "v1_col_introduction": "introduction  : GABA is the major inhibitory neurotransmitter in mammalian brain. Its fast effects are mediated by synaptic and extrasynaptic GABAA receptors. These receptors are composed of five subunits that surround a central chloride ion channel (Macdonald & Olsen, 1994; Sieghart, 1995; Sieghart\n& Sperk, 2002; Sigel & Steinmann, 2012). The major receptor isoform consists of \u03b11, \u03b22, and \u03b32 subunits (Olsen & Sieghart, 2008). Different approaches led to the widely accepted 2\u03b1:2\u03b2:1\u03b3 subunit stoichiometry (Chang et al., 1996; Tretter et al., 1997; Farrar et al., 1999; Baumann et al.,\n2001, Baumann et al., 2002; Baur et al., 2006) with a subunit arrangement \u03b3\u03b2\u03b1\u03b2\u03b1 anti-clockwise as seen from the synaptic cleft (Baumann et al., 2001, Baumann et al., 2002; Baur et al., 2006). The pharmacological properties are dependent on subunit composition (Sigel et al., 1990) and arrangement (Minier and Sigel, 2004).\nNeurosteroids (Belelli & Lambert, 2005) and the endocannabinoid 2-AG (Sigel et al.,\n2011) have been documented as endogenous ligands of GABAA receptors. A binding site for 2-AG has been shown to be present specifically on each of the two \u03b22 subunits present in a pentameric receptor. Moreover it has been proposed that 2-AG dips into the membrane and binds\nto the fourth trans-membrane sequence (M4) of \u03b22 subunits (Baur et al., 2013).\nNA-glycine has been shown to be present in amounts of about 50 pmol/g dry weight in rat\nbrain tissue and about 140 pmol/g dry weight in spinal cord (Huang et al., 2001). This can be compared to the levels of 2-AG that have been reported to be 4 and 50 nmol/g wet weight in brain (Sigel et al., 2011) and spinal cord (Guasti et al., 2009), respectively. Thus, NA-glycine is about 1000-fold less abundant than 2-AG. NA-glycine has no functional affinity for CB1 receptors (Sheskin et al., 1997), but may activate the G protein coupled receptors GPR18 (Kohno et al., 2006) and GPR92 (Oh et al., 2008) and target Na+/Ca2+-exchanger NCX (Bondarenko et al.,\n21\n22 23 24\n25\n26 27\n28 29 30 31 32 33\n34 35\n36 37 38 39 40 41 42 43\nPeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013)\nR ev ie w in g M an\nus cr ip t\n2013) and T-type Ca2+-channels (Barbara et al., 2009). Interestingly, NA-glycine exerts analgesic properties in different rodent models of pain (Huang et al., 2001; Succar, et al., 2007). Recently,\nit has been reported that NA-glycine can also interact with glycine receptors, potentiating \u03b11 and interestingly inhibiting \u03b12 and \u03b13-containing glycine receptors (Y\u00e9venes & Zeilhofer, 2011). It should be noted that NA-glycine functionally interacts with glycine receptors with an apparent affinity >10 \u00b5M. This should be compared to the functional affinities to GABAA receptors, NCX and T-type Ca2+-channels that are estimated <10 \u00b5M.\nWe have previously demonstrated that NA-glycine allosterically potentiates GABAA\nreceptors (Baur et al., 2013), but it remained unclear whether this occurred via the same binding site as 2-AG. The aim of the present study was therefore to establish whether NA-glycine acts at the identical site on GABAA receptors as 2-AG. In most studied aspects Na-glycine acts similar to 2-AG, indicating a common binding site. However, some point mutations that abrogate modulation by 2-AG leave the initial modulation by NA-glycine nearly unaffected and only over time the modulation is gradually decreased to zero. Differential solubilization effects of 2-AG and NA-glycine may account for this phenomena. Thus, at least during initial phases of interaction with GABAA receptors the mode of binding is different for the two compounds.",
+    "v2_Abstract": "NA-glycine is an endogenous lipid molecule with analgesic properties, which is structurally similar to the endocannabinoids 2-AG and anandamide but does not interact with cannabinoid receptors. NA-glycine is a full agonist at the G-protein coupled receptor GPR18. Recently, we have described that NA-glycine can also modulate recombinant \u03b11\u03b22\u03b32 GABAA receptors. Here we characterize in more detail this modulation and investigate the relationship of its binding site with that of the endocannabinoid 2-AG.",
+    "v2_col_introduction": "introduction  : The inhibitory GABAA receptors are composed of five subunits surrounding a central chloride ion selective channel (Macdonald & Olsen, 1994; Sieghart, 1995; Sieghart & Sperk, 2002; Sigel & Steinmann, 2012). A variety of subunit isoforms of the GABAA receptor has been cloned, leading to a multiplicity of receptor subtypes (Macdonald & Olsen, 1994; Sigel & Steinmann, 2012; Barnard et al., 1998; Olsen & Sieghart, 2008).\nThe major receptor isoform in mammalian brain consists of \u03b11, \u03b22, and \u03b32 subunits\n(Olsen & Sieghart, 2008). Different approaches have indicated a 2\u03b1:2\u03b2:1\u03b3 subunit\nstoichiometry for this receptor (Chang et al., 1996; Tretter et al., 1997; Farrar et al., 1999; Baumann et al., 2001, Baumann et al., 2002; Baur et al., 2006) with a subunit\narrangement \u03b3\u03b2\u03b1\u03b2\u03b1 anti-clockwise as seen from the synaptic cleft (Baumann et al.,\n2001, Baumann et al., 2002; Baur et al., 2006). The pharmacological properties depend both on subunit composition (Sigel et al., 1990) and arrangement (Minier and Sigel, 2004).\nNeurosteroids (Belelli & Lambert, 2005) and the endocannabinoid 2-AG (Sigel\net al., 2011) have been documented as endogenous ligands of GABAA receptors. Two\nbinding sites for 2-AG have been shown to be present specifically on the two \u03b22 subunit containing receptor pentamers. Moreover it has been shown that 2-AG dips into the\nmembrane and binds to the trans-membrane sequence (M4) of \u03b22 subunits (Baur et al., 2013).\nNA-glycine has been shown to be present in amounts of about 50 pmol/g dry\nweight in rat brain tissue and about 140 pmol/g dry weight in spinal cord (Huang et al., 2001). It has a poor affinity for CB1 receptors (Sheskin et al., 1997), but fully activates the G protein coupled receptor GPR18 (Kohno et al., 2006). Interestingly, NA-glycine\n3 3\n38\n39\n40\n41\n42\n43\n44\n45\n46\n47\n48\n49\n50\n51\n52\n53\n54\n55\n56\n57\n58\n59\n60\n61\nPeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013)\nR ev ie w in g M an\nus cr ip t\nexerts analgesic properties in different rodent models of pain (Huang et al., 2001; Succar, et al., 2007). Recently, it has been reported that NA-glycine can also interact with glycine receptors (Yvenes & Zeilhofer, 2011). We have previously demonstrated that NA-glycine allosterically potentiates GABAA receptors, but it remained unclear whether this occurred via the same binding site as 2-AG (Baur et al., 2013).\nHere we studied the interaction of NA-glycine with GABAA receptors. In most\nstudied aspects Na-glycine acts similar to 2-AG, indicating a common binding site. However, some point mutations that abrogate modulation by 2-AG leave initial modulation by NA-glycine nearly unaffected and only with time modulation is gradually decreased to zero. Differential solubilization effects of 2-AG and NA-glycine may account for this phenomena. Thus, at least during early phases of interaction with GABAA receptors the mode of binding is different for the two compounds.",
+    "v1_text": "results : Both NA-glycine and 2-AG allosterically potentiate recombinant \u03b11\u03b22\u03b32 GABAA receptors expressed in Xenopus oocytes. Both compounds share the arachidonoyl tail structure but differ in their hydrophilic head groups (Fig. 1). Please note that at physiological pH, NA-glycine is negatively charged. Here we wanted to compare the GABAA receptor binding site for NA-glycine with the well-characterized binding site for 2-AG. Figure 2a shows current traces of a cumulative concentration-response curve of the allosteric potentiation of \u03b11\u03b22\u03b32 GABAA receptors at a GABA concentration of 1 \u00b5M. At the highest concentration used the current trace displayed the typical signs of an open channel block, rapid apparent desensitization and an off current. This phenomenon made a precise curve fit impossible as maximal potentiation could not be determined precisely. The averaged concentration-response curve (Fig. 2b) was fitted with the assumption of different maximal potentiation. From these fits it was estimated that the EC50 was between 1 and 10 \u00b5M (not shown). Direct activation by 3 \u00b5M NA-glycine elicited no significant current (< 2 nA) in oocytes where 100 \u00b5M GABA elicited a current > 7 \u00b5A. Allosteric potentiation by 3 \u00b5M NA-glycine was determined at different concentrations of the endogenous agonist GABA. Fig. 3 shows that the degree of potentiation was rapidly decreasing with increasing concentrations of GABA. The comparable properties of 2-AG are also shown in Fig. 3. We tried to rationalize these findings using a model that has previously been proposed on the basis of other observations (Baumann et al., 2003; Fig. 4a). This model assumes binding of GABA to two sites differing in their binding affinity and transition to the open state with low propensity of singly ligated states and high propensity of the doubly ligated state. Fig. 4b shows computed current amplitudes in dependence of the GABA concentration. In addition, a 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t predicted curve is shown where it is assumed that NA-glycine promotes transition of the singly ligated receptor from the closed to the open state. Fig. 4c shows a computed GABA concentration-dependence of the current potentiation expected in this case. The model predicts that sizeable potentiation is limited to very low concentrations of GABA. We have previously shown that the CB1 receptor ligand DEA antagonizes potentiation by 2-AG (Baur et al., 2013). Therefore, we compared the ability of DEA to antagonize potentiation by NA-glycine and 2-AG. Based on the structural similarity of the three compounds we assumed a competitive behaviour. Fig. 5 compares the cumulative concentration inhibition curves for 3 \u00b5M NA-glycine and for 3 \u00b5M 2-AG. Potentiation by NA-glycine was inhibited half-maximally at 72 \u00b1 36 \u00b5M (n = 4) DEA and potentiation by 2-AG at 1.4 \u00b1 0.6 \u00b5M (n = 6). If the two ligands displayed a similar apparent affinity for potentiation at the same site where DEA acts, providing equal water solubility and lipid solubilization, a similar inhibitory potency of DEA would have been expected. In order to investigate if DEA and NA-glycine act competitively, we repeated a concentration inhibition curves at 6-times lower concentration (0.5 \u00b5M) of NA-glycine. Half-maximal inhibition was observed at 96 \u00b1 41 \u00b5M (n = 4) DEA (Fig. 5). This could be interpreted as non-competitive interaction of DEA with NA-glycine. Similarly, we performed concentration inhibition curves at 1 and 15 \u00b5M 2-AG. The higher concentration of 2-AG elicits direct current that amount to less than 0.1% of the maximal current amplitude elicited by GABA in the same oocytes. As expected for a competitive interaction between 2-AG and DEA, half-maximal inhibition was shifted to the left upon decrease of the 2-AG concentration from 3 \u00b5M to 1 \u00b5M, but the inhibition curve became much flatter, indicating the DEA becomes partially inactive at higher concentrations. Half-maximal inhibition for 15 \u00b5M was not reached at concentrations up to 100 \u00b5M (n = 4) DEA (Fig. 5). In case of a non-competitive interaction of 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t DEA with 2-AG an IC50 of about 1.4 \u00b5M and in case of a competitive interaction an IC50 of about 7 \u00b5M would be expected in the latter case. The observed results cannot be explained by classical receptor theory and we therefore speculate that 2-AG and NA-glycine exhibit a different water solubility and lipid solubilisation in the experimental setup. In the discussion we mention possible explanations. NA-glycine shows higher efficacy than 2-AG for potentiation of currents elicited by GABA. 2-AG is metabolically stable in Xenopus oocytes as no degradation by serine hydrolases was found (not shown). In case NA-glycine competes for the same binding site as 2-AG and both molecules have a similar apparent affinity to this binding site, it would be anticipated that the degree of potentiation by both agents at the same concentration would result in an intermediate potentiation as compared to the individual agents. The apparent affinity of 2-AG has been determined as 2 \u00b5M, while the apparent affinity of NA-glycine is estimated 1-10 \u00b5M here. As shown in Fig. 6a, combined application results surprisingly in nearly the same extent of potentiation as application of NA-glycine alone. Again, this may be caused by a differential water solubility and membrane solubilisation behaviour of NA-glycine and 2-AG. We therefore measured the critical micelle concentrations of both molecules. The apparent CMC was > 100 \u00b5M for NA-glycine and 4.2 \u00b1 0.5 \u00b5M for 2-AG, pointing to significant self-assembly and detergent behaviour of 2-AG. As 2-AG fails to potentiate in GABAA receptors where the \u03b22 subunit is replaced by \u03b21, we tested potentiation by NA-glycine in \u03b11\u03b21\u03b32 receptors (Fig. 6b). Similarly to 2-AG, potentiation by NA-glycine depends on the presence of \u03b22 subunits. We studied potentiation by NA-glycine in concatenated receptors containing either two \u03b22 subunits, two \u03b21 subunits or one each \u03b21 and \u03b22 in different positions in the receptor pentamer (Fig. 7). Receptors containing two 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t \u03b22 subunits exhibited strong potentiation while receptors containing two \u03b21 subunits showed very weak potentiation. Intermediate potentiation was observed in receptors containing one each, \u03b21 and \u03b22. This strongly indicates that the NA-glycine binding site is located on the \u03b22 subunit as previously shown with 2-AG (Sigel et al., 2011). A number of point mutations have been described to interfere with the potentiation by 2-AG. We tested the effect of the point mutations \u03b22W428C, \u03b22S429C, \u03b22F432C, \u03b22F439L and \u03b22V443C. Original current traces are shown for the mutant receptor \u03b11\u03b22S429C\u03b32. These traces are compared with traces from wild type receptors (Fig. 8a). While wild type receptors show a time-independent potentiation by NA-glycine, mutant receptors showed initially a potentiation that rapidly decayed over time. As these mutant receptors show a similar dependence on GABA as wild type receptors, and the experiment were carried out at very low GABA concentrations this current transient is not due to desensitization. In the case of 2-AG the effect of these mutations is a reduction of the potentiation independent of the time of exposure to 2-AG. This behaviour is observed with NA-glycine for the potentiation of \u03b11\u03b22F432C\u03b32 receptors, but not the other mutant receptors studied. The mutation studies indicate a site of action in the inner leaflet of M4 of the \u03b22 subunit. In this case NA-glycine has to traverse the lipid bilayer either by diffusion or mediated by a transport system and this may require some time. In order to test the time-dependence of action of NA-glycine we exposed an oocyte to GABA followed by GABA and NA-glycine (Fig. 9). Indeed, onset of modulation was slow and did not reach a steady level within 1 min. Upon switch of the medium to GABA only, a slow decay of the potentiation was observed. Discussion 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t NA-glycine allosterically potentiates GABAA receptors like the major endocannabinoid 2-AG. We aimed at localizing the site of interaction of Na-glycine with recombinant \u03b11\u03b22\u03b32 GABAA receptors relative to the site for 2-AG. An interpretation of our results is hampered by the fact that the apparent affinity for the potentiation by NA-glycine could not be determined accurately. However, we can estimate this value to be in the range of 1-10 \u00b5M, which compares well with the value of 2 \u00b5M for 2-AG (Sigel et al., 2011). The fact that we find significant potentiation of GABAA receptors by > 0.1 \u00b5M NA-glycine may reflect the better water solubility of NA-glycine over 2-AG at low concentrations and even indicate biological relevance as the average in vivo concentration in the central nervous system may be estimated from the dry tissue content as about 15-50 nM and NA-glycine is unlikely to be randomly distributed. The following observations argue for a similar mode of action of NA-glycine and 2-AG. First, both substances only act exclusively at low GABA concentration (Fig. 3), putatively by enhancing the opening of singly ligated receptor channels (Fig. 4). A leftward shift of the concentration response curve for GABA as observed in the case of benzodiazepines does not abrogate potentiation below EC50 (Sigel and Steinmann, 2012). To our knowledge, this is a new mode of action of a ligand. Second, investigation of receptors with different \u03b2 subunits (Fig. 5) and experiments with concatenated receptors containing either no, one, or two \u03b22 subunits (Fig. 6) strongly indicate that both ligand binding sites are located on the \u03b22 subunit. A common binding site in the inner leaflet of the fourth trans-membrane region (M4) of this subunit is suggested by the fact that modulation by both agents is either reduced or abolished in five identical mutant receptors, at least in the late phase of action of NA-glycine (Fig. 8b). The onset of action for both substances was found to be slow (Fig. 9; Baur et al., 2013). On the basis of these observations, it is tempting to assume a common binding site for the two ligands. 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t The following observation cannot be explained by classical receptor theory in case NA-glycine and 2-AG use an identical binding site, i.e. display a similar apparent affinity and interact with each other in a competitive way. Combined application of two compounds with similar affinities at identical concentrations is then expected to result in an intermediate potentiation as compared to that by individual compounds. Instead, the observed potentiation is similar to the one by NA-glycine alone. A second observation is difficult to reconcile with a common binding site for NA-glycine, 2-AG and the inhibitor of the potentiation of 2-AG, DEA. Namely, DEA prevents potentiation by NA-glycine only at 50-fold higher concentrations as that caused by 2-AG. As mentioned in the result section, the interaction between DEA and 2-AG cannot be explained by classical receptor theory. Since NA-glycine exerts a significant higher CMC than 2-AG, differential solubilisation of NA-glycine and 2-AG with Xenopus oocytes may account for some of the effects observed in this study. The way these lipids are organized in an aqueous environment will affect entry of the molecules into the bilayer, binding equilibrium, and the way the receptor is occupied. If this holds true the observations with co-application of NA-glycine and 2-AG as well as the inhibition of NA-glycine and DEA have to be seen in a new light. In spite of our observations the three agents could still all bind to largely overlapping sites within an extended surface able to bind flexible hydrophobic structures. The mutant receptors \u03b11\u03b22S428C\u03b32, \u03b11\u03b22R429C\u03b32, \u03b11\u03b22F439L\u03b32 and \u03b11\u03b22Y443C\u03b32 all largely abrogate modulation by NA-glycine after 1 min of combined application of GABA with NA-glycine. This abrogation is not present at the beginning of the combined application, but sets in rather slowly. We have no explanation for this observation. Solubility considerations do not help to explain this phenomenon. 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t Overall, most arguments point to a similar action and possibly overlapping binding site for NA-glycine and 2-AG. No matter what the exact mode of interaction of NA-glycine with the GABAA receptor is, this agent represents by far the stronger positive allosteric modulator than 2-AG, although the latter is more abundant in brain. The implications of our findings for the analgesic effect of NA-glycine remain to be studied. methods : material : 2-AG and NA-glycine were obtained from Cayman Chemical (Chemie Brunschwig, Basel, Switzerland). All other chemicals, unless mentioned otherwise below, were from Sigma (Buchs, Switzerland). Expression of GABAA receptors in Xenopus oocytes 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t Capped cRNAs were synthesized (Ambion, Austin, TX, USA) from the linearized plasmids with a cytomegalovirus promotor (pCMV vectors) containing the different subunits, respectively. A poly-A tail of about 400 residues was added to each transcript using yeast poly-A polymerase (United States Biologicals, Cleveland, OH, USA). The concentration of the cRNA was quantified on a formaldehyde gel using Radiant Red stain (Bio-Rad Laboratories, Reinach, Switzerland) for visualization of the RNA. Known concentrations of RNA ladder (Invitrogen, Life Technologies, Zug, Switzerland) were loaded as standard on the same gel. cRNAs were precipitated in ethanol/ isoamylalcohol 19 : 1, the dried pellet dissolved in water and stored at -80\u00b0C. cRNA mixtures were prepared from these stock solutions and stored at -80\u00b0C. Xenopus laevis oocytes were prepared, injected and defolliculated as described previously (Sigel, 1987; Sigel & Minier, 2005; Animal Permit No. BE98/12, Kantonaler Verterin\u00e4rdienst, Kanton Bern). They were injected with 50 nL of the cRNA solution containing rat wild type \u03b11 and wild type or mutated \u03b22 and wild type \u03b32 subunits at a concentration of 10 nM : 10 nM : 50 nM (Boileau et al., 2002) and then incubated in modified Barth\u2019s solution at +18\u00b0C for at least 24 h before the measurements. Where indicated concatenated subunits \u03b11-\u03b22-\u03b11 / \u03b32-\u03b22 or \u03b11-\u03b22-\u03b11 / \u03b32-\u03b21 or \u03b11-\u03b21-\u03b11 / \u03b32-\u03b22 or \u03b11-\u03b21-\u03b11 / \u03b32-\u03b21 were used at a concentration of 25 nM : 25 nM, each. Functional characterization of the GABAA receptors Currents were measured using a modified two-electrode voltage clamp amplifier Oocyte clamp OC-725 (Warner Instruments, Camden, CT, USA) in combination with a XY-recorder (90% response time 0.1s) or digitized at 100 Hz using a PowerLab 2/20 (AD Instruments) using the computer programs Chart (ADInstruments GmbH, Spechbach, Germany). Tests with a model oocyte were performed to ensure linearity in the larger current range. The response was linear up to 15 \u00b5A. 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t Electrophysiological experiments were performed using the two-electrode voltage clamp method at a holding potential of -80 mV. The perfusion medium contained 90 mM NaCl, 1 mM KCl, 1 mM MgCl2, 1 mM CaCl2, 5 mM Na-HEPES (pH 7.4) and 0.5 % DMSO and was applied by gravity flow 6 ml/min. The perfusion medium was applied through a glass capillary with an inner diameter of 1.35 mm, the mouth of which was placed about 0.4 mm from the surface of the oocyte. Allosteric modulation via the 2-AG site was measured at a GABA concentration eliciting about 1% of the maximal GABA current amplitude (EC1). In each experiment, 1 mM GABA was applied to determine the maximal current amplitude. Subsequently increasing concentrations of GABA were applied until 0.5-1 % of the maximal current amplitude was elicited (0.3-3 \u00b5M). For modulation experiments, GABA was applied for 20 s alone or in combination with 2-AG or NA-glycine. 2-AG or NA-glycine were pre-applied for 30 s. Modulation of GABA currents was expressed as (I(modulator + GABA) / IGABA \u2013 1) * 100 %. Inhibition by DEA was determined at the end of a 1 min co-application with either NA-glycine or 2-AG following a 30 s pre-application of both compounds. The perfusion system was cleaned between drug applications by washing with dimethylsulfoxide to avoid contamination. Determination of critical micelle concentrations (CMC) Assays were performed as reported previously (Raduner et al., 2007). In brief, compounds (from 2 mM stock solutions) were incubated at increasing concentrations with 0.1 nM fluorescein (free acid, 99%, Fluka, Switzerland) for 90 min at 30 \u00b0C in Nanopure distilled water. Experiments were carried out on 96-well microtiter plates (excitation at 485 nm, emission at 535 nm) on a TECAN Farcyte reader. Experiments were performed in triplicates in three independent experiments and data are mean values \u00b1 S.D. 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t abstract : NA-glycine is an endogenous lipid molecule with analgesic properties, which is structurally similar to the endocannabinoids 2-AG and anandamide but does not interact with cannabinoid receptors. NA-glycine has been suggested to act at the G-protein coupled receptors GPR18 and GPR92. Recently, we have described that NA-glycine can also modulate recombinant \u03b11\u03b22\u03b32 GABAA receptors. Here we characterize in more detail this modulation and investigate the relationship of its binding site with that of the endocannabinoid 2-AG. Keywords: GABAA receptors; GABA; endocannabinoids; 2-AG Running title: The binding site for NA-glycine 12 13 14 15 16 17 18 19 20 PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t acknowledgments : We thank Dr. V. Niggli for carefully reading the manuscript and Dr. A. Chicca for determining stability of 2-AG in Xenopus oocytes. an oocyte expressing a1\u03b22\u03b32 receptors was sequentially exposed to medium alone, to 1 m \u03bc : GABA, to the same concentration of GABA in combination with 5 M NA-glycine, to 1 microM\u03bc GABA alone and the to medium. This experiment was repeated two more times with similar results. PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t do n-arachidonyl-glycine (na-glycine) and 2-arachidonoyl glycerol (2-ag) share mode of action and the binding site on the \u03b22 subunit of gabaa receptors? : NA-glycine is an endogenous lipid molecule with analgesic properties, which is structurally similar to the endocannabinoids 2-AG and anandamide but does not interact with cannabinoid receptors. NA-glycine has been suggested to act at the G-protein coupled receptors GPR18 and GPR92. Recently, we have described that NA-glycine can also modulate recombinant \u03b1 1 \u03b2 2 \u03b3 2 GABAA receptors. Here we characterize in more detail this modulation and investigate the relationship of its binding site with that of the endocannabinoid 2-AG. PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t action and the binding site on the \u03b22 subunit of GABAA receptors? Roland Baur, J\u00fcrg Gertsch, Erwin Sigel Institute of Biochemistry and Molecular Medicine, University of Bern, CH-3012 Bern, Switzerland Address correspondence to: Erwin Sigel, Institute of Biochemistry and Molecular Medicine, University of Bern, B\u00fchlstrasse 28, Ch-3012 Bern, Switzerland. Fax: +41 31 631 3737, E-mail: erwin.sigel@ibmm.unibe.ch Abbreviations: GABA, \u03b3-aminobutyric acid; GABAA receptor, \u03b3-aminobutyric acid type A receptor; NA-glycine: N-arachidonyl-glycine; 2-AG: 2-arachidonoyl glycerol; DEA, docosatetraenylethanolamide. 1 2 3 4 5 6 7 8 9 10 11 PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t legends : Figure 1 Chemical structure of NA-glycine and 2-AG. Figure 2 Concentration-dependent potentiation of currents mediated by recombinant \u03b11\u03b22\u03b32 GABAA receptors. a) Receptors were expressed in Xenopus oocytes and currents were measured by using electrophysiological techniques at a GABA concentration eliciting 0.5-1.0 % of the maximal current amplitude (EC0.5-1.0). GABA was applied twice (single bars) and subsequently in combination with increasing concentrations of NA-glycine. The numbers above the double bars indicate the concentration of NA-glycine in \u00b5M. NA-glycine was pre-applied for 30 s. Original current traces are shown. b) shows the averaged concentration-dependent potentiation of currents elicited by GABA by NA-glycine. Four experiments as shown under a) were averaged. Data are shown as mean \u00b1 SD (n = 4). Such an averaged curve has been shown before based on 3 experiments and missing the point at 10 \u00b5M NA-glycine (Baur et al., 2013). Figure 3 Influence of the GABA concentration. Current potentiation by 3 \u00b5M NA-glycine (closed circles) or 2-AG (closed squares) was determined at different concentrations of GABA. Potentiation decreased with increasing concentrations of GABA. The GABA concentration response curve was fitted with a mean EC50 of 35 \u00b5M and a mean Hill coefficient of 1.5 (not shown). Figure 4 Mode of action of NA-glycine. Simplified model a) The model assumes two agonist binding sites 1 and 2 with different affinities. 2-AG affects the closed / open transition of 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t the two singly ligated states. The receptor R can first bind GABA (A) either to the site 1 (AR) or the site 2 (RA). The receptor occupied by two agonist molecules ARA can isomerize to the open state ARA*, the receptors occupied by a single agonist molecule can isomerize to the open states AR* and RA*. Binding is described with K as dissociation constants and gating with L as closed state / open state equlibrium. c) Theoretical GABA concentration response curves in the absence and presence of NA-glycine. The following parameters were assumed: 0.24 for L, 10 and 2.2 for L1 in the absence and presence of NA-glycine, respectively, 11 and 2.4 for L2 in the absence (line) and presence (dashed line) of NA-glycine, respectively, 30 \u00b5M for K1, 90 \u00b5M for K2. c) Dependence of the potentiation by NA-glycine on the concentration of GABA obtained by the ratio of the computed current in the presence of NA-glycine divided by the current in its absence. Figure 5 Concentration inhibition curve of DEA. Increasing concentrations of DEA were co-applied with 1 \u00b5M 2-AG (open circles), 3 \u00b5M 2-AG (open squares), 15 \u00b5M 2-AG (open diamonds), 0.5 \u00b5M NA-glycine (filled circles), or 3 \u00b5M NA-glycine (filled squares). Data are shown as mean \u00b1 SEM (n = 4). Figure 6 Effect of subunit combination and co-application with 2-AG. a) Current potentiation by the combined application of 3 \u00b5M NA-glycine and 3 \u00b5M 2-AG is compared with the individual application of the two substances. b) Current potentiation by 3 \u00b5M NA-glycine in \u03b11\u03b22\u03b32 receptors and \u03b11\u03b21\u03b32 receptors. Potentiation is strongly dependent on the presence of the \u03b22 subunit. Figure 7 Concentration-dependent potentiation of currents mediated by concatenated 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t GABAA receptors. Concatenated \u03b11-\u03b21-\u03b11/\u03b32-\u03b21, \u03b11-\u03b21-\u03b11/\u03b32-\u03b22, \u03b11-\u03b22-\u03b11/\u03b32-\u03b21 or \u03b11-\u03b22-\u03b11/\u03b32-\u03b22 receptors were expressed in Xenopus oocytes and currents were measured at a GABA concentration eliciting 0.5-1.0 % of the maximal current amplitude (EC0.5-1.0). Current potentiation by increasing concentrations of NA-glycine was determined. Four such experiments were averaged. Data are shown as mean \u00b1 SD (n = 4). Figure 8 Effect of point mutations that reduced potentiation by 2-AG on the potentiation of NA-glycine. a) Potentiation by 3 \u00b5M NA-glycine is compared between wild type receptors and receptors containing the point mutation S429C in the \u03b22 subunit. This mutation results at the beginning of the drug application in an about 50% reduction of potentiation and after 1 min drug application potentiation is abolished. b) Wild type receptors are compared with mutant receptors. Current potentiation is indicated at the beginning of the drug application (filled bars) and after 1 min drug exposure (open bars). Figure 9 Time course of the potentiation by NA-glycine. An oocyte expressing \u03b11\u03b22\u03b32 receptors was sequentially exposed to medium alone, to 1 \u00b5M GABA, to the same concentration of GABA in combination with 5 \u00b5M NA-glycine, to 1 \u00b5M GABA alone and the to medium. This experiment was repeated two more times with similar results. 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t Figure 1 chemical structure of na-glycine and 2-ag : PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t Figure 2 Concentration-dependent potentiation of currents mediated by recombinant \u03b1 1 \u03b2 2 \u03b3 2 GABAA receptors a) Receptors were expressed in Xenopus oocytes and currents were measured by using electrophysiological techniques at a GABA concentration eliciting 0.5-1.0 % of the maximal current amplitude (EC0.5-1.0). Original current traces are shown. The experiment was carried out with a GABA concentration of 1 M. b) Four such experiments were averaged. Data are \u03bc shown as mean \u00b1 SD (n = 4). PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t Figure 3 influence of the gaba concentration : Current potentiation by 3 micro M NA-glycine (closed circles) or 2-AG (closed squares) was determined at different concentrations of GABA. Potentiation decreased with increasing concentrations of GABA. PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t Figure 4 mode of action of na-glycine : Simplified model a) The model assumes two agonist binding sites 1 and 2 with different affinities. 2-AG affects the closed / open transition of the two singly ligated states. The receptor R can first bind GABA (A) either to the site 1 (AR) or the site 2 (RA). The receptor occupied by two agonist molecules ARA can isomerize to the open state ARA*, the receptors occupied by a single agonist molecule can isomerize to the open states AR* and RA*. Constants are taken as dissociation constants and gating constants as closed state / open state. c) Theoretical GABA concentration response curves in the absence and presence of NA-glycine. The following parameters were assumed: 0.24 for L, 10 and 2.2 for L1 in the absence and presence of NA-glycine, respectively, 11 and 2.4 for L2 in the absence (line) and presence (dashed line) of NA-glycine, respectively, 30 \u00b5M for K1, 90 \u00b5M for K2 . c) Dependence of the potentiation by NA-glycine on the concentration of GABA obtained by the ratio of the computed current in the presence of NA-glycine divided by the current in its absence. PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t Figure 5 Concentration inhibition curve of DEA. Increasing concentrations of DEA were co-applied with 1 M 2-AG (open circles), 3 M 2-AG\u03bc \u03bc (open squares), 15 M 2-AG (open diamonds), 0.5 M NA-glycine (filled circles), or 3 M NA-\u03bc \u03bc \u03bc glycine (filled squares). Data are shown as mean \u00b1 SEM (n = 4). PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t Figure 6 Effect of subunit combination and co-application with 2-AG a) Current potentiation by the combined application of 3 M NA-glycine and 3 mM 2-AG is \u03bc compared with the individual application of the two substances. b) Current potentiation by 3 M NA-glycine in \u03bc \u03b11\u03b22\u03b32 receptors and \u03b11\u03b21\u03b32 receptors. Potentiation is strongly dependent on the presence of the \u03b22 subunit. PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t Figure 7 Concentration-dependent potentiation of currents mediated by concatenated GABAA receptors Concatenated \u03b11-\u03b21-\u03b11 / \u03b32-\u03b21 , \u03b11-\u03b21-\u03b11 / \u03b32-\u03b22 , \u03b11 -\u03b22-\u03b11 / \u03b32-\u03b21 or \u03b11-\u03b22-\u03b11 / \u03b32-\u03b22 receptors receptors were expressed in Xenopus oocytes and currents were measured at a GABA concentration eliciting 0.5-1.0 % of the maximal current amplitude (EC0.5-1.0). Current potentiation by increasing concentrations of NA-glycine was determined. Four such experiments were averaged. Data are shown as mean \u00b1 SD (n = 4). PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t Figure 8 Effect of point mutations that reduced potentiation by 2-AG on the potentiation of NAglycine. a) Potentiation by 3 M NA-glycine is compared between wild type receptors and receptors \u03bc containing the point mutation S429C in the \u03b22 subunit. This mutation results at the beginning of the drug application in an about 50% reduction of potentiation and after 1 min drug application potentiation is abolished. b) Wild type receptors are compared with mutant receptors. Current potentiation is indicated at the beginning of the drug application (filled bars) and after 1 min drug exposure (open bars). PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:615:1:0:ACCEPTED 22 Aug 2013) R ev ie w in g M an us cr ip t Figure 9 Time course of the potentiation by NA-glycine",
+    "v2_text": "abstract : NA-glycine is an endogenous lipid molecule with analgesic properties, which is structurally similar to the endocannabinoids 2-AG and anandamide but does not interact with cannabinoid receptors. NA-glycine is a full agonist at the G-protein coupled receptor GPR18. Recently, we have described that NA-glycine can also modulate recombinant \u03b11\u03b22\u03b32 GABAA receptors. Here we characterize in more detail this modulation and investigate the relationship of its binding site with that of the endocannabinoid 2-AG. Keywords: GABAA receptors; GABA; endocannabinoids; 2-AG Running title: The binding site for NA-glycine 2 2 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t results : Both NA-glycine and 2-AG allosterically potentiate recombinant \u03b11\u03b22\u03b32 GABAA receptors expressed in Xenopus oocytes. Both compounds share the arachidonoyl tail structure but differ in the structure of their hydrophilic head group (Fig. 1). We wanted 6 6 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t to compare the GABAA receptor binding site for NA-glycine with the well-characterized binding site for 2-AG. Figure 2a shows current traces of a cumulative concentration-response curve of the allosteric potentiation of \u03b11\u03b22\u03b32 GABAA receptors at a GABA concentration of 1 \u00b5M. At the highest concentration used here the current trace displayed the typical signs of an open channel block, rapid apparent desensitization and an off current. This phenomenon made a precise curve fit impossible as maximal potentiation could not be determined precisely. The averaged concentration-response curve (Fig. 2b) was fitted with the assumption of different maximal potentiation. From these fits it was estimated that the EC50 was between 1 and 10 \u00b5M (not shown). Direct activation by 3 \u00b5M NA-glycine elicited no significant current (< 2 nA) in oocytes where 100 \u00b5M GABA elicited > 7 \u00b5A. Allosteric potentiation by 3 \u00b5M NA-glycine was determined at different concentrations of the endogenous agonist GABA. Fig. 3 shows that the degree of potentiation was rapidly decreasing with increasing concentrations of GABA. The comparable properties of 2-AG are also shown in Fig. 3. We tried to rationalize these findings using a model that has previously been proposed on the basis of other observations (Baumann et al., 2003; Fig. 4a). This model assumes binding of GABA to two sites differing in their binding affinity and isomerization to the open state with low propensity of singly ligated states and high propensity of the doubly ligated state. Fig. 4b shows computed current amplitudes in dependence of the GABA concentration. In addition, a predicted curve is shown, where it is assumed that NA-glycine promotes isomerization of the singly ligated receptor from the closed to the open state. Fig. 4c shows a computed GABA concentration-dependence of the current potentiation 7 7 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t expected in this case. The model predicts that sizeable potentiation is limited to very low concentrations of GABA. The CB1 receptor ligand DEA antagonizes potentiation by 2-AG (Baur et al., 2013). Therefore, we compared the ability of DEA to antagonize potentiation by NA-glycine with 2-AG. Based on the structural similarity of the three compounds we assumed a competitive behaviour. Fig. 5 compares the cumulative concentration inhibition curves for 3 \u00b5M NA-glycine and for 3 \u00b5M 2-AG. Potentiation by NA-glycine was inhibited half-maximally at 72 \u00b1 36 \u00b5M (n = 4) DEA and potentiation by 2-AG at 1.4 \u00b1 0.6 \u00b5M (n = 6). If the two drugs displayed a similar apparent affinity for potentiation at the same site where DEA acts, providing equal water solubility and lipid solubilization, a similar inhibitory potency of DEA would have been expected. In order to investigate if DEA and NA-glycine act competitively, we repeated a concentration inhibition curves at 6-times lower concentration (0.5 \u00b5M) of NA-glycine. Half-maximal inhibition was observed at 96 \u00b1 41 \u00b5M (n = 4) DEA (Fig. 5). This could be interpreted as non-competitive interaction of DEA with NA-glycine. Similarly, we performed concentration inhibition curves at 15 \u00b5M 2-AG. Half-maximal inhibition was not reached at concentrations up to 100 \u00b5M (n = 4) DEA (Fig. 5). In case of a non-competitive interaction of DEA with 2-AG an IC50 of about 1.4 \u00b5M and in case of a competitive interaction an IC50 of about 7 \u00b5M would be expected. The observed results cannot be explained by classical receptor theory and we therefore speculate that 2-AG and NA-glycine exhibit a different water solubility and lipid solubilisation in the experimental setup. In the discussion we mention possible explanations. 8 8 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t NA-glycine has higher efficacy than 2-AG for potentiation of currents elicited by GABA. 2-AG is metabolically stable in Xenopus oocytes as no degradation by serine hydrolases was found (not shown). In case NA-glycine competes for the same binding site as 2-AG and both molecules have a similar apparent affinity to the binding site, it would be anticipated that the degree of potentiation by both agents at the same concentration would result in an intermediate potentiation as compared to the individual agents. The apparent affinity of 2-AG has been determined as 2 \u00b5M, while the apparent affinity of NA-glycine is estimated 1-10 \u00b5M here. As shown in Fig. 6a, combined application results surprisingly in nearly the same extent of potentiation as application of NA-glycine alone. Again, this may be caused by a differential water solubility and membrane solubilisation behaviour of NA-glycine and 2-AG.. We therefore measured the critical micelle concentrations of both molecules. The apparent CMC was > 100 \u00b5M for NA-glycine and 4.2 \u00b1 0.5 \u00b5M for 2-AG, pointing to significant self-assembly and detergent behaviour of 2-AG. As 2-AG fails to potentiate in GABAA receptors where the \u03b22 subunit is replaced by \u03b21, we tested potentiation by NA-glycine in \u03b11\u03b21\u03b32 (Fig. 6b). Similarly to 2-AG, potentiation by NA-glycine depends on the presence of \u03b22 subunits. We studied potentiation by NA-glycine in concatenated receptors containing either two \u03b22 subunits, two \u03b21 subunits or one each \u03b21 and \u03b22 in different positions in the receptor pentamer (Fig. 7). In receptors containing either two \u03b22 subunits strong potentiation, two \u03b21 subunits very weak potentiation and in both receptors containing one each, \u03b21 and \u03b22, intermediate potentiation was observed. This strongly indicates that the NA-glycine 9 9 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t binding site is located on the \u03b22 subunit as previously shown with 2-AG (Sigel et al., 2011). A number of point mutations have been described to interfere with potentiation by 2-AG. We tested the effect of the point mutations \u03b22S428C, \u03b22R429C, \u03b22F432C, \u03b22F439L and \u03b22Y443C. Original current traces are shown for the mutant receptors \u03b11\u03b22R429C\u03b32. These traces are compared with traces from wild type receptors (Fig. 8a). While wild type receptors show a time-independent potentiation by NA-glycine, mutant receptors showed initially a potentiation that rapidly decayed over time. As these mutant receptors show a similar dependence to GABA as wild type receptors and the experiment were carried out at very low GABA concentrations this current transient is not due to desensitization. In the case of 2-AG the effect of these mutations is to reduce the potentiation independent of the time of exposure to 2-AG. This behaviour is observed with NA-glycine for the potentiation of \u03b11\u03b22F432C\u03b32 receptors, but not the other mutant receptors studied. The mutation studies indicate a site of action in the inner leaflet of M4 of the \u03b22 subunit. In this case NA-glycine has to traverse the lipid bilayer either by diffusion or mediated by a transport system and this may require some time. In order to test this we exposed an oocyte to GABA followed by GABA and NA-glycine (Fig. 9). Indeed, onset of modulation was slow and did not reach a steady level within 1 min. Upon switch of the medium to GABA only, a slow decay of the potentiation was observed. This again points to the differences of 2-AG and NA-glycine in the way they dissolve and segregate into the membranes. Discussion 10 10 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t NA-glycine allosterically potentiates GABAA receptors like the major endocannabinoid 2-AG. We aimed at localizing the site of interaction of Na-glycine with recombinant \u03b11\u03b22\u03b32 GABAA receptors relative to the site for 2-AG. An interpretation of our results is made more difficult by the fact that the apparent affinity for the potentiation by NA-glycine could not be determined accurately. However, we can estimate this value to be in the range of 1-10 \u00b5M, which compares well with the value of 2 \u00b5M for 2-AG (Sigel et al., 2011). The fact that we find significant potentiation of GABAA receptors by > 0.1 \u00b5M NA-glycine may reflect the better water solubility of NA-glycine over 2-AG at low concentrations and even indicate biological relevance as the average in vivo concentration in the central nervous system may be estimated from the dry tissue content as about 15-50 nM and NA-glycine is unlikely to be randomly distributed. The following observations argue for a similar mode of action of NA-glycine and 2-AG. First, both drugs only act exclusively at low GABA concentration (Fig. 3), putatively by enhancing the opening of singly ligated receptor channels (Fig. 4). A leftward shift of the concentration response curve for GABA as observed in the case of benzodiazepines does not abrogate potentiation below EC50 (Sigel and Steinmann, 2012). To our knowledge, this is a new mode of action of a ligand. Second, investigation of receptors with different \u03b2 subunits (Fig. 5) and experiments with concatenated receptors containing either no, one, or two \u03b22 subunits (Fig. 6) strongly indicate that both ligand binding sites are located on the \u03b22 subunit. A common binding site in the inner leaflet of the fourth trans-membrane region (M4) of this subunit is suggested by the fact that modulation by both agents is either reduced or abolished in five identical 11 11 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t mutant receptors, at least in the late phase of action of NA-glycine (Fig. 8b). The onset of action for both drugs was found to be slow (Fig. 9; Baur et al., 2013). On the basis of these observations, it is tempting to assume a common binding site for the two drugs. The following observation cannot be explained by classical receptor theory in case NA-glycine and 2-AG use the identical binding site, thus display a similar apparent affinity and interact with each other in a competitive way. Combined application at identical concentrations is then expected to result in an intermediate potentiation as compared to that by individual compounds. Instead the observed potentiation is similar to the one by NA-glycine alone. A second observation is difficult to reconcile with a common binding site for NA-glycine, 2-AG and the inhibitor of the potentiation of 2-AG, DEA. Namely, DEA prevents potentiation by NA-glycine only at 50-fold higher concentrations as that caused by 2-AG. As mentioned in the result section, the interaction between DEA and 2-AG cannot be explained by classical receptor theory. Since NA-glycine exerts a significant higher CMC than 2-AG, differential solubilisation of NA-glycine and 2-AG with Xenopus oocytes may account for some of the effects observed in this study. The way these lipids are organized in an aqueous environment will affect entry of the molecules into the bilayer, binding equilibrium, and the way the receptor is occupied. If this holds true the observations with co-application of NA-glycine and 2-AG as well as the inhibition of NA-glycine and DEA have to be seen in a new light. In spite of our observations the three agents could still all bind to largely overlapping sites and the receptor site may be a surface able to bind flexible hydrophobic structures. The mutant receptors \u03b11\u03b22S428C\u03b32, \u03b11\u03b22R429C\u03b32, \u03b11\u03b22F439L\u03b32 and \u03b11\u03b22Y443C\u03b32, all largely abrogate modulation by NA-glycine after 1 min of combined 12 12 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t application of GABA with NA-glycine. This abrogation is not present at the beginning of the combined application, but sets in rather slowly. We have no explanation for these observations. Solubility considerations do not help to explain here. Overall, most arguments point to a similar action and possibly overlapping binding site for NA-glycine and 2-AG. No matter what the exact mode of interaction of NA-glycine with the GABAA receptor is, this agent represents by far the more potent positive allosteric modulator than 2-AG, although the latter is more abundant in brain. The implications of our findings for the analgesic effect of NA-glycine remain to be studied. methods : material : 2-AG and NA-glycine were obtained from Cayman Chemical (Chemie Brunschwig, Basel, Switzerland). All other chemicals, unless mentioned otherwise below, were from Sigma (Buchs, Switzerland). Expression of GABAA receptors in Xenopus oocytes Capped cRNAs were synthesized (Ambion, Austin, TX, USA) from the linearized plasmids with a cytomegalovirus promotor (pCMV vectors) containing the different subunits, respectively. A poly-A tail of about 400 residues was added to each transcript using yeast poly-A polymerase (United States Biologicals, Cleveland, OH, USA). The concentration of the cRNA was quantified on a formaldehyde gel using 4 4 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t Radiant Red stain (Bio-Rad Laboratories, Reinach, Switzerland) for visualization of the RNA. Known concentrations of RNA ladder (Invitrogen, Life Technologies, Zug, Switzerland) were loaded as standard on the same gel. cRNAs were precipitated in ethanol/ isoamylalcohol 19 : 1, the dried pellet dissolved in water and stored at -80\u00b0C. cRNA mixtures were prepared from these stock solutions and stored at -80\u00b0C. Xenopus laevis oocytes were prepared, injected and defolliculated as described previously (Sigel, 1987; Sigel & Minier, 2005; Animal Permit No. BE98/12, Kantonaler Verterin\u00e4rdienst, Kanton Bern). They were injected with 50 nL of the cRNA solution containing wild type \u03b11 and wild type or mutated \u03b22 and wild type \u03b32 subunits at a concentration of 10 nM : 10 nM : 50 nM (Boileau et al., 2002) and then incubated in modified Barth\u2019s solution at +18\u00b0C for at least 24 h before the measurements. Where indicated concatenated subunits \u03b11-\u03b22-\u03b11 / \u03b32-\u03b22 or \u03b11-\u03b22-\u03b11 / \u03b32-\u03b21 or \u03b11-\u03b21-\u03b11 / \u03b32-\u03b22 or \u03b11-\u03b21-\u03b11 / \u03b32-\u03b21 were used at a concentration of 25 nM : 25 nM, each. Functional characterization of the GABAA receptors Currents were measured using a modified two-electrode voltage clamp amplifier Oocyte clamp OC-725 (Warner Instruments, Camden, CT, USA) in combination with a XY-recorder (90% response time 0.1s) or digitized at 100 Hz using a PowerLab 2/20 (AD Instruments) using the computer programs Chart (ADInstruments GmbH, Spechbach, Germany). Tests with a model oocyte were performed to ensure linearity in the larger current range. The response was linear up to 15 \u00b5A. Electrophysiological experiments were performed using the two-electrode voltage clamp method at a holding potential of -80 mV. The perfusion medium contained 90 mM NaCl, 1 mM KCl, 1 mM MgCl2, 1 mM CaCl2, 5 mM Na-HEPES (pH 7.4) and 0.5 % DMSO and was applied by gravity flow 6 ml/min. The perfusion 5 5 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t medium was applied through a glass capillary with an inner diameter of 1.35 mm, the mouth of which was placed about 0.4 mm from the surface of the oocyte. Allosteric modulation via the 2-AG site was measured at a GABA concentration eliciting about 1% of the maximal GABA current amplitude (EC1). In each experiment, 1 mM GABA was applied to determine the maximal current amplitude. Subsequently increasing concentrations of GABA were applied until 0.5-1 % of the maximal current amplitude was elicited (0.3-3 \u00b5M). For modulation experiments, GABA was applied for 20 s alone or in combination with 2-AG or NA-glycine. 2-AG or NA-glycine were pre-applied for 30 s. Modulation of GABA currents was expressed as (I(modulator + GABA) / IGABA \u2013 1) * 100 %. The perfusion system was cleaned between drug applications by washing with dimethylsulfoxide to avoid contamination. Determination of critical micelle concentrations (CMC) Assays were performed as reported previously (Raduner et al., 2007). In brief, compounds (from 2 mM stock solutions) were incubated at increasing concentrations with 0.1 nM fluorescein (free acid, 99%, Fluka, Switzerland) for 90 min at 30 \u00b0C in Nanopure distilled water. Experiments were carried out on 96-well microtiter plates (excitation at 485 nm, emission at 535 nm) on a TECAN Farcyte reader. Experiments were performed in triplicates in three independent experiments and data are mean values \u00b1 S.D. acknowledgments : We thank Dr. V. Niggli for carefully reading the manuscript. an oocyte expressing a1\u03b22\u03b32 receptors was sequentially exposed to medium alone, to 1 \u03bcm gaba, to : the same concentration of GABA in combination with 5 \u03bcM NA-glycine, to 1 microM GABA alone and the to medium. This experiment was repeated two more times with similar results. PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t do n-arachidonyl-glycine (na-glycine) and 2-arachidonoyl glycerol (2-ag) share mode of action and the binding site on the \u03b22 subunit of gabaa receptors? : NA-glycine is an endogenous lipid molecule with analgesic properties, which is structurally similar to the endocannabinoids 2-AG and anandamide but does not interact with cannabinoid receptors. NA-glycine is a full agonist at the G-protein coupled receptor GPR18. Recently, we have described that NA-glycine can also modulate recombinant \u03b1 1 \u03b2 2 \u03b3 2 GABAA receptors. Here we characterize in more detail this modulation and investigate the relationship of its binding site with that of the endocannabinoid 2-AG. PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t mode of action and the binding site on the \u03b22 subunit of GABAA receptors? Roland Baur, J\u00fcrg Gertsch, Erwin Sigel Institute of Biochemistry and Molecular Medicine, University of Bern, CH-3012 Bern, Switzerland Address correspondence to: Erwin Sigel, Institute of Biochemistry and Molecular Medicine, University of Bern, B\u00fchlstrasse 28, Ch-3012 Bern, Switzerland. Fax: +41 31 631 3737, E-mail: erwin.sigel@ibmm.unibe.ch Abbreviations: GABA, \u03b3-aminobutyric acid; GABAA receptor, \u03b3-aminobutyric acid type A receptor; NA-glycine: N-arachidonyl-glycine; 2-AG: 2-arachidonoyl glycerol; DEA, docosatetraenylethanolamide. 1 1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t legends : Figure 1 Chemical structure of NA-glycine and 2-AG. Figure 2 Concentration-dependent potentiation of currents mediated by recombinant \u03b11\u03b22\u03b32 GABAA receptors. a) Receptors were expressed in Xenopus oocytes and currents were measured by using electrophysiological techniques at a GABA concentration eliciting 0.5-1.0 % of the maximal current amplitude (EC0.5-1.0). Original current traces are shown. The experiment was carried out with a GABA concentration of 1 \u00b5M. b) Four such experiments were averaged. Data are shown as mean \u00b1 SD (n = 4). Figure 3 Influence of the GABA concentration. Current potentiation by 3 \u00b5M NA-glycine (closed circles) or 2-AG (closed squares) was determined at different concentrations of GABA. Potentiation decreased with increasing concentrations of GABA. 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 R ev ie w in g M an us cr ip t following parameters were assumed: 0.24 for L, 10 and 2.2 for L1 in the absence and presence of NA-glycine, respectively, 11 and 2.4 for L2 in the absence (line) and presence (dashed line) of NA-glycine, respectively, 30 \u00b5M for K1, 90 \u00b5M for K2. c) Dependence of the potentiation by NA-glycine on the concentration of GABA obtained by the ratio of the computed current in the presence of NA-glycine divided by the current in its absence. Figure 5 Concentration inhibition curve of DEA. Increasing concentrations of DEA were co-applied with 3 \u00b5M 2-AG (open circles), 15 \u00b5M 2-AG (open squares), 0.5 \u00b5M NA-glycine (filled circles), or 3 \u00b5M NA-glycine (filled squares). Data are shown as mean \u00b1 SEM (n = 4). Figure 6 Effect of subunit combination and co-application with 2-AG. a) Current potentiation by 3 \u00b5M NA-glycine in \u03b11\u03b22\u03b32 receptors and \u03b11\u03b21\u03b32 receptors. Potentiation is strongly dependent on the presence of the \u03b22 subunit. b) Current potentiation by the combined application of 3 \u00b5M NA-glycine and 3 \u00b5M 2-AG is compared with the individual application of the two drugs. Figure 7 Concentration-dependent potentiation of currents mediated by concatenated GABAA receptors. Concatenated \u03b11\u03b21\u03b11/\u03b32\u03b21, \u03b11\u03b21\u03b11/\u03b32\u03b22, \u03b11\u03b22\u03b11/\u03b32\u03b21 or \u03b11\u03b22\u03b11/\u03b32\u03b22 receptors receptors were expressed in Xenopus oocytes and currents were measured at a GABA concentration eliciting 0.5-1.0 % of the maximal current amplitude (EC0.5-1.0). Current potentiation by increasing concentrations 18 18 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t of NA-glycine was determined. Four such experiments were averaged. Data are shown as mean \u00b1 SD (n = 4). Figure 8 Effect of point mutations that reduced potentiation by 2-AG on the potentiation of NA-glycine. a) Potentiation by 3 \u00b5M NA-glycine is compared between wild type receptors and receptors containing the point mutation R429C in the \u03b22 subunit. This mutation results at the beginning of the drug application in an about 50% reduction of potentiation and after 1 min drug application potentiation is abolished. b) Wild type receptors are compared with mutant receptors. Current potentiation is indicated at the beginning of the drug application (filled bars) and after 1 min drug exposure (open bars). Figure 9 Time course of the potentiation by NA-glycine. An oocyte expressing \u03b11\u03b22\u03b32 receptors was sequentially exposed to medium alone, to 1 \u00b5M GABA, to the same concentration of GABA in combination with 5 \u00b5M NA-glycine, to 1 \u00b5M GABA alone and the to medium. This experiment was repeated two more times with similar results. 19 19 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t Figure 1 chemical structure of na-glycine and 2-ag : PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t Figure 2 Concentration-dependent potentiation of currents mediated by recombinant \u03b1 1 \u03b2 2 \u03b3 2 GABAA receptors a) Receptors were expressed in Xenopus oocytes and currents were measured by using electrophysiological techniques at a GABA concentration eliciting 0.5-1.0 % of the maximal current amplitude (EC0.5-1.0). Original current traces are shown. The experiment was carried out with a GABA concentration of 1 \u03bcM. b) Four such experiments were averaged. Data are shown as mean \u00b1 SD (n = 4). PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t Figure 3 influence of the gaba concentration : Current potentiation by 3 micro M NA-glycine (closed circles) or 2-AG (closed squares) was determined at different concentrations of GABA. Potentiation decreased with increasing concentrations of GABA. PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t Figure 4 mode of action of na-glycine : Simplified model a) The model assumes two agonist binding sites 1 and 2 with different affinities. 2-AG affects the closed / open transition of the two singly ligated states. The receptor R can first bind GABA (A) either to the site 1 (AR) or the site 2 (RA). The receptor occupied by two agonist molecules ARA can isomerize to the open state ARA*, the receptors occupied by a single agonist molecule can isomerize to the open states AR* and RA*. Constants are taken as dissociation constants and gating constants as closed state / open state. c) Theoretical GABA concentration response curves in the absence and presence of NA-glycine. The following parameters were assumed: 0.24 for L, 10 and 2.2 for L1 in the absence and presence of NA-glycine, respectively, 11 and 2.4 for L2 in the absence (line) and presence (dashed line) of NA-glycine, respectively, 30 \u00b5M for K1, 90 \u00b5M for K2 . c) Dependence of the potentiation by NA-glycine on the concentration of GABA obtained by the ratio of the computed current in the presence of NA-glycine divided by the current in its absence. PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t Figure 5 concentration inhibition curve of dea : Increasing concentrations of DEA were co-applied with 3 \u03bcM 2-AG (open circles), 15 \u03bcM 2-AG (open squares), 0.5 \u03bcM NA-glycine (filled circles), or 3 \u03bcM NA-glycine (filled squares). Data are shown as mean \u00b1 SEM (n = 4). PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t Figure 6 Effect of subunit combination and co-application with 2-AG a) Current potentiation by 3 \u03bcM NA-glycine in \u03b11\u03b22\u03b32 receptors and \u03b11\u03b21\u03b32 receptors. Potentiation is strongly dependent on the presence of the \u03b22 subunit. b) Current potentiation by the combined application of 3 \u03bcM NA-glycine and 3 \u03bcM 2-AG is compared with the individual application of the two drugs. PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t Figure 7 Concentration-dependent potentiation of currents mediated by concatenated GABAA receptors Concatenated \u03b11-\u03b21-\u03b11 / \u03b32-\u03b21 , \u03b11-\u03b21-\u03b11 / \u03b32-\u03b22 , \u03b11 -\u03b22-\u03b11 / \u03b32-\u03b21 or \u03b11-\u03b22-\u03b11 / \u03b32-\u03b22 receptors receptors were expressed in Xenopus oocytes and currents were measured at a GABA concentration eliciting 0.5-1.0 % of the maximal current amplitude (EC0.5-1.0). Current potentiation by increasing concentrations of NA-glycine was determined. Four such experiments were averaged. Data are shown as mean \u00b1 SD (n = 4). PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t Figure 8 Effect of point mutations that reduced potentiation by 2-AG on the potentiation of NA-glycine a) Potentiation by 3 \u03bcM NA-glycine is compared between wild type receptors and receptors containing the point mutation R429C in the \u03b22 subunit. This mutation results at the beginning of the drug application in an about 50% reduction of potentiation and after 1 min drug application potentiation is abolished. b) Wild type receptors are compared with mutant receptors. Current potentiation is indicated at the beginning of the drug application (filled bars) and after 1 min drug exposure (open bars). PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:615:0:1:NEW 25 Jun 2013) R ev ie w in g M an us cr ip t Figure 9 Time course of the potentiation by NA-glycine",
+    "url": "https://peerj.com/articles/151/reviews/",
+    "review_1": "Robert Toonen \u00b7 Aug 14, 2013 \u00b7 Academic Editor\nACCEPT\nThank you for clearly addressing the suggestions of the referees in your revisions. I am happy to say that I believe your manuscript is a valuable contribution to the field, and can now be accepted for publication. Keep up the good work.",
+    "review_2": "Robert Toonen \u00b7 Jul 28, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nOverall both referees were very positive about your paper and recommend acceptance. I concur with them, and am pleased to be able to tell you that your paper is acceptable for publication pending minor revisions in response to the comments raised by the referees. In particular, I concur with the comment about the quality of the figures - I am not sure if it is an issue of the PDF conversion or not, but the figures were hard to read and Figure has grey/white in a & c, but b & d appear all grey with just text \"before\" and \"after\" - some color-coding here would really help the readability of the figure.\n\nReferee #2 suggests some additional references and asks exactly which literature to date has mischaracterized the susceptibility of deeper reefs to storm damage. I think this comment is easily dealt with by a slight revision of the text to something along the lines of it being intuitive that depth provides some buffer to storm damage or adding a reference such as Hughes & Connell (1999 L&0 44:932\u2013940) or Scoffin (1993 Coral Reefs 12:203-221) to support your statement that there is a perception that mesophotic reefs are less susceptible to storm damage. In addition to the Harmelin-Vivien (1994) review pointed out by the referee, a more recent missing reference seems to be Bongaerts et al. 2013 Coral Reefs (Cyclone damage at mesophotic depths on Myrmidon Reef).\n\nFinally, I also wanted to pass along the confidential comment from one suggested referee who declined to review the paper, because they are similar in tone to those of Reviewer #2. They said that it seemed to be a fine study, but they felt it was \"over-sold in an attempt to ride the mesophotic bandwagon and imply that the study is more novel than it really is.\" They felt that the work should stand on it's own and that the novelty pitch severly detracted from the paper and they would reject it based solely on that rather than the content. I tell you this because I believe that one of the benefits of review is to get feedback about how some readers view the work, so that you have the option to revise before final publication. I am fine with whatever you decide in this regard, but wanted to pass along the information so that you can consider toning it down if you choose prior to final publication.\n\nAloha,\nRob",
+    "review_3": "Reviewer 1 \u00b7 Jul 25, 2013\nBasic reporting\nThis study is very interesting to show that corals at deeper depth are less resistant to the storm than we expected.\nExperimental design\nno problem\nValidity of the findings\ngood\nAdditional comments\nI think that Typhoon 16 also affected the deep coral communities. I wonder why the authors consider it.\n\nFigure 1: it is hard to see any words in this figure. Also should put \"a\" and \"b\" on the figure.\n\nFigure 4: need much more resoluiion. Also need to explain more how to see the data. I do not understand the meaning \"each circle displays relative abundance of the species\" from this figure.\nAnd what do you mean the number of coral rubble?\nAlso it is very complicated to see this because coral rubbles were shown by full circles, for both \"before\" and \"after\".\n\nMinor comments\n\"Montastraea\" is correct, not \"Montastrea\".\nAlso recently \"Montastraea\" was separated into three genera, Montastraea (for M. cavernosa), Phymastrea (for Indo-Pacific species), Orbicella (for M. annularis complex). If possible, consider it when the authors use \"Montastraea\".\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Typhoon damage on a shallow mesophotic reef in Okinawa, Japan (v0.1)\". PeerJ https://doi.org/10.7287/peerj.151v0.1/reviews/1",
+    "review_4": "Samuel Kahng \u00b7 Jul 22, 2013\nBasic reporting\nThe authors present novel data measuring and analyzing coral community structure at two intermediate depths before and after a major storm/disturbance. Given the relatively depauperate nature of the primary literature on this topic, this data set provide valuable insight into the susceptibility of coral reef communities at intermediate depths. The experimental design and data analyses are sound, and the manuscript is well written. The fundamentals of this manuscript warrant publication, however, minor revisions are warranted in the opinion of this reviewer. Detailed comments have been included in the pdf using the comments feature.\n\nThe context of this study has been overstated and overgeneralized. The data for this study was collected at two similar depths (27 & 31m) boarding the upper mesophotic. To no fault of the authors, the upper depth limit of term \u201cmesophotic\u201d as coined by NOAA is poorly defined to correspond to the lower depth limits of recreational Scuba diving at 30-40m. This upper depth limit has no ecologically meaningful relevance to corals but pertains to a human (or socio-economic) limitation. The use of this data to generalize the susceptibility of mesophotic reefs (which can extend to depths >3X the depth of this study) to storm damage is a bit of a stretch. Understandably, while the authors attempted to collected data from a larger range of depths, the absence of data from shallower depths weakens their ability to generalize conclusions comparing the storm impact on \u201cshallow\u201d versus deep coral reefs.\n\nWhile this study is valuable and detailed, previous studies have recorded storm induces damage to coral reefs at this depth and deeper (reviewed by Harmelin-Vivien 1994). The abstract of this manuscript implies that most literature to date has incorrectly characterized the susceptibility of deeper coral reefs to storm damage. The authors should reference the specific papers which have mischaracterized this issue to substantiate this implied generalization. The seminal reference on this topic (also referenced by the authors) clearly reviews these prior studies.\n\nfrom Harlmelin-Vivien 1994\nOn several Pacific reefs, particularly in Guam and Hawaii, shallow outer reef areas (0-10 m) suffer less than deeper areas (15-25 m), as corals growing there are well adjusted in size and shape to strong hydrodynamic conditions, and endure better an increase in wave energy (RANDALL and ELDREDGE, 1977; 0GG and KosLow, 1978; DoLLAR, 1982; PFEFFER and TRIBBLE, 1985). Cyclone-induced damage generally decreases with depth but bt coral destruction was observed down to a depth of 25 m in Belize (HIGHSMITH et al., 1980), to 30m on the Great Barrier Reef (WoESIK et al., 1991), Guam (RANDALL and ELDREDGE, 1977) and Hawaii (WALSH, 1983), to 50 m in Jamaica (WooDLEY et al., 1981) and Hawaii (PFEFFER and TRIBBLE, 1985), and to 90-100 m in French Polynesia (LABOUTE, 1985; HAR\u00ac MELIN-VIVIEN and LABOUTE, 1986).\nExperimental design\nno comments\nValidity of the findings\nno comments\nCite this review as\nKahng SE (2013) Peer Review #2 of \"Typhoon damage on a shallow mesophotic reef in Okinawa, Japan (v0.1)\". PeerJ https://doi.org/10.7287/peerj.151v0.1/reviews/2",
+    "pdf_1": "https://peerj.com/articles/151v0.2/submission",
+    "pdf_2": "https://peerj.com/articles/151v0.1/submission",
+    "all_reviews": "Review 1: Robert Toonen \u00b7 Aug 14, 2013 \u00b7 Academic Editor\nACCEPT\nThank you for clearly addressing the suggestions of the referees in your revisions. I am happy to say that I believe your manuscript is a valuable contribution to the field, and can now be accepted for publication. Keep up the good work.\nReview 2: Robert Toonen \u00b7 Jul 28, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nOverall both referees were very positive about your paper and recommend acceptance. I concur with them, and am pleased to be able to tell you that your paper is acceptable for publication pending minor revisions in response to the comments raised by the referees. In particular, I concur with the comment about the quality of the figures - I am not sure if it is an issue of the PDF conversion or not, but the figures were hard to read and Figure has grey/white in a & c, but b & d appear all grey with just text \"before\" and \"after\" - some color-coding here would really help the readability of the figure.\n\nReferee #2 suggests some additional references and asks exactly which literature to date has mischaracterized the susceptibility of deeper reefs to storm damage. I think this comment is easily dealt with by a slight revision of the text to something along the lines of it being intuitive that depth provides some buffer to storm damage or adding a reference such as Hughes & Connell (1999 L&0 44:932\u2013940) or Scoffin (1993 Coral Reefs 12:203-221) to support your statement that there is a perception that mesophotic reefs are less susceptible to storm damage. In addition to the Harmelin-Vivien (1994) review pointed out by the referee, a more recent missing reference seems to be Bongaerts et al. 2013 Coral Reefs (Cyclone damage at mesophotic depths on Myrmidon Reef).\n\nFinally, I also wanted to pass along the confidential comment from one suggested referee who declined to review the paper, because they are similar in tone to those of Reviewer #2. They said that it seemed to be a fine study, but they felt it was \"over-sold in an attempt to ride the mesophotic bandwagon and imply that the study is more novel than it really is.\" They felt that the work should stand on it's own and that the novelty pitch severly detracted from the paper and they would reject it based solely on that rather than the content. I tell you this because I believe that one of the benefits of review is to get feedback about how some readers view the work, so that you have the option to revise before final publication. I am fine with whatever you decide in this regard, but wanted to pass along the information so that you can consider toning it down if you choose prior to final publication.\n\nAloha,\nRob\nReview 3: Reviewer 1 \u00b7 Jul 25, 2013\nBasic reporting\nThis study is very interesting to show that corals at deeper depth are less resistant to the storm than we expected.\nExperimental design\nno problem\nValidity of the findings\ngood\nAdditional comments\nI think that Typhoon 16 also affected the deep coral communities. I wonder why the authors consider it.\n\nFigure 1: it is hard to see any words in this figure. Also should put \"a\" and \"b\" on the figure.\n\nFigure 4: need much more resoluiion. Also need to explain more how to see the data. I do not understand the meaning \"each circle displays relative abundance of the species\" from this figure.\nAnd what do you mean the number of coral rubble?\nAlso it is very complicated to see this because coral rubbles were shown by full circles, for both \"before\" and \"after\".\n\nMinor comments\n\"Montastraea\" is correct, not \"Montastrea\".\nAlso recently \"Montastraea\" was separated into three genera, Montastraea (for M. cavernosa), Phymastrea (for Indo-Pacific species), Orbicella (for M. annularis complex). If possible, consider it when the authors use \"Montastraea\".\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Typhoon damage on a shallow mesophotic reef in Okinawa, Japan (v0.1)\". PeerJ https://doi.org/10.7287/peerj.151v0.1/reviews/1\nReview 4: Samuel Kahng \u00b7 Jul 22, 2013\nBasic reporting\nThe authors present novel data measuring and analyzing coral community structure at two intermediate depths before and after a major storm/disturbance. Given the relatively depauperate nature of the primary literature on this topic, this data set provide valuable insight into the susceptibility of coral reef communities at intermediate depths. The experimental design and data analyses are sound, and the manuscript is well written. The fundamentals of this manuscript warrant publication, however, minor revisions are warranted in the opinion of this reviewer. Detailed comments have been included in the pdf using the comments feature.\n\nThe context of this study has been overstated and overgeneralized. The data for this study was collected at two similar depths (27 & 31m) boarding the upper mesophotic. To no fault of the authors, the upper depth limit of term \u201cmesophotic\u201d as coined by NOAA is poorly defined to correspond to the lower depth limits of recreational Scuba diving at 30-40m. This upper depth limit has no ecologically meaningful relevance to corals but pertains to a human (or socio-economic) limitation. The use of this data to generalize the susceptibility of mesophotic reefs (which can extend to depths >3X the depth of this study) to storm damage is a bit of a stretch. Understandably, while the authors attempted to collected data from a larger range of depths, the absence of data from shallower depths weakens their ability to generalize conclusions comparing the storm impact on \u201cshallow\u201d versus deep coral reefs.\n\nWhile this study is valuable and detailed, previous studies have recorded storm induces damage to coral reefs at this depth and deeper (reviewed by Harmelin-Vivien 1994). The abstract of this manuscript implies that most literature to date has incorrectly characterized the susceptibility of deeper coral reefs to storm damage. The authors should reference the specific papers which have mischaracterized this issue to substantiate this implied generalization. The seminal reference on this topic (also referenced by the authors) clearly reviews these prior studies.\n\nfrom Harlmelin-Vivien 1994\nOn several Pacific reefs, particularly in Guam and Hawaii, shallow outer reef areas (0-10 m) suffer less than deeper areas (15-25 m), as corals growing there are well adjusted in size and shape to strong hydrodynamic conditions, and endure better an increase in wave energy (RANDALL and ELDREDGE, 1977; 0GG and KosLow, 1978; DoLLAR, 1982; PFEFFER and TRIBBLE, 1985). Cyclone-induced damage generally decreases with depth but bt coral destruction was observed down to a depth of 25 m in Belize (HIGHSMITH et al., 1980), to 30m on the Great Barrier Reef (WoESIK et al., 1991), Guam (RANDALL and ELDREDGE, 1977) and Hawaii (WALSH, 1983), to 50 m in Jamaica (WooDLEY et al., 1981) and Hawaii (PFEFFER and TRIBBLE, 1985), and to 90-100 m in French Polynesia (LABOUTE, 1985; HAR\u00ac MELIN-VIVIEN and LABOUTE, 1986).\nExperimental design\nno comments\nValidity of the findings\nno comments\nCite this review as\nKahng SE (2013) Peer Review #2 of \"Typhoon damage on a shallow mesophotic reef in Okinawa, Japan (v0.1)\". PeerJ https://doi.org/10.7287/peerj.151v0.1/reviews/2\nReview 5: nan\nReview 6: \n",
+    "LLM_extracted_review": "['Review 1: Thank you for clearly addressing the suggestions of the referees in your revisions. I am happy to say that I believe your manuscript is a valuable contribution to the field, and can now be accepted for publication. Keep up the good work.\\n\\nReview 2: Overall both referees were very positive about your paper and recommend acceptance. I concur with them, and am pleased to be able to tell you that your paper is acceptable for publication pending minor revisions in response to the comments raised by the referees. In particular, I concur with the comment about the quality of the figures - I am not sure if it is an issue of the PDF conversion or not, but the figures were hard to read and Figure has grey/white in a & c, but b & d appear all grey with just text \"before\" and \"after\" - some color-coding here would really help the readability of the figure. Referee #2 suggests some additional references and asks exactly which literature to date has mischaracterized the susceptibility of deeper reefs to storm damage. I think this comment is easily dealt with by a slight revision of the text to something along the lines of it being intuitive that depth provides some buffer to storm damage or adding a reference such as Hughes & Connell (1999 L&0 44:932\u2013940) or Scoffin (1993 Coral Reefs 12:203-221) to support your statement that there is a perception that mesophotic reefs are less susceptible to storm damage. In addition to the Harmelin-Vivien (1994) review pointed out by the referee, a more recent missing reference seems to be Bongaerts et al. 2013 Coral Reefs (Cyclone damage at mesophotic depths on Myrmidon Reef). Finally, I also wanted to pass along the confidential comment from one suggested referee who declined to review the paper, because they are similar in tone to those of Reviewer #2. They said that it seemed to be a fine study, but they felt it was \"over-sold in an attempt to ride the mesophotic bandwagon and imply that the study is more novel than it really is.\" They felt that the work should stand on its own and that the novelty pitch severely detracted from the paper and they would reject it based solely on that rather than the content. I tell you this because I believe that one of the benefits of review is to get feedback about how some readers view the work, so that you have the option to revise before final publication. I am fine with whatever you decide in this regard, but wanted to pass along the information so that you can consider toning it down if you choose prior to final publication.\\n\\nReview 3: This study is very interesting to show that corals at deeper depth are less resistant to the storm than we expected. I think that Typhoon 16 also affected the deep coral communities. I wonder why the authors consider it. Figure 1: it is hard to see any words in this figure. Also should put \"a\" and \"b\" on the figure. Figure 4: need much more resolution. Also need to explain more how to see the data. I do not understand the meaning \"each circle displays relative abundance of the species\" from this figure. And what do you mean the number of coral rubble? Also it is very complicated to see this because coral rubbles were shown by full circles, for both \"before\" and \"after\". Minor comments: \"Montastraea\" is correct, not \"Montastrea\". Also recently \"Montastraea\" was separated into three genera, Montastraea (for M. cavernosa), Phymastrea (for Indo-Pacific species), Orbicella (for M. annularis complex). If possible, consider it when the authors use \"Montastraea\".\\n\\nReview 4: The authors present novel data measuring and analyzing coral community structure at two intermediate depths before and after a major storm/disturbance. Given the relatively depauperate nature of the primary literature on this topic, this data set provides valuable insight into the susceptibility of coral reef communities at intermediate depths. The experimental design and data analyses are sound, and the manuscript is well written. The fundamentals of this manuscript warrant publication, however, minor revisions are warranted in the opinion of this reviewer. The context of this study has been overstated and overgeneralized. The data for this study was collected at two similar depths (27 & 31m) boarding the upper mesophotic. To no fault of the authors, the upper depth limit of term \u201cmesophotic\u201d as coined by NOAA is poorly defined to correspond to the lower depth limits of recreational Scuba diving at 30-40m. This upper depth limit has no ecologically meaningful relevance to corals but pertains to a human (or socio-economic) limitation. The use of this data to generalize the susceptibility of mesophotic reefs (which can extend to depths >3X the depth of this study) to storm damage is a bit of a stretch. Understandably, while the authors attempted to collect data from a larger range of depths, the absence of data from shallower depths weakens their ability to generalize conclusions comparing the storm impact on \u201cshallow\u201d versus deep coral reefs. While this study is valuable and detailed, previous studies have recorded storm-induced damage to coral reefs at this depth and deeper (reviewed by Harmelin-Vivien 1994). The abstract of this manuscript implies that most literature to date has incorrectly characterized the susceptibility of deeper coral reefs to storm damage. The authors should reference the specific papers which have mischaracterized this issue to substantiate this implied generalization. \\n\\nReview 5: nan\\n\\nReview 6: ']"
+}

peerj_json_files/PeerJ_Json_108.json ADDED Viewed

	@@ -0,0 +1,17 @@

+{
+    "v1_Abstract": "Little is known about effects of large storm systems on mesophotic reefs. This study reports on how Typhoon 17 (Jelawat) affected Ryugu Reef on Okinawa-jima, Japan in September 2012. Benthic communities were surveyed before and after the typhoon using line intercept transect method. Comparison of the benthic assemblages showed highly significant differences in coral coverage at depths of 25-32 m before and after Typhoon 17. A large deep stand of Pachyseris foliosa was apparently less resistant to the storm than the shallower high diversity area of this reef. Contradictory to common perception, this research shows that large foliose corals at deeper depths are just as susceptible to typhoon damage as shallower branching corals. However, descriptive functional group analyses resulted in only minor changes after the disturbance, suggesting the high likelihood of recovery and the high resilience capacity of this mesophotic reef.",
+    "v1_col_introduction": "introduction  : Typhoon damage from direct physical disturbances, turbidity, sedimentation, and salinity\nchanges can be destructive to shallow coral reefs and has been well studied (Van Woesik, Ayling & Mapstone, 1991; Harmelin-Vivien, 1994; Ninio et al., 2000; Cheal et al., 2002; Hongo, Kawamata & Goto, 2012). Declines in coral cover on shallow reefs (<25 m in depth) has been documented, specifically in genera such as Acropora, Montastraea, Porites, Agaracia, Diploria, Millepora, Siderastrea, Pocillopora, Pachyseris, Montipora, and Merulina (Harmelin-Vivien, 1994; Van Woesik, De Vantier & Glazebrook, 1995; Fabricius et al., 2008). Although massive corals such as Montipora, Montastraea, Siderastrea, and Diploria can be overturned during typhoons, they are often the most resistant to storms and therefore tend to dominate or increase in cover after a disturbance (Harmelin-Vivien, 1994; Fabricius et al., 2008; Hongo, Kawamata & Goto, 2012). Increase in cover of genera Porites, Montipora, and Lobophyllia, unattached fungiids and even Acropora species with high regeneration efficiencies have been documented after disturbances (Harmelin-Vivien, 1994; Van Woesik , De Vantier & Glazebrook, 1995; Fabricius et al., 2008; Kuo et al., 2010).\nRecent observational data of a mesophotic reef (35\u201340 m) near Kume-jima, Okinawa,\nJapan before and after a typhoon reported that newly broken Acropora pieces fused to new branches, producing clones, and the reef made a quick recovery after typhoon damage (Fujita, Kimura, & Atsuo, 2012).\nHowever, comparatively little has been published on the effects of tropical cyclones on\ndeeper reefs (Randall & Eldredge, 1977; Woodley et al., 1981; Walsh, 1983; Pfeffer & Tribble, 1985; Harmelin-Vivien & Laboute, 1986; Van Woesik, Ayling & Mapstone, 1991). In general, reefs at depths greater than 25 meters appear to be less affected by tropical cyclones than shallower reefs (Harmelin-Vivien, 1994; Bongaerts et al., 2011; Bridge & Guinotte, 2012). Harmelin-Vivien (1994) reported that most physical damage to deep reefs was due to rolling\n16\n17\n18\n19\n20\n21\n22\n23\n24\n25\n26\n27\n28\n29\n30\n31\n32\n33\n34\n35\n36\n37\n38\n39\n40\nPeerJ reviewing PDF | (v2013:07:651:1:0:REVIEW 13 Aug 2013)\nR ev ie w in g M an\nus cr ip t\ncolonies dislodged from shallower areas. However, as reported in Harmelin-Vivien (1994), tropical cyclone induced coral destruction was observed to depths of 25 m in Belize (Highsmith, Riggs, & D\u2019Antonio,1980), to 30 m on the Great Barrier Reef (Van Woesik, Ayling & Mapstone, 1991), to 30 m in Guam and Hawaii (Walsh, 1983), to 50 m in Jamaica (Woodley et al., 1981), to 50 m in Hawaii (Pfeffer & Tribble, 1985), to 50-65 m on the Great Barrier Reef (Bongaerts et al., 2013), and to 90-100 m in French Polynesia (Laboute, 1985; Harmelin-Vivien & Laboute, 1986). Bridge and Guinotte (2012) concluded that although depth does have an impact on coral community survival, rugosity and angle of slope play large roles in the protection of species on a reef. Large communities of of broadcast spawning species are important to the recovery of reefs and are more likely to survive in deeper waters than in shallower waters during a typhoon (Madin & Connolly, 2006; Bridge et al., 2012), and recruitment of coral larvae is likely the most effective method of recovery for disturbed reefs (Harmelin-Vivien, 1994).\nTyphoon 17 (Jelawat) struck the west coast of Okinawa-jima Island on 29 September\n2012 (Fig. 1)., heading from the southwest to the northeast with a general wind direction of northwest. The Japan Meteorological Agency (JMA) and the United States Navy Joint Typhoon Warning Center (JTW) documented this record-breaking typhoon as the third strongest typhoon to hit Okinawa-jima Island since weather radar observations were started in 1954 with maximum wave heights of 12 m. The Okinawa Meteorological Observatory (OMO) recorded the following maximum measurements for Typhoon 17 at the northern end of Okinawa-jima Island (Nago Meteorological Station): 32.2 m/s sustained winds, 57.4 m/s wind gusts, 947.4 hPa atmospheric pressure.\nRecently, Ohara et al. (2013) reported on a previously undiscovered shallow mesophotic\ncoral reef in Okinawa, Japan. The Japanese name for this reef is \u201cRyugu,\u201d based on its resemblance to the undersea palace of Ry\u016bjin, the dragon god of the sea. The deeper sections of Ryugu (32\u201342 m) were reported to be primarily composed of Pachyseris foliosa Veron, 1990,\n41\n42\n43\n44\n45\n46\n47\n48\n49\n50\n51\n52\n53\n54\n55\n56\n57\n58\n59\n60\n61\n62\n63\n64\n65\nPeerJ reviewing PDF | (v2013:07:651:1:0:REVIEW 13 Aug 2013)\nR ev ie w in g M an\nus cr ip t\nwith shallower sections (25\u201332 m) showing much higher diversity (Ohara et al., 2013). Little is known about Pachyseris foliosa, although the depth range of this species appears to be deeper than previously reported by Hoeksema, Rogers & Quibilan in 2008 (25-30 m).\nThis study reports on how Typhoon 17 affected Ryugu Reef with before and after transect\ndata. Objectives of this study include examination of coral communities at different depths and the identification of species and functional groups most affected by typhoons. One hypothesis tested was that shallow mesophotic reefs with large monospecific stands are more resistant to storm damage.",
+    "v2_Abstract": "Little is known about effects of large storm systems on mesophotic reefs. This study reports on how Typhoon 17 (Jelawat) affected Ryugu Reef on Okinawa-jima, Japan in September 2012. Benthic communities were surveyed before and after the typhoon using line intercept transect method. Comparison of the benthic assemblages showed highly significant differences in coral coverage at depths of 25-32 m before and after Typhoon 17. A large deep stand of Pachyseris foliosa was apparently less resistant to the storm than the shallower high diversity area of this reef. Contradictory to most current literature, this research shows that large foliose corals at deeper depths are just as susceptible to typhoon damage as shallower branching corals. However, descriptive functional group analyses resulted in only minor changes after the disturbance, suggesting the high likelihood of recovery and the high resilience capacity of this mesophotic reef.",
+    "v2_col_introduction": "introduction  : Typhoon damage from direct physical disturbances, turbidity, sedimentation, and salinity\nchanges can be destructive to shallow coral reefs and has been well studied (Van Woesik, Ayling & Mapstone, 1991; Harmelin-Vivien, 1994; Ninio et al., 2000; Cheal et al., 2002; Hongo, Kawamata & Goto, 2012). Declines in coral cover on shallow reefs (<25 m in depth) has been documented, specifically in genera such as Acropora, Montastrea, Porites, Agaracia, Diploria, Millepora, Siderastrea, Pocillopora, Pachyseris, Montipora, and Merulina (Harmelin-Vivien, 1994; Van Woesik, De Vantier & Glazebrook, 1995; Fabricius et al., 2008). Although massive corals such as Montipora, Montastrea, Siderastrea, and Diploria can be overturned during\nPeerJ reviewing PDF | (v2013:07:651:0:0:NEW 16 Jul 2013)\nR ev ie w in g M an\nus cr ip t\ntyphoons, they are often the most resistant to storms and therefore tend to dominate or increase in cover after a disturbance (Harmelin-Vivien, 1994; Fabricius et al., 2008; Hongo, Kawamata & Goto, 2012). Increase in cover of genera Porites, Montipora, and Lobophyllia, unattached fungiids and even Acropora species with high regeneration efficiencies have been documented after disturbances (Harmelin-Vivien, 1994; Van Woesik , De Vantier & Glazebrook, 1995; Fabricius et al., 2008; Kuo et al., 2010).\nRecent observational data of a mesophotic reef (35\u201340 m) near Kume-jima, Okinawa,\nJapan before and after a typhoon reported that newly broken Acropora pieces fused to new branches, producing clones, and the reef made a quick recovery after typhoon damage (Fujita, Kimura, & Atsuo, 2012).\nHowever, comparatively little has been published on the effects of tropical cyclones on\ndeeper reefs (Randall & Eldredge, 1977; Woodley et al., 1981; Walsh, 1983; Pfeffer & Tribble, 1985; Harmelin-Vivien & Laboute, 1986; Van Woesik, Ayling & Mapstone, 1991). In general, reefs at depths greater than 25 meters appear to be less affected by tropical cyclones than shallower reefs (Harmelin-Vivien, 1994; Bongaerts et al., 2011; Bridge & Guinotte, 2012). Harmelin-Vivien (1994) reported that most physical damage to deep reefs was due to rolling colonies dislodged from shallower areas. Bridge and Guinotte (2012) concluded that although depth does have an impact on coral community survival, rugosity and angle of slope play large roles in the protection of species on a reef. Large communities of plating corals that are important to the recovery of broadcast spawning species are more likely to survive in deeper waters than in shallower waters during a typhoon (Madin & Connolly, 2006; Bridge et al., 2012), and recruitment of coral larvae is likely the most effective method of recovery for disturbed reefs (Harmelin-Vivien, 1994).\nPeerJ reviewing PDF | (v2013:07:651:0:0:NEW 16 Jul 2013)\nR ev ie w in g M an\nus cr ip t\nTyphoon 17 (Jelawat) struck the west coast of Okinawa-jima Island on 29 September\n2012 (Fig. 1)., heading from the southwest to the northeast with a general wind direction of northwest. The Japan Meteorological Agency (JMA) and the United States Navy Joint Typhoon Warning Center (JTW) documented this record-breaking typhoon as the third strongest typhoon to hit Okinawa-jima Island since weather radar observations were started in 1954 with maximum wave heights of 12 m. The Okinawa Meteorological Observatory (OMO) recorded the following maximum measurements for Typhoon 17 at the northern end of Okinawa-jima Island (Nago Meteorological Station): 32.2 m/s sustained winds, 57.4 m/s wind gusts, 947.4 hPa atmospheric pressure.\nRecently, Ohara et al. (2013) reported on a previously undiscovered shallow mesophotic\ncoral reef in Okinawa, Japan. The Japanese name for this reef is \u201cRyugu,\u201d based on its resemblance to the undersea palace of Ry\u016bjin, the dragon god of the sea. The deeper sections of Ryugu (32\u201342 m) were reported to be primarily composed of Pachyseris foliosa Veron, 1990, with shallower sections (25\u201332 m) showing much higher diversity (Ohara et al., 2013). Little is known about Pachyseris foliosa, although the depth range of this species appears to be deeper than previously reported by Hoeksema, Rogers & Quibilan in 2008 (25-30 m).\nThis study reports on how Typhoon 17 affected Ryugu Reef with before and after transect\ndata. Objectives of this study include examination of coral communities at different depths and the identification of species and functional groups most affected by typhoons. One hypothesis tested was that deeper reefs with large monospecific stands are more resistant to storm damage.",
+    "v1_text": "materials and methods : Five stations were designated at the Ryugu site (Fig 1). Station 1 was the deepest (42 m) and station 5 the shallowest (17 m). Stations are shown in Fig. 1 and summarized in Table 1. Temperature was recorded every 30 minutes using temperature loggers (HOBO U22 Temp Pro v2 logger, Onset Corp., Massachusetts, U.S.A.) placed approximately 30-50 cm from the substrate at each station from 12 September 2012 to 10 January 2013. Water motion was estimated at each station (except station 1) by the dissolution of plaster balls. Plaster balls (10.5 cm diameter) were made following Komatsu and Kawai (1992). The balls were set approximately 50 cm above the substrate at stations 2\u20135 on 11 January 2013 and removed on 16 January 2013. Water speeds for each station were calculated following the equations provided in Yokoyama, Inoue & Abo (2004). Ten meter line transects were surveyed both before (17 April 2012 and 11\u201312 September 2012) and after (14 December 2012) Typhoon 17 at locations near stations 2 (7 transects before, 9 transects after) and 3 (10 transects before, 8 transects after). Based on the amount of data available, only stations 2 and 3 were included in analyses. For each line transect, a 10 meter tape measure was laid out along a constant dedpth contour and overlapping photographs or video was 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 PeerJ reviewing PDF | (v2013:07:651:1:0:REVIEW 13 Aug 2013) R ev ie w in g M an us cr ip t taken along the line. Photographs or videos taken along the transects were used to report the total distance occupied by each operational taxonomic unit (OTU) identified. When feasible, OTUs were identified to species level following Hoeksema (1989) and Gittenberger, Reignen & Hoeksema (2011) for Fungiidae and Veron (2000) for other species. Community data were analyzed using PRIMER 6 statistical software in order to find differences in coral communities before and after the typhoon (Clarke and Warwick, 2001). All percent cover data were square-root transformed prior to analysis to moderately down-weight the importance of large space occupying operational taxonomic units (OTU). Bray-Curtis similarity matrices were calculated at stations 2 and 3. A one-way analysis of similarities test (ANOSIM) was performed to determine the difference and magnitude of difference in the assemblages before and after Typhoon 17. Non-Metric Multidimensional scaling (nMDS) was used to visualize multivariate patterns on the basis of the Bray-Curtis matrix. Bubble plot (square-root transformed cover data) was added to the plots to visualize variation in relevant OTUs. Each circle in figure 4 displays relative abundance of livecoral species (based on square root transformed data of the species' occurrence along each transect). Finally, the percentage contributions of each benthic grouping for observed differences between locations were assessed with the SIMPER routine. To assess potential changes in the functionality of the coral community after the typhoon coral taxa were classified into functional groups according to the shapes of the colonies following Denis et al. (2013). Each OTU was assigned to one or more of eight functional groups: massive, encrusting, foliose, columnar, plate-like, bushy, arborescent, and unattached (Supplementary Table 1). These were defined by each colony\u2019s growth form as described in Veron (2000), Wallace (1999) and by visual observation. Functional composition of the coral assemblages were calculated based on the relative abundance of coral OTU and plot for stations 2 and 3 before and after the typhoon. 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 PeerJ reviewing PDF | (v2013:07:651:1:0:REVIEW 13 Aug 2013) R ev ie w in g M an us cr ip t results : Temperature sensors showed that the temperature was typically 0.1\u20130.2 degrees lower at station 5 (shallowest station) than at any of the other stations, although the temperature appeared to fluctuate the most at this station. Temperature drops were observed at all stations during and directly after typhoons, with the largest decrease in temperature at station 1 (20.9\u00b0C after Typhoon 17) (Fig. 2). Based on plaster ball weight loss, station 2 had the lowest amount of water movement compared to the other stations. The weight loss of each plaster ball and water speed for each station were as follows: Station 2: lost 254 g, 9.2 cm/s; Station 3: lost 284 g, 11.2 cm/s; Station 4: lost 292 g, 11.7 cm/s; Station 5: lost 294 g, 12.5 cm/s. Supplementary table 1 lists all OTUs documented on transects and their percent cover change before and after Typhoon 17. Live coral cover decreased and coral rubble increased by 33.3% at station 2 and by 11.4% at station 3 after Typhoon 17. Figure 3 shows before and after images at stations 2 and 3. Composition of the benthic communities before and after Typhoon 17 (Fig. 4) presented a significant difference at both station 2 (ANOSIM test, R = 0.572, p = 0.001) and station 3 (ANOSIM test, R = 0.24, p = 0.009). At station 2, the change in the occurrence of coral rubble on the transects contributed the most to this difference (Simper-test, 33.0% - Fig. 4), followed by the coverage of Lithophyllon repanda (12.7%), Pachyseris foliosa (11.6%), then Galaxea sp. 1 (11.2%). At station 3, every OTU contributed to <10% of the difference observed. Change in the occurrence of coral rubble contributed to only 6% of this difference (Fig. 4). Interestingly, the difference observed at station 3 was not significant (R = 0.07, p = 0.137) when the effects of the dominant OTU were not reduced using square root transformation. Functionality of the coral communities (Fig. 5) at both stations seems only slightly affected by the typhoon. Among the major differences observed at station 2, the encrusting group decreased by 8%, while the foliose group increased by 15%. At station 3, bushy (8%), columnar (6%) and plate-like (3%) groups 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 PeerJ reviewing PDF | (v2013:07:651:1:0:REVIEW 13 Aug 2013) R ev ie w in g M an us cr ip t suffered the most from the path of the typhoon while encrusting (7%) and foliose (11%) corals were more resistant to this disturbance. acknowledgements : Thanks go to the boat captain, Tokunobu Toyama and diving staff, Sakiko Kawabata and Yoko Fudesaka, for their assistance during transect surveys. V. Denis is the recipient of a Post-Doctoral fellowship by the National Science Council of Taiwan. J.D. Reimer was funded by the Rising Star Program, and International Research Hub Project for Climate Change and Coral Reef/Island Dynamics, both at the University of the Ryukyus. 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 PeerJ reviewing PDF | (v2013:07:651:1:0:REVIEW 13 Aug 2013) R ev ie w in g M an us cr ip t discussion : Typhoon 17 resulted in highly significant changes to the live coral abundance at Ryugu stations 2 and 3. The depth of this reef does not appear to have sheltered corals from drastic damage with notable increases in coral rubble in many of the study areas. Most interestingly, P. foliosa was among those species most affected by Typhoon 17. Apparently more diverse and complex communities, such as at station 3, are more resistant to typhoons in terms of survivability and functional group distribution, perhaps due to the lower impact on individual species. Based on SIMPER tests, there were many small differences in diverse OTUs at station 3 compared to station 2, where only 4 OTUs contributed to 70% of the typhoon effects. Station 2, primarily composed of P. foliosa, was heavily impacted by this storm, despite the fact that it was deeper than station 3, suggesting that the foliose structure of P. foliosa is vulnerable to physical disturbances. The large monospecific stand found here is also in a more stable environment, likely making it more sensitive to disturbance (Hughes, 1989; Rogers, 1992, 1993; HarmelinVivien, 1994). Therefore, our hypothesis that shallow mesophotic reefs with large monospecific stands are more resistant to storm damage is rejected. Consistent with previous shallow water typhoon damage studies (Harmelin-Vivien, 1994; Van Woesik, De Vantier & Glazebrook, 1995; Fabricius et al., 2008; Kuo et al., 2010) corals in the genus Acropora were strongly affected and were mostly dead at Ryugu after Typhoon 17 whereas unattached Fungiidae corals were mostly healthy. The fungiid corals may have been hidden under other living corals and after Typhoon 17 became more visible with the other corals having been damaged. Away from the transect locations; however, several Fungiidae corals were completely buried by newly generated Acropora rubble and other branching coral 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 PeerJ reviewing PDF | (v2013:07:651:1:0:REVIEW 13 Aug 2013) R ev ie w in g M an us cr ip t rubble. Accordingly, despite minor changes in the functionality of the coral community observed before and after Typhoon 17, the groups the most affected at station 3 were the bushy, columnar and plate-like corals. At station 2, the dominance of P. foliosa may have masked any differences. Most damaged colonies of P. foliosa were still alive after Typhoon 17, suggesting that species composition of this area may not change. This opposes the common idea that only massive corals would remain after disturbance (Harmelin-Vivien, 1994), suggesting a strong potential for the recovery and resilience of Ryugu Reef. Pachyseris species are typically gonochoric spawners that are most likely unable to fuse and create clones (Richmond and Hunter, 1990). If Pachyseris species reproduce only by spawning, Pachyseris-dominated reefs such as Ryugu should have a much slower recovery rate than Acropora-dominated reefs such as the one found near Kumejima. However, recruitment of coral larvae may allow this reef to recover relatively quickly. During this study, new P. foliosa polyps were observed growing two to three months after Typhoon 17, suggesting that the Pachyseris portion of the reef had already started to recover from the damage it incurred. Based on plaster ball data, Ryugu is a fairly calm reef and the lower currents at station 2 may be due to less tides or wave impacts, and is worth investigating further in future studies. The lower temperature observed during Typhoon 17 at station 1 (42 m) may be due to upwelling or thermal averaging due to wind driven vertical mixing with deeper cooler water that was enhanced by the onset of the typhoon, as seen during other large storms. Figure 2 shows large changes in temperature on 18 September (drop to 26.0\u00b0C) and 30 September (drop to 20.9\u00b0C), both of which correspond with large typhoon systems (Fig. 1; Typhoons 16 and 17, respectively). Many studies have found that increasing sea surface temperatures and global climate change have and will continue to cause increases in typhoon frequency, power dissipation, and storm intensity (Emanuel, 2005; Trenberth, 2005; Webster et al., 2005; Emanuel, Sundararajan & Williams, 2008; Tu et al., 2009). Tu et al. (2009) have documented a northward shift in typhoon 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 PeerJ reviewing PDF | (v2013:07:651:1:0:REVIEW 13 Aug 2013) R ev ie w in g M an us cr ip t tracks in the western North Pacific-East Asia region with an increase in typhoon frequency in the Taiwan/East China Sea region (3.3 per year from 1970-1999; 5.7 per year from 2000-2006). Emanuel (2005) documented an increase in destructiveness of cyclones since the 1970s and has predicted a continued increase with global climate change. Global climate change is expected to bring larger and stronger typhoons to Okinawa, which will likely affect the survivability of some coral populations. A potential increase in storms makes it even more important to understand their effect on mesophotic reefs, which have been thought to act as refugia for many marine organisms during disturbances on shallow reefs. This study has shown that despite their depth, shallow mesophotic reefs may also be strongly affected by disturbances. It is, therefore, critical to document the succession of this reef after disturbances such as Typhoon 17 to understand its resilience and the role that mesophotic reefs may play in the future of coral reefs. 2013; station 3, c) 12 september 2012, d) 01 january 2013 : PeerJ reviewing PDF | (v2013:07:651:1:0:REVIEW 13 Aug 2013) R ev ie w in g M an us cr ip t Figure 4 Non-metric dimensional scaling of the benthic communities at Ryugu Reef Non-metric dimensional scaling of the benthic communities at Ryugu Reef based on the Bray-Curtis similarities matrices. Each circle in the bubble plot displays relative abundance of live coral or percent cover of coral rubble (based on square root transformed data of the species' occurrence along each transect) a) Station 2: live coral species, b) Station 2: coral rubble, c) Station 3: live coral species, d) Station 3: coral rubble. Red circle: before typhoon, blue circle: after typhoon. PeerJ reviewing PDF | (v2013:07:651:1:0:REVIEW 13 Aug 2013) R ev ie w in g M an us cr ip t Figure 5 Functional composition of the coral assemblage before and after Typhoon 17 at Ryugu Reef Functional composition of the coral assemblage before and after Typhoon 17 at Ryugu Reef, Stations 2 and 3 based on relative abundance of the coral OTUs. Axes represent the relative contribution of each of the 8 functional groups. PeerJ reviewing PDF | (v2013:07:651:1:0:REVIEW 13 Aug 2013) R ev ie w in g M an us cr ip t Table 1(on next page) stations at ryugu reef : PeerJ reviewing PDF | (v2013:07:651:1:0:REVIEW 13 Aug 2013) R ev ie w in g M an us cr ip t Station Location/Description Depth (m) 1 Outside outer edge of dense Pachyseris foliosa area, sandy 42 2 Upper edge of dense Pachyseris foliosa area 31.2 3 Upper edge of high diversity area 26.5 4 Upper edge of Fungiidae/rubble area 21.3 5 Sand, coral rubble 17 PeerJ reviewing PDF | (v2013:07:651:1:0:REVIEW 13 Aug 2013) R ev ie w in g M an us cr ip t",
+    "v2_text": "abstract : Little is known about effects of large storm systems on mesophotic reefs. This study reports on how Typhoon 17 (Jelawat) affected Ryugu Reef on Okinawa-jima, Japan in September 2012. Benthic communities were surveyed before and after the typhoon using line intercept transect method. Comparison of the benthic assemblages showed highly significant differences in coral coverage at depths of 25\u201332 m before and after Typhoon 17. A large deep stand of Pachyseris foliosa was apparently less resistant to the storm than the shallower high diversity area of this reef. Contradictory to most current literature, this research shows that large foliose corals at deeper depths are just as susceptible to typhoon damage as shallower branching corals. However, descriptive functional group analyses resulted in only minor changes after the disturbance, suggesting the high likelihood of recovery and the high resilience capacity of this mesophotic reef. materials and methods : PeerJ reviewing PDF | (v2013:07:651:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t Five stations were designated at the Ryugu site (Fig 1). Station 1 was the deepest (42 m) and station 5 the shallowest (17 m). Stations are shown in Fig. 1 and summarized in Table 1. Temperature was recorded every 30 minutes using temperature loggers (HOBO U22 Temp Pro v2 logger, Onset Corp., Massachusetts, U.S.A.) placed approximately 30-50 cm from the substrate at each station from 12 September 2012 to 10 January 2013. Water motion was estimated at each station (except station 1) by the dissolution of plaster balls. Plaster balls (10.5 cm diameter) were made following Komatsu and Kawai (1992). The balls were set approximately 50 cm above the substrate at stations 2\u20135 on 11 January 2013 and removed on 16 January 2013. Water speeds for each station were calculated following the equations provided in Yokoyama, Inoue & Abo (2004). Ten meter line transects were surveyed both before (17 April 2012 and 11\u201312 September 2012) and after (14 December 2012) Typhoon 17 at locations near stations 2 (7 transects before, 9 transects after) and 3 (10 transects before, 8 transects after). Based on the amount of data available, only stations 2 and 3 were included in analyses. For each line transect, a 10 meter tape measure was laid out and overlapping photographs or video was taken along the line. Photographs or videos taken along the transects were used to report the total distance occupied by each operational taxonomic unit (OTU) identified. When feasible, OTUs were identified to species level following Hoeksema (1989) and Gittenberger, Reignen & Hoeksema (2011) for Fungiidae and Veron (2000) for other species. Community data were analyzed using PRIMER 6 statistical software in order to find differences in coral communities before and after the typhoon (Clarke and Warwick, 2001). All percent cover data were square-root transformed prior to analysis to moderately down-weight the importance of large space occupying operational taxonomic units (OTU). Bray-Curtis similarity PeerJ reviewing PDF | (v2013:07:651:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t matrices were calculated at stations 2 and 3. A one-way analysis of similarities test (ANOSIM) was performed to determine the difference and magnitude of difference in the assemblages before and after Typhoon 17. Non-Metric Multidimensional scaling (nMDS) was used to visualize multivariate patterns on the basis of the Bray-Curtis matrix. Bubble plot (square-root transformed cover data) was added to the plots to visualize variation in relevant OTUs. Each circle in figure 4 displays relative abundance of the species (based on square root transformed data of the species' occurrence along each transect). Finally, the percentage contributions of each benthic grouping for observed differences between locations were assessed with the SIMPER routine. To assess potential changes in the functionality of the coral community after the typhoon coral taxa were classified into functional groups according to the shapes of the colonies following Denis et al. (2013). Each OTU was assigned to one or more of eight functional groups: massive, encrusting, foliose, columnar, plate-like, bushy, arborescent, and unattached (Supplementary Table 1). These were defined by each colony\u2019s growth form as described in Veron (2000), Wallace (1999) and by visual observation. Functional composition of the coral assemblages were calculated based on the relative abundance of coral OTU and plot for stations 2 and 3 before and after the typhoon. results : Temperature sensors showed that the temperature was typically 0.1\u20130.2 degrees lower at station 5 (shallowest station) than at any of the other stations, although the temperature appeared to fluctuate the most at this station. Temperature drops were observed at all stations during and directly after typhoons, with the largest decrease in temperature at station 1 (20.9\u00b0C after PeerJ reviewing PDF | (v2013:07:651:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t Typhoon 17) (Fig. 2). Based on plaster ball weight loss, station 2 had the lowest amount of water movement compared to the other stations. The weight loss of each plaster ball and water speed for each station were as follows: Station 2: lost 254 g, 9.2 cm/s; Station 3: lost 284 g, 11.2 cm/s; Station 4: lost 292 g, 11.7 cm/s; Station 5: lost 294 g, 12.5 cm/s. Supplementary table 1 lists all OTUs documented on transects and their percent cover change before and after Typhoon 17. Coral rubble increased by 33.3% at station 2 and by 11.4% at station 3 after Typhoon 17. Figure 3 shows before and after images at stations 2 and 3. Composition of the benthic communities before and after Typhoon 17 (Fig. 4) presented a significant difference at both station 2 (ANOSIM test, R = 0.572, p = 0.001) and station 3 (ANOSIM test, R = 0.24, p = 0.009). At station 2, the change in the occurrence of coral rubble on the transects contributed the most to this difference (Simper-test, 33.0% - Fig. 4), followed by the coverage of Lithophyllon repanda (12.7%), Pachyseris foliosa (11.6%), then Galaxea sp. 1 (11.2%). At station 3, every OTU contributed to <10% of the difference observed. Change in the occurrence of coral rubble contributed to only 6% of this difference (Fig. 4). Interestingly, the difference observed at station 3 was not significant (R = 0.07, p = 0.137) when the effects of the dominant OTU were not reduced using square root transformation. Functionality of the coral communities (Fig. 5) at both stations seems only slightly affected by the typhoon. Among the major differences observed at station 2, the encrusting group decreased by 8%, while the foliose group increased by 15%. At station 3, bushy (8%), columnar (6%) and plate-like (3%) groups suffered the most from the path of the typhoon while encrusting (7%) and foliose (11%) corals were more resistant to this disturbance. figure legends : Figure 1. a)Map showing track of Typhoon 17 (Jelawat) around Okianwa-jima Island with position and b) studied stations of Ryugu Reef. Data on the track indicates date, time, central atmospheric pressure and maximum wind speed at each \"X\" mark. In b), dotted areas indicate positions of transects. Figure 2. Graph showing daily temperature change for Ryugu Reef, Stations 1\u20135, 12 September 2012 to 10 January 2013. Figure 3. Ryugu Reef photographs before and after Typhoon 17. Station 2, a) 12 September 2012, b) 01 January 2013; Station 3, c) 12 September 2012, d) 01 January 2013. Figure 4. Non-metric dimensional scaling of the benthic communities at Ryugu Reef based on the Bray-Curtis similarities matrices. Each circle displays relative abundance of the species (based on square root transformed data of the species' occurrence along each transect) a) Station 2, b) Station 2: bubble plot of coral rubble, c) Station 3, d) Station 3: bubble plot of coral rubble. Empty circle: before typhoon, full circle: after typhoon. Figure 5. Functional composition of the coral assemblage before and after Typhoon 17 at Ryugu Reef, Stations 2 and 3 based on relative abundance of the coral OTUs. Axes represent the relative contribution of each of the 8 functional groups. PeerJ reviewing PDF | (v2013:07:651:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t Figure 1 Map of typhoon track and Ryugu Reef stations a) Map showing track of Typhoon 17(Jelawat) around Okinawa-jima Island with position and b) studied stations of Ryugu Reef. Data on the track indicate date, time, central atmospheric pressure and maximum wind speed at each \"X\" mark. In b), dotted areas indicate positions of transects. PeerJ reviewing PDF | (v2013:07:651:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t Table 1(on next page) acknowledgements : Thanks go to the boat captain, Tokunobu Toyama and diving staff, Sakiko Kawabata and Yoko Fudesaka, for their assistance during transect surveys. V. Denis is the recipient of a Post-Doctoral fellowship by the National Science Council of Taiwan. J.D. Reimer was funded by the Rising Star Program, and International Research Hub Project for Climate Change and Coral Reef/Island Dynamics, both at the University of the Ryukyus. PeerJ reviewing PDF | (v2013:07:651:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t discussion : PeerJ reviewing PDF | (v2013:07:651:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t Typhoon 17 resulted in highly significant changes to the coral communities at Ryugu stations 2 and 3. The depth of this reef does not appear to have sheltered corals from drastic damage with notable increases in coral rubble in many of the study areas. Most interestingly, P. foliosa was among those species most affected by Typhoon 17. It appears that more diverse and complex communities, such as at station 3, are more resistant to typhoons, perhaps due to the lower impact on individual species. Based on SIMPER tests, there were many small differences in diverse OTUs at station 3 compared to station 2, where only 4 OTUs contributed to 70% of the typhoon effects. Station 2, primarily composed of P. foliosa, was heavily impacted by this storm, despite the fact that it was deeper than station 3, suggesting that the foliose structure of P. foliosa is vulnerable to physical disturbances. The large monospecific stand found here is also in a more stable environment, likely making it more sensitive to disturbance (Hughes, 1989; Rogers, 1992, 1993; Harmelin-Vivien, 1994). Therefore, our hypothesis that deeper reefs with large monospecific stands are more resistant to storm damage is rejected. Consistent with previous shallow water typhoon damage studies (Harmelin-Vivien, 1994; Van Woesik, De Vantier & Glazebrook, 1995; Fabricius et al., 2008; Kuo et al., 2010) corals in the genus Acropora were strongly affected and were mostly dead at Ryugu after Typhoon 17 whereas unattached Fungiidae corals were mostly healthy. The fungiid corals may have been hidden under other living corals and after Typhoon 17 became more visible with the other corals having been damaged. Away from the transect locations; however, several Fungiidae corals were completely buried by newly generated Acropora rubble and other branching coral rubble. Accordingly, despite minor changes in the functionality of the coral community observed before and after Typhoon 17, the groups the most affected at station 3 were the bushy, columnar and plate-like corals. At station 2, the dominance of P. foliosa may have masked any differences. PeerJ reviewing PDF | (v2013:07:651:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t Most damaged colonies of P. foliosa were still alive after Typhoon 17, suggesting that species composition of this area may not change. This opposes the common idea that only massive corals would remain after disturbance (Harmelin-Vivien, 1994), suggesting a strong potential for the recovery and resilience of Ryugu Reef. Pachyseris species are typically gonochoric spawners (Richmond and Hunter, 1990), suggesting that they are most likely unable to fuse and create clones. If this is the case, Pachyseris-dominated reefs such as Ryugu should have a much slower recovery rate than Acropora-dominated reefs such as the one found near Kume-jima. However, recruitment of coral larvae may allow this reef to recover relatively quickly. During this study, new P. foliosa polyps were observed growing two to three months after Typhoon 17, suggesting that the Pachyseris portion of the reef had already started to recover from the damage it incurred. Based on plaster ball data, Ryugu is a fairly calm reef and the lower currents at station 2 may be due to less tides or wave impacts, and is worth investigating further in future studies. It is likely that the lower temperature observed during Typhoon 17 at station 1 (42 m) is due to upwelling or a thermocline that was enhanced by the onset of the typhoon, as seen during other large storms. Figure 2 shows large changes in temperature on 18 September (drop to 26.0\u00b0C) and 30 September (drop to 20.9\u00b0C), both of which correspond with large typhoon systems (Fig. 1; Typhoons 16 and 17, respectively). Many studies have found that increasing sea surface temperatures and global climate change have and will continue to result in increases in typhoon frequency, power dissipation, and storm intensity (Emanuel, 2005; Trenberth, 2005; Webster et al., 2005; Emanuel, Sundararajan & Williams, 2008; Tu et al., 2009). Tu et al. (2009) have documented a northward shift in typhoon tracks in the western North Pacific-East Asia region with an increase in typhoon frequency in the Taiwan/East China Sea region (3.3 per year from 1970-1999; 5.7 per year from 2000-2006). PeerJ reviewing PDF | (v2013:07:651:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t Emanuel (2005) documented an increase in destructiveness of cyclones since the 1970s and has predicted a continued increase with global climate change. With global climate change expected to bring larger and stronger typhoons to Okinawa, and the likelihood of these storms affecting the survivability of some coral populations, it is even more important to understand their effect on mesophotic reefs, which have been thought to act as refugia for many marine organisms during disturbances on shallow reefs. This study has shown that despite their depth, these deeper reefs may also be strongly affected by disturbances. It is, therefore, critical to document the succession of this mesophotic reef after disturbances such as Typhoon 17 to understand its resilience and the role that mesophotic reefs may play in the future of coral reefs. stations at ryugu reef : PeerJ reviewing PDF | (v2013:07:651:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t Station Location/Description Depth (m) 1 Outside outer edge of dense Pachyseris foliosa area, sandy 42 2 Upper edge of dense Pachyseris foliosa area 31.2 3 Upper edge of high diversity area 26.5 4 Upper edge of Fungiidae/rubble area 21.3 5 Sand, coral rubble 17 PeerJ reviewing PDF | (v2013:07:651:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t Figure 2 temperature change graph : Graph showing daily temperature change for Ryugu Reef, Stations 1-5, 12 September 2012 to 10 January 2013. PeerJ reviewing PDF | (v2013:07:651:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t Figure 3 Ryugu Reef photographs before and after Typhoon 17. Station 2, a) 12 September 2012, b) 01 January 2013; Station 3, c) 12 September 2012, d) 01 January 2013 PeerJ reviewing PDF | (v2013:07:651:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t Figure 4 Non-metric dimensional scaling of benthic communities at Ryugu Reef Non-metric dimensional scaling of the benthic communities at Ryugu Reef based on the Bray-Curtis similarities matrices. Each circle displays relative abundance of the species a) Station 2, bubble plot of live coral; b) Station 2, bubble plot of coral rubble; c) Station 3, bubble plot of live coral; d)Station 3, bubble plot of coral rubble. Empty circle: before typhoon, full circle: after typhoon PeerJ reviewing PDF | (v2013:07:651:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t Figure 5 Functional composition of coral assemblage before and after typhoon Functional composition of the coral assemblage before and after Typhoon 17 at Ryugu Reef, Stations 2 and 3 based on relative abundance of the coral OYUs. Axes represent the relative contribution of each of the 8 functional groups PeerJ reviewing PDF | (v2013:07:651:0:0:NEW 16 Jul 2013) R ev ie w in g M an us cr ip t",
+    "url": "https://peerj.com/articles/152/reviews/",
+    "review_1": "Howard Young \u00b7 Aug 14, 2013 \u00b7 Academic Editor\nACCEPT\nYou changes were accepted and the manuscript is now ready for publication. Congratulations.",
+    "review_2": "Howard Young \u00b7 Jun 27, 2013 \u00b7 Academic Editor\nMAJOR REVISIONS\nDear Authors,\n\nI believe that your paper is worthy of publication but please try to especially address the concerns of the first reviewer. I am not concerned about comment 3 but I do believe that a response to the first 2 concerns is needed.",
+    "review_3": "Balamurugan Kuppusamy \u00b7 Jun 26, 2013\nBasic reporting\nSee below\nExperimental design\nSee below\nValidity of the findings\nSee below\nAdditional comments\nThe manuscript by Baysal et al studies the effect of hypoxia on C to U RNA editing of SDHB gene in monocytes. Previous studies show that SDH is a heterotetrameric tumor suppressor complex mutated in paraganglioma tumors due to hypoxia-inducible activation pathways. The authors studied regulation, extent and cell type origin of SDHB RNA editing. The authors found that RNA editing is low in monocyte-enriched PBMCs under normal condition, however, markedly increased under hypoxic conditions. This leads to down regulation of SDHB gene and consequently an increase in monocyte cell survival. Overall these findings add the mechanistic details to our understanding of the RNA editing under physiologically relevant conditions such as hypoxia, however additional data would support and strengthen their conclusions.\n\nMajor concerns:\n\n1. The authors show that hypoxia, one of the physiological conditions, increased the C136U RNA editing of SDHB transcripts in CD14+ monocytes.\n- Are HIF-factors (HIF-1alpha/HIF-2alpha) involved in this process?\n- Does the knockdown of these would rescue the change in the C136U RNA editing of SDHB transcripts in CD14+ monocytes?\n- Is the RNA editing pattern changed when these CD14+ monocytes are treated with inflammatory conditions (for example LPS) which is also one of the microenvironmental conditions.\n-\n2. On page 22, the authors state that \u201cthe suppression of SDH might be an evolutionarily conserved metabolic adaptation to hypoxia\u2026.. due to enhanced glycolysis\u201d. Though the data suggest that the C136U RNA editing is increased which should supposedly decrease the express, however, SDHB protein levels were unchanged under hypoxic conditions compared to normoxia (Figure 4D). I do not think that changes in the RNA editing of this gene is sufficient to conclude that SDHB expression is downregulated leading to enhanced glycolysis.\n\n3. The authors showed RNA editing pattern in peripheral blood mononuclear cells (lymphocytes and monocyte-enriched population). Did the authors by chance look at the status of RNA editing in polymorphonuclear cells, especially neutrophils, which are also part of the immune system and have key role(s) in the immunological processes.\n\nMinor issues:\n1. Provide p values for the figures: 1C-E, 2C and 4A, B and E\n2. Figure 4 legends, line 8, change C13U into C136U RNA editing\nCite this review as\nKuppusamy B (2013) Peer Review #1 of \"Hypoxia-inducible C-to-U coding RNA editing downregulates SDHB in monocytes (v0.1)\". PeerJ https://doi.org/10.7287/peerj.152v0.1/reviews/1",
+    "review_4": "Jianyun Liu \u00b7 Jun 24, 2013\nBasic reporting\n1. In line 322, \u201c>= fold change\u201d, a \u201c3\u201d is missing.\n2. In Fig.1A, 9 lymphoblastoid cell lines and one fibroblastic cell line were included, but the authors did not provide names and sources of these cell lines.\nExperimental design\nNo Comments\nValidity of the findings\nNo Comments\nAdditional comments\nIn this manuscript, Baysal investigated RNA editing of SDHB (C136U) in different cell types and human tissues using both bioinformatics tools and wet lab tools. They provided massive amount of data to support their hypothesis that hypoxia induces RNA editing of SDHB and downregulates its expression in monocytes. The authors have performed many hypothesis-driven and well-designed experiments. Overall, this manuscript is well-written and the results are properly interpreted with some speculations. Two minor points are listed below:\n1. In line 322, \u201c>= fold change\u201d, a \u201c3\u201d is missing.\n2. In Fig.1A, 9 lymphoblastoid cell lines and one fibroblastic cell line were included, but the authors did not provide names and sources of these cell lines.\nCite this review as\nLiu J (2013) Peer Review #2 of \"Hypoxia-inducible C-to-U coding RNA editing downregulates SDHB in monocytes (v0.1)\". PeerJ https://doi.org/10.7287/peerj.152v0.1/reviews/2",
+    "pdf_1": "https://peerj.com/articles/152v0.2/submission",
+    "pdf_2": "https://peerj.com/articles/152v0.1/submission",
+    "all_reviews": "Review 1: Howard Young \u00b7 Aug 14, 2013 \u00b7 Academic Editor\nACCEPT\nYou changes were accepted and the manuscript is now ready for publication. Congratulations.\nReview 2: Howard Young \u00b7 Jun 27, 2013 \u00b7 Academic Editor\nMAJOR REVISIONS\nDear Authors,\n\nI believe that your paper is worthy of publication but please try to especially address the concerns of the first reviewer. I am not concerned about comment 3 but I do believe that a response to the first 2 concerns is needed.\nReview 3: Balamurugan Kuppusamy \u00b7 Jun 26, 2013\nBasic reporting\nSee below\nExperimental design\nSee below\nValidity of the findings\nSee below\nAdditional comments\nThe manuscript by Baysal et al studies the effect of hypoxia on C to U RNA editing of SDHB gene in monocytes. Previous studies show that SDH is a heterotetrameric tumor suppressor complex mutated in paraganglioma tumors due to hypoxia-inducible activation pathways. The authors studied regulation, extent and cell type origin of SDHB RNA editing. The authors found that RNA editing is low in monocyte-enriched PBMCs under normal condition, however, markedly increased under hypoxic conditions. This leads to down regulation of SDHB gene and consequently an increase in monocyte cell survival. Overall these findings add the mechanistic details to our understanding of the RNA editing under physiologically relevant conditions such as hypoxia, however additional data would support and strengthen their conclusions.\n\nMajor concerns:\n\n1. The authors show that hypoxia, one of the physiological conditions, increased the C136U RNA editing of SDHB transcripts in CD14+ monocytes.\n- Are HIF-factors (HIF-1alpha/HIF-2alpha) involved in this process?\n- Does the knockdown of these would rescue the change in the C136U RNA editing of SDHB transcripts in CD14+ monocytes?\n- Is the RNA editing pattern changed when these CD14+ monocytes are treated with inflammatory conditions (for example LPS) which is also one of the microenvironmental conditions.\n-\n2. On page 22, the authors state that \u201cthe suppression of SDH might be an evolutionarily conserved metabolic adaptation to hypoxia\u2026.. due to enhanced glycolysis\u201d. Though the data suggest that the C136U RNA editing is increased which should supposedly decrease the express, however, SDHB protein levels were unchanged under hypoxic conditions compared to normoxia (Figure 4D). I do not think that changes in the RNA editing of this gene is sufficient to conclude that SDHB expression is downregulated leading to enhanced glycolysis.\n\n3. The authors showed RNA editing pattern in peripheral blood mononuclear cells (lymphocytes and monocyte-enriched population). Did the authors by chance look at the status of RNA editing in polymorphonuclear cells, especially neutrophils, which are also part of the immune system and have key role(s) in the immunological processes.\n\nMinor issues:\n1. Provide p values for the figures: 1C-E, 2C and 4A, B and E\n2. Figure 4 legends, line 8, change C13U into C136U RNA editing\nCite this review as\nKuppusamy B (2013) Peer Review #1 of \"Hypoxia-inducible C-to-U coding RNA editing downregulates SDHB in monocytes (v0.1)\". PeerJ https://doi.org/10.7287/peerj.152v0.1/reviews/1\nReview 4: Jianyun Liu \u00b7 Jun 24, 2013\nBasic reporting\n1. In line 322, \u201c>= fold change\u201d, a \u201c3\u201d is missing.\n2. In Fig.1A, 9 lymphoblastoid cell lines and one fibroblastic cell line were included, but the authors did not provide names and sources of these cell lines.\nExperimental design\nNo Comments\nValidity of the findings\nNo Comments\nAdditional comments\nIn this manuscript, Baysal investigated RNA editing of SDHB (C136U) in different cell types and human tissues using both bioinformatics tools and wet lab tools. They provided massive amount of data to support their hypothesis that hypoxia induces RNA editing of SDHB and downregulates its expression in monocytes. The authors have performed many hypothesis-driven and well-designed experiments. Overall, this manuscript is well-written and the results are properly interpreted with some speculations. Two minor points are listed below:\n1. In line 322, \u201c>= fold change\u201d, a \u201c3\u201d is missing.\n2. In Fig.1A, 9 lymphoblastoid cell lines and one fibroblastic cell line were included, but the authors did not provide names and sources of these cell lines.\nCite this review as\nLiu J (2013) Peer Review #2 of \"Hypoxia-inducible C-to-U coding RNA editing downregulates SDHB in monocytes (v0.1)\". PeerJ https://doi.org/10.7287/peerj.152v0.1/reviews/2\nReview 5: nan\nReview 6: \n",
+    "LLM_extracted_review": "['Review 1: You changes were accepted and the manuscript is now ready for publication. Congratulations.\\n\\nReview 2: I believe that your paper is worthy of publication but please try to especially address the concerns of the first reviewer. I am not concerned about comment 3 but I do believe that a response to the first 2 concerns is needed.\\n\\nReview 3: The authors found that RNA editing is low in monocyte-enriched PBMCs under normal condition, however, markedly increased under hypoxic conditions. This leads to down regulation of SDHB gene and consequently an increase in monocyte cell survival. Overall these findings add the mechanistic details to our understanding of the RNA editing under physiologically relevant conditions such as hypoxia, however additional data would support and strengthen their conclusions. Major concerns include the involvement of HIF-factors in the RNA editing process, the sufficiency of RNA editing changes to conclude downregulation of SDHB expression, and whether RNA editing status was examined in polymorphonuclear cells. Minor issues include the need for p values in certain figures and a correction in the figure legend.\\n\\nReview 4: The authors have performed many hypothesis-driven and well-designed experiments. Overall, this manuscript is well-written and the results are properly interpreted with some speculations. Minor points include a missing \"3\" in line 322 regarding fold change and the lack of names and sources for the included cell lines in Fig.1A.']"
+}

peerj_json_files/PeerJ_Json_109.json ADDED Viewed

The diff for this file is too large to render. See raw diff

peerj_json_files/PeerJ_Json_11.json ADDED Viewed

	@@ -0,0 +1,17 @@

+{
+    "v1_col_introduction": "introduction  : Magic is one of the oldest art forms, and magicians have manipulated audiences' perceptual and cognitive processes for much longer than cognitive scientists have (Martinez-Conde & Macknik 2007; Martinez-Conde & Macknik 2008; Macknik et al. 2008), Thus, classic and contemporary magic illusions provide scientists with methodological refinements and testable hypotheses about the building blocks of conscious experience (Cui et al. 2011; Otero-Millan et al. 2011). The \u201cCups and Balls\u201d is a sleight-of-hand magic trick that was performed by Roman conjurers as far back as two thousand years ago (Christopher & Christopher 2006). The trick has many variations, but the most common one uses three balls and three cups. The magician makes the balls pass through the bottom of cups, jump from cup to cup, disappear from a cup and turn up elsewhere, turn into other objects, and so on. The cups are usually opaque and the balls brightly colored. Here we examined a version of this trick inspired by a routine performed by the entertainment duo Penn & Teller, conducted with three opaque and subsequently with three transparent cups.\nMagician Teller devised this variation while fiddling with an empty water glass and wadded-up paper napkins for balls, at a Midwestern diner (Macknik, Martinez-Conde & Blakeslee 2010). He turned the glass upside down and put a ball on top, then tilted the glass so that the ball fell into his other hand. The falling ball was so compelling that it even drew his own attention away from his other hand, which was deftly and automatically loading a second ball under the glass. (He was so well practiced that he no longer needed to consciously control his hands.) In fact, Teller found that the sleight happened so quickly he himself did not realize he had loaded the transparent cup. Teller further realized that all of this took place despite the fact that he should have been able to see the secret ball as it was loaded under the cup. Its image was on his retina, but he nevertheless missed it because his attention was so enthralled with the falling ball. He surmised that if it worked for him with a transparent cup, it would work with an audience. The transparency of the cups would make the trick all the more magical to the audience. Penn & Teller claim that their version of the trick violates four rules of magic: don\u2019t tell the audience how the trick is done, don\u2019t perform the same trick twice, don\u2019t show the audience the secret preparation, and never perform cups and balls with clear plastic cups.\nHere we set up to investigate whether the falling ball in Penn & Teller's \"Cups and Balls\" generated stronger misdirection, as hypothesized by Teller, than alternative manipulations. Teller used his right hand to load (i.e. introduce surreptitiously) a small ball inside each of two upside-down cups, one at a time, while using his left hand to remove a different ball from the upside-down bottom of the cup. The third cup sleight involved one of six manipulations: a) standard maneuver (i.e. ball falling to the magicians' hand), b) standard maneuver without a third ball, c) ball placed on the table before going to the pocket, d) ball lifted before going to the pocket, e) ball dropped to the floor, and f) ball stuck to the cup. See Supplemental Movies S1-6. Seven subjects watched the videos of the performances while reporting, via button press, whenever balls were removed from the cups/table (button \u201c1\u201d) or placed inside the cups/on the table (button \u201c2\u201d).\nSubjects\u2019 perception was more accurate with transparent than with opaque cups. Perceptual performance was worse for the conditions where the ball was placed on the table, or stuck to the cup, than for the standard maneuver. The condition in which the ball was lifted displaced the\nPre Prin ts Pre Prin ts\nsubjects\u2019 gaze position the most, whereas the condition in which there was no ball caused the smallest gaze displacement. Thus, neither the \"Standard\" falling ball or the enhanced falling ball condition (where the ball fell to the floor) generated stronger misdirection, either in terms of perceptual performance or gaze position, contrary to the magician's expectation.\nTraining improved the subjects\u2019 perceptual performance. Occlusion of the magician\u2019s face did not affect the subjects\u2019 perception, suggesting that gaze misdirection does not play a strong role in the \"Cups and Balls\" illusion. Our results have implications for how to optimize the performance of this classic magic trick, and for the types of hand and object motion that maximize magic misdirection.",
+    "v2_col_introduction": "introduction  : Magic is one of the oldest art forms, and magicians have manipulated audiences' perceptual and cognitive processes for much longer than cognitive scientists have (Martinez-Conde & Macknik 2007; Martinez-Conde & Macknik 2008; Macknik et al. 2008), Thus, classic and contemporary magic illusions provide scientists with methodological refinements and testable hypotheses about the building blocks of conscious experience (Cui et al. 2011; Otero-Millan et al. 2011). The \u201cCups and Balls\u201d is a sleight-of-hand magic trick that was performed by Roman conjurers as far back as two thousand years ago (Christopher & Christopher 2006). The trick has many variations, but the most common one uses three balls and three cups. The magician makes the balls pass through the bottom of cups, jump from cup to cup, disappear from a cup and turn up elsewhere, turn into other objects, and so on. The cups are usually opaque and the balls brightly colored. Here we examined a version of this trick inspired by a routine performed by the entertainment duo Penn & Teller, conducted with three opaque and subsequently with three transparent cups.\nMagician Teller devised this variation while fiddling with an empty water glass and wadded-up paper napkins for balls, at a Midwestern diner (Macknik, Martinez-Conde & Blakeslee 2010). He turned the glass upside down and put a ball on top, then tilted the glass so that the ball fell into his other hand. The falling ball was so compelling that it even drew his own attention away from his other hand, which was deftly and automatically loading a second ball under the glass. (He was so well practiced that he no longer needed to consciously control his hands.) In fact, Teller found that the sleight happened so quickly he himself did not realize he had loaded the transparent cup. Teller further realized that all of this took place despite the fact that he should have been able to see the secret ball as it was loaded under the cup. Its image was on his retina, but he nevertheless missed it because his attention was so enthralled with the falling ball. He surmised that if it worked for him with a transparent cup, it would work with an audience. The transparency of the cups would make the trick all the more magical to the audience. Penn & Teller claim that their version of the trick violates four rules of magic: don\u2019t tell the audience how the trick is done, don\u2019t perform the same trick twice, don\u2019t show the audience the secret preparation, and never perform cups and balls with clear plastic cups.\nHere we set up to investigate whether the falling ball in Penn & Teller's \"Cups and Balls\" generated stronger misdirection, as hypothesized by Teller, than alternative manipulations. Teller used his right hand to load (i.e. introduce surreptitiously) a small ball inside each of two upside-down cups, one at a time, while using his left hand to remove a different ball from the upside-down bottom of the cup. The third cup sleight involved one of six manipulations: a) standard maneuver (i.e. ball falling to the magicians' hand), b) standard maneuver without a third ball, c) ball placed on the table before going to the pocket, d) ball lifted before going to the pocket, e) ball dropped to the floor, and f) ball stuck to the cup. See Supplemental Movies S1-6. Seven subjects watched the videos of the performances while reporting, via button press, whenever balls were removed from the cups/table (button \u201c1\u201d) or placed inside the cups/on the table (button \u201c2\u201d).\nSubjects\u2019 perception was more accurate with transparent than with opaque cups. Perceptual performance was worse for the conditions where the ball was placed on the table, or stuck to the cup, than for the standard maneuver. The condition in which the ball was lifted displaced the\nPre Prin ts Pre Prin ts\nsubjects\u2019 gaze position the most, whereas the condition in which there was no ball caused the smallest gaze displacement. Thus, neither the \"Standard\" falling ball or the enhanced falling ball condition (where the ball fell to the floor) generated stronger misdirection, either in terms of perceptual performance or gaze position, contrary to the magician's expectation.\nTraining improved the subjects\u2019 perceptual performance. Occlusion of the magician\u2019s face did not affect the subjects\u2019 perception, suggesting that gaze misdirection does not play a strong role in the \"Cups and Balls\" illusion. Our results have implications for how to optimize the performance of this classic magic trick, and for the types of hand and object motion that maximize magic misdirection.",
+    "v3_col_introduction": "introduction  : Magic is one of the oldest art forms, and magicians have manipulated audiences' perceptual and cognitive processes for much longer than cognitive scientists have (Martinez-Conde & Macknik 2007; Martinez-Conde & Macknik 2008; Macknik et al. 2008), Thus, classic and contemporary magic illusions provide scientists with methodological refinements and testable hypotheses about the building blocks of conscious experience (Cui et al. 2011; Otero-Millan et al. 2011). The \u201cCups and Balls\u201d is a sleight-of-hand magic trick that was performed by Roman conjurers as far back as two thousand years ago (Christopher & Christopher 2006). The trick has many variations, but the most common one uses three balls and three cups. The magician makes the balls pass through the bottom of cups, jump from cup to cup, disappear from a cup and turn up elsewhere, turn into other objects, and so on. The cups are usually opaque and the balls brightly colored. Here we examined a version of this trick inspired by a routine performed by the entertainment duo Penn & Teller, conducted with three opaque and subsequently with three transparent cups.\nMagician Teller devised this variation while fiddling with an empty water glass and wadded-up paper napkins for balls, at a Midwestern diner (Macknik, Martinez-Conde & Blakeslee 2010). He turned the glass upside down and put a ball on top, then tilted the glass so that the ball fell into his other hand. The falling ball was so compelling that it even drew his own attention away from his other hand, which was deftly and automatically loading a second ball under the glass. (He was so well practiced that he no longer needed to consciously control his hands.) In fact, Teller found that the sleight happened so quickly he himself did not realize he had loaded the transparent cup. Teller further realized that all of this took place despite the fact that he should have been able to see the secret ball as it was loaded under the cup. Its image was on his retina, but he nevertheless missed it because his attention was so enthralled with the falling ball. He surmised that if it worked for him with a transparent cup, it would work with an audience. The transparency of the cups would make the trick all the more magical to the audience. Penn & Teller claim that their version of the trick violates four rules of magic: don\u2019t tell the audience how the trick is done, don\u2019t perform the same trick twice, don\u2019t show the audience the secret preparation, and never perform cups and balls with clear plastic cups.\nHere we set up to investigate whether the falling ball in Penn & Teller's \"Cups and Balls\" generated stronger misdirection, as hypothesized by Teller, than alternative manipulations. Magician Teller used his right hand to load (i.e. introduce surreptitiously) a small ball inside each of two upside-down cups, one at a time, while using his left hand to remove a different ball from the upside-down bottom of the cup. The third cup sleight involved one of six manipulations: a) standard maneuver (i.e. ball falling to the magicians' hand), b) standard maneuver without a third ball, c) ball placed on the table before going to the pocket, d) ball lifted before going to the pocket, e) ball dropped to the floor, and f) ball stuck to the cup. Seven subjects watched the videos of the performances while reporting, via button press, whenever balls were removed from the cups/table (button \u201c1\u201d) or placed inside the cups/on the table (button \u201c2\u201d).\nSubjects\u2019 perception was more accurate with transparent than with opaque cups. Perceptual performance was worse for the conditions where the ball was placed on the table, or stuck to the cup, than for the standard maneuver. The condition in which the ball was lifted displaced the subjects\u2019 gaze position the most, whereas the condition in which there was no ball caused the\nPre Prin ts Pre Prin ts\nsmallest gaze displacement. Thus, neither the \"Standard\" falling ball or the enhanced falling ball condition (where the ball fell to the floor) generated stronger misdirection, either in terms of perceptual performance or gaze position, contrary to the magician's expectation.\nTraining improved the subjects\u2019 perceptual performance. Occlusion of the magician\u2019s face did not affect the subjects\u2019 perception, suggesting that gaze misdirection does not play a strong role in the \"Cups and Balls\" illusion. Our results have implications for how to optimize the performance of this classic magic trick, and for the types of hand and object motion that maximize magic misdirection.",
+    "v1_text": "results : Perceptual reports Subjects reported the placement and removal of balls: they pressed \u201c1\u201d whenever a ball was removed and \u201c2 whenever a new ball was placed on the table or under a cup (Figure 2; see Methods for details). We analyzed subject performance using a logistic regression model (Hosmer-Lemeshow statistic \u03c72=2.02, area under ROC curve AUC = 0.77). Subjects\u2019 performance in reporting the loading of the third cup was at chance level in the conditions with opaque cups (p > 0.05), and significantly improved in the transparent cups trials, when taking all the experimental trials into account (p < 10-7) (Figure 3A). Performance was also better for simulated rather than real loads in the opaque cups (p < 10-6), due to skipped loading reports in the opaque cups condition, which impaired perceptual performance for the real loads, but not for the simulated loads (p < 0.001) (Figure 3C). From the various sleight-of-hand maneuvers tested, the last-ball loading reports were significantly worse for the \u201cTable\u201d p<0.05) and the \u201cStuck\u201d (p < 0.05) conditions than for the \u201cStandard\u201d condition (Figure 3A). Subjects\u2019 performance was equivalent when the magician\u2019s face was visible and when it was blocked (Figure 3B). Subjects\u2019 reaction times were comparable for all three cups, across the six different sleight-of-hand manipulations (for each individual condition and for the six conditions together as a whole), and for visible vs. blocked faces. Gaze dynamics We studied the subjects\u2019 gaze dynamics during the viewing of each video clip (Figure 4; see Methods for details) using a linear regression model (R2 = 0.19). Gaze distance to the third cup was highest for the \u201cLift\u201d condition (p < 0.0001) and lowest for the \u201cNo ball\u201d condition (p < 0.0001), suggesting that the \u201cLift\u201d manipulation causes the largest gaze displacement (i.e. overt misdirection (Macknik et al. 2008)), whereas the \u201cNo ball\u201d manipulation produced the smallest gaze displacement/misdirection (possibly because in the absence of a ball, subjects may allocate stronger attention to the cup) (Figure 4A). We used a different linear regression model (R2 = 0.18) to correlate gaze distance and reaction times, we found that increased gaze distance resulted in higher reaction times (p < 0.001), with a significant effect of sleight-of-hand manipulation after controlling for the effect of gaze distance (p < 0.05). To study the potential effect of saccadic suppression on the perceptual differences we found across conditions, we estimated the saccade production rate in the same movie frames used to Pre Prin ts Pre Prin ts measure the gaze distance to the bottom of the cup (Supplemental Figure S1). Saccade production was equivalent across the tested conditions. Learning effects Subjects\u2019 performance improved over the course of the experiment (Figure 5). In the opaque cups conditions, the number of \u201clate findings\u201d (i.e. ball placement reports after the magician showed the contents of the cups) decreased with trial number (logistic regression, p < 0.01 Hosmer-Lemeshow statistic \u03c72=17.23, area under ROC curve AUC = 0.92). In the transparent cups conditions, there were few \u201clate findings\u201d, even in the initial trials (Figure 5A). In the clear cups conditions, correct loading reports for the third cup increased as the experiment progressed (logistic regression, p < 0.001, Hosmer-Lemeshow statistic \u03c72=15.02, area under ROC curve AUC = 0.73) (Figure 5B). In the opaque cup conditions, subjects did not have any information about the load of the last cup, and performed at chance, therefore we found no apparent learning effect, as expected. Reaction times decreased (linear regression, p < 0.05, R2 = 0.34) (Figure 5C) with trial number in the clear cups conditions, but remained constant in the trials with opaque cups, indicating that subjects were guessing during this condition. Gaze distance to the bottom of the third cup decreased with trial number for transparent and opaque cups (linear regression, p < 0.01, R2=0.31). To ensure that this learning effect did not affect our other conclusions about the experimental conditions, we conducted an additional analysis of subject performance as a function of the first viewing of each condition (Supplemental Figure S2). The results are comparable to those in Figure 3, indicating that the learning effect did not affect subject performance as a function of condition. Further, because the sequence of conditions was random and different for each subject, a systematic learning effect could not have biased our other results. acknowledgements : We thank the NOVA production crew for providing the filming of the videos used in the experiment, Penn & Teller from providing us with their theater and Teller for performing the magic tricks we used. methods : Subjects Seven naive subjects participated in the experiment. All participants had normal or corrected-to-normal vision, and were paid $15 dollars for a single experimental session. The experiment was carried out under the guidelines of the Barrow Neurological Institute\u2019s Internal Review Board (protocol 04BN039), and written informed consent was obtained from each participant. Eye movement recordings During the experiment, subjects rested their head on a chin/forehead-rest 57 cm away from a video monitor (Barco Reference Calibrator V), while free viewing the video clips. Their eye movements were non-invasively recorded with video-based eye tracker (Eyelink 1000, SR Research), at 500 samples per second. From the eye tracker recordings, we identified and removed blink periods as the portions of the recorded data were the pupil information was missing. Furthermore, we removed the 200 ms before and after each identified blink period, to eliminate periods of time in which the pupil is partially occluded. We identified saccades using an objective algorithm (Engbert and Kliegl 2003). To reduce spurious positives due to noise, we analyzed only binocular saccades (i.e. saccades with at least one sample of overlap in both eyes). Furthermore, we ensured that overshoot corrections were not counted as saccades by imposing a minimum intersaccadic interval of 20 ms (Otero-Millan et al. 2011). Experimental design Subjects sat in a dark, quiet room and watched video clips of 10 to 12 seconds each, in which Teller performed different variations of a \u201ccups and balls\u201d magic routine. The videos had a resolution of 720x480 pixels and subtended an area of 28x19 degrees of visual angle inside the visual field. The average luminance of the clips was 23 cd/m2, and their contrast ratio (full on/full off) was 128:1. Areas of the screen not occupied by the video were white. In each clip, Teller performed the manipulation sequentially in each of three different cups, located from left to right on the screen. The manipulation in the first two cups was identical in all the clips (Standard load, see below), whereas the routine used in the third cup varied in each video clip (Figure 1). After the third cup\u2019s sleight was complete, Teller individually lifted all Pre Prin ts Pre Prin ts three cups to show the balls hidden underneath them. Subjects were instructed to report, as fast as possible, the removal and placing of each ball as soon as they were aware of them, by pressing one of two different buttons on a gamepad with their left and right index fingers (button \u201c1\u201d for removals, button \u201d2\u201d for placings, see Figure 2). A removal was defined as the moment each ball stopped touching either the table or a cup, and a placing was defined as when each ball made physical contact with a cup or the table. The different routines tested were: a) Standard (Supplemental Movie S1): the standard maneuver, identical to the one performed in the first two cups b) No ball (Supplemental Movie S2): similar to the \u201cStandard\u201d routine, but there was no ball on top of the third cup c) Lift (Supplemental Movie S3): the ball on top of the third cup was lifted to eye level before loading the cup d) Table (Supplemental Movie S4): the top ball on the third cup was placed on the table before the cup was loaded e) Drop (Supplemental Movie S5): the top ball on the third cup was dropped to the floor before the cup was loaded f) Stuck (Supplemental Movie S6): similar to the \u201cDrop\u201d condition, but the ball was stuck to the cup and therefore it did not fall. For each of these different routines, we tested other variables concerning the magician\u2019s performance. We tested \u201cClear Cups\u201d, in which the cups were transparent, versus \u201cOpaque Cups\u201d, in which the cups were not. We also tested \u201cLoad\u201d versus \u201cNo load\u201d conditions, in which the third cup was either \u201cloaded\u201d (i.e. a ball was surreptitiously placed under it), or not. Finally, we compared a \u201cNo face\u201d condition, in which the magician\u2019s face was occluded by a static black rectangle, versus the unmodified \u201cFace\u201d-visible video clips. This yielded a total of 48 conditions. Each subject saw each condition twice. The order of conditions was blocked and randomized for each subject. Each participant saw all the 48 conditions first in random order, and then the same conditions again in a new random sequence. Data analysis We defined a correct report of ball placement or removal as an appropriate button press in the 2000 ms immediately following the first movie frame in a movie in which the ball had been placed or removed. We also coded correct reports when subject did not indicate a placement after the magician performed a faked load. The reaction time of each report was measured in the conditions in which the cup was loaded. For each placing, gaze distance and saccade rate was calculated as the average distance between the subjects\u2019 gaze and the point where the cup sat on the table, during the 400 ms immediately subsequent to, and following, the first movie frame in which the load occurs (or the equivalent frame in the \u201cNo load\u201d condition). We varied the duration of these two time windows and found that the results were similar. Subjects were allowed to report ball placings during the reveal sequence at the end of each trial, in which the magician lifted the cups to show their contents. We counted the number of reports the subjects made during this period in each trial and considered them \u201clate findings\u201d. Pre Prin ts Pre Prin ts Statistical testing employed a logistic regression fit to correct reports of placings and removals, and a linear regression fit to the reaction times and the gaze distances. The different magic routines, the load or no load of the third cup, the visibility or occlusion of the face, and the clear or opaque cups were factors in the main analyses. The analyses to determine the evolution of responses and gaze positions throughout the experiment used only the trial number as predictor. The statistical models determined main effects and first order interactions, when applicable. Only significant effects are reported in the text. Pairwise comparisons across different routines were tested with the Newman-Keuls post hoc test. discussion : We investigated the potential contribution of several perceptual elements in Penn & Teller's version of the classic \"Cups and balls\" magic trick. We measured the perceptual performance and gaze behavior of naive observers as Teller surreptitiously introduced balls inside opaque and transparent upside down plastic cups. Contrary to the magician's intuition, a gravity-driven drop of a ball into his hand (or to the floor) caused less misdirection, both in terms of gaze displacement and impaired perception, than alternative manipulations such as lifting the ball, or placing it on the table. Thus, perception of (the effects of gravity on) falling objects does not enhance magic misdirection, at least in the performance of this particular sleight-of-hand trick. The contradiction between our results and the magician\u2019s original perception may have been caused by one or more of several possible sources. One possibility is that performing the trick in a new way may have drawn his attention towards the new element (the ball dropping), and away from the common element (the loading of the cup). Successive, non-controlled repetitions of the procedure could have given the impression of a worse detection of the loading because of confirmation bias. Our results confirm that controlled experiments give valuable insight to reject (Cui et al. 2011) or accept (Otero-Millan et al. 2011) intuitive judgments about attention and misdirection formulated by the magicians. Further, the three consecutive sleight-of-hand Pre Prin ts Pre Prin ts manipulations (actual or simulated loads) were presented in isolation, rather than as part of a complete \"Cups and balls\" magic routine (an arrangement of tricks organized in logical fashion as part of a magic performance). Finally, because an actual magician (i.e. rather than a cartoon or computer simulation) performed all maneuvers, motion features such as timing, duration, etc. could not be exactly equated across all experimental conditions. Future research using computer simulations of the magician\u2019s sleight-of-hand movements should conducted with the goal of replicating and generalizing the current findings to other sleights-of-hand and magic tricks. Blocking or unblocking the magician's face did not affect the observers' perception or oculomotor behavior, suggesting that the \"Cups and balls\" magic trick does not rely on social misdirection (for instance, due to the magician's head or eye position/movements). These results are surprising\u2014the belief among magician's that social misdirection, generated by the face, is one of their most powerful tools, is pervasive\u2014though they agree with those reported by Cui et al (Cui et al. 2011) with a different magic trick. Together they suggest that different magic illusions may differentially be enhanced, unchanged or lessened by social misdirection. Also in agreement with Cui et al's research (Cui et al. 2011), we found significant effects of learning on the perception and gaze behavior or initially naive observers--the more spectators see a trick the less effective the misdirection. Our combined results have implications for how to optimize the performance of the \"Cups and balls\" magic trick, and for the types of hand and object motion that maximize magic misdirection. in \u201cstandard\u201d, the routine performed on the third cup was identical to that performed for the previous : two cups. In \u201cNo ball\u201d, the routine was again the same as in the first two cups, but there was no ball initially placed on top of the cup. In \u201clift\u201d, the ball initially on top of the third cup was lifted to approximately eye level before the cup was loaded. In \u201cTable\u201d, the ball originally on top of the third cup was placed on the table before the cup was loaded. In \u201cDrop\u201d, the ball was dropped out of the screen before the third cup was loaded. In \u201cStuck\u201d, the ball was attached to the top of the third cup but the magician follows it with his gaze as in the \u201cDrop\u201d condition. Pre Prin ts Pre Prin ts Pre Prin ts Pre Prin ts Figure 2 example trial : Schematic of a single trial for the standard routine in which the third cup is loaded. The spikes in the time courses show when a load or removal happens (dashed lines) or is reported (solid lines). Blue color symbolizes the removals of balls, and red color indicates the loadings. Pre Prin ts Pre Prin ts Pre Prin ts Pre Prin ts Figure 3 Summary of subjects\u2019 performance across the different conditions a) Subjects performance in reporting the load of the third cup across the different routines, for the conditions with clear and opaque cups. Performance is uniform across the different routines for the opaque cups, and worse than in the conditions with clear cups (logistic regression, p < 10-7). When the cups are clear, and the loading or no loading of the cup is therefore visible, performance is worse for the \u201cTable\u201d and \u201cStuck\u201d routines (logistic regression, p < 0.05). b) Performance was similar regardless of the face being visible or not. c) Performance was better for the \u201cNo load\u201d condition with opaque cups (logistic regression, p < 0.001). Dashed lines show the expected chance performance level. Error bars indicate the standard error from the mean across subjects.Pre Prin ts Pre Prin ts Pre Prin ts Pre Prin ts Figure 4 Gaze displacement from the bottom of the cup at the time the cup is being loaded a) Gaze distance across the different routines. The \u201cLift\u201d routine causes the biggest displacement from the bottom of the cup (linear regression, p <0.0001), while the \u201cNo ball\u201d routine produces the smallest one (linear regression, p < 0.0001). b) Gaze displacement was similar for the \u201cClear Cups\u201d and \u201cOpaque Cups\u201d conditions. c) Gaze displacement was similar for the \u201cFace\u201d and \u201cNo face\u201d conditions. Distance is reported in units of degrees of visual angle, and error bars indicate the standard error from the mean across subjects. Pre Prin ts Pre Prin ts Pre Prin ts Pre Prin ts Figure 5 Effects of learning in perceptual reports and gaze distance a)The number of late findings (placings reported after the magician shows what is under the cups) goes down with the trial number in the conditions with the opaque cups, while is very low during the experiment for the conditions with clear cups. The correlation between the trial number and the number of late findings is statistically significant in the conditions with opaque cups (logistic regression, p < 0.01). b) Probability of subjects reporting correctly the loading of the third cup in the conditions with clear cups as a function of the trial number. The relationship is statistically significant (logistic regression, p < 0.001). c) The reaction times of the subjects reporting the loading of the third cup (in the conditions with clear cups) decreased with the number of trials (linear regression, p < 0.05). d) Similarly, the gaze distance (in degrees of visual angle) to the bottom of the cup decreased with the trial number (linear regression, p < 0.01). Error bars indicate standard error from the mean across subjects. Pre Prin ts Pre Prin ts Pre Prin ts Pre Prin ts",
+    "v2_text": "results : Perceptual reports Subjects reported the placement and removal of balls: they pressed \u201c1\u201d whenever a ball was removed and \u201c2 whenever a new ball was placed on the table or under a cup (Figure 2; see Methods for details). Subjects\u2019 performance in reporting the loading of the third cup was at chance level in the conditions with opaque cups (logistic regression, p > 0.05, Hosmer-Lemeshow statistic \u03c72=2.02, area under ROC curve AUC = 0.77), and significantly improved in the transparent cups trials, when taking all the experimental trials into account (logistic regression, p < 10-7, Hosmer-Lemeshow statistic \u03c72=2.02, area under ROC curve AUC = 0.77) (Figure 3A). Performance was also better for simulated rather than real loads in the opaque cups (logistic regression, p < 10-6, Hosmer-Lemeshow statistic \u03c72=2.02, area under ROC curve AUC = 0.77), due to skipped loading reports in the opaque cups condition, which impaired perceptual performance for the real loads, but not for the simulated loads (logistic regression, p < 0.001, Hosmer-Lemeshow statistic \u03c72=2.02, area under ROC curve AUC = 0.77) (Figure 3C). From the various sleight-of-hand maneuvers tested, the last-ball loading reports were significantly worse for the \u201cTable\u201d (logistic regression, p<0.05, Hosmer-Lemeshow statistic \u03c72=2.02, area under ROC curve AUC = 0.77) and the \u201cStuck\u201d (logistic regression, p < 0.05, Hosmer-Lemeshow statistic \u03c72=2.02, area under ROC curve AUC = 0.77) conditions than for the \u201cStandard\u201d condition (Figure 3A). Subjects\u2019 performance was equivalent when the magician\u2019s face was visible and when it was blocked (Figure 3B). Subjects\u2019 reaction times were comparable for all three cups, across the six different sleight-of-hand manipulations (for each individual condition and for the six conditions together as a whole), and for visible vs. blocked faces. Gaze dynamics We studied the subjects\u2019 gaze dynamics during the viewing of each video clip (Figure 4; see Methods for details). Gaze distance to the third cup was highest for the \u201cLift\u201d condition (linear regression, p < 0.0001, R2 = 0.19) and lowest for the \u201cNo ball\u201d condition (linear regression, p < 0.0001, R2 = 0.19), suggesting that the \u201cLift\u201d manipulation causes the largest gaze displacement (i.e. overt misdirection (Macknik et al. 2008)), whereas the \u201cNo ball\u201d manipulation produced the smallest gaze displacement/misdirection (possibly because in the absence of a ball, subjects may allocate stronger attention to the cup) (Figure 4A). Increased gaze distance resulted in higher reaction times (linear regression, p < 0.001, R2 = 0.18), with a significant effect of sleight-of-hand manipulation after controlling for the effect of gaze distance (linear regression, p < 0.05, R2 = 0.18). Pre Prin ts Pre Prin ts To study the potential effect of saccadic suppression on the perceptual differences we found across conditions, we estimated the saccade production rate in the same movie frames used to measure the gaze distance to the bottom of the cup (Supplemental Figure S1). Saccade production was equivalent across the tested conditions. Learning effects Subjects\u2019 performance improved over the course of the experiment (Figure 5). In the opaque cups conditions, the number of \u201clate findings\u201d (i.e. ball placement reports after the magician showed the contents of the cups) decreased with trial number (logistic regression, p < 0.01 Hosmer-Lemeshow statistic \u03c72=17.23, area under ROC curve AUC = 0.92). In the transparent cups conditions, there were few \u201clate findings\u201d, even in the initial trials (Figure 5A). In the clear cups conditions, correct loading reports for the third cup increased as the experiment progressed (logistic regression, p < 0.001, Hosmer-Lemeshow statistic \u03c72=15.02, area under ROC curve AUC = 0.73) (Figure 5B). In the opaque cup conditions, subjects did not have any information about the load of the last cup, and performed at chance, therefore we found no apparent learning effect, as expected. Reaction times decreased (linear regression, p < 0.05, R2 = 0.34) (Figure 5C) with trial number in the clear cups conditions, but remained constant in the trials with opaque cups, indicating that subjects were guessing during this condition. Gaze distance to the bottom of the third cup decreased with trial number for transparent and opaque cups (linear regression, p < 0.01, R2=0.31). To ensure that this learning effect did not affect our other conclusions about the experimental conditions, we conducted an additional analysis of subject performance as a function of the first viewing of each condition (Supplemental Figure S2). The results are comparable to those in Figure 3, indicating that the learning effect did not affect subject performance as a function of condition. Further, because the sequence of conditions was random and different for each subject, a systematic learning effect could not have biased our other results. acknowledgements : We thank the NOVA production crew for providing the filming of the videos used in the experiment, Penn & Teller from providing us with their theater and Teller for performing the magic tricks we used. discussion : We investigated the potential contribution of several perceptual elements in Penn & Teller's version of the classic \"Cups and balls\" magic trick. We measured the perceptual performance and gaze behavior of naive observers as Teller surreptitiously introduced balls inside opaque and transparent upside down plastic cups. Contrary to the magician's intuition, a gravity-driven drop of a ball into his hand (or to the floor) caused less misdirection, both in terms of gaze displacement and impaired perception, than alternative manipulations such as lifting the ball, or placing it on the table. Thus, perception of (the effects of gravity on) falling objects does not enhance magic misdirection, at least in the performance of this particular sleight-of-hand trick. The contradiction between our results and the magician\u2019s original perception may have been caused by one or more of several possible sources. One possibility is that performing the trick in a new way may have drawn his attention towards the new element (the ball dropping), and away from the common element (the loading of the cup). Successive, non-controlled repetitions of the procedure could have given the impression of a worse detection of the loading because of Pre Prin ts Pre Prin ts confirmation bias. Our results confirm that controlled experiments give valuable insight to reject (Cui et al. 2011) or accept (Otero-Millan et al. 2011) intuitive judgments about attention and misdirection formulated by the magicians. Further, the three consecutive sleight-of-hand manipulations (actual or simulated loads) were presented in isolation, rather than as part of a complete \"Cups and balls\" magic routine (an arrangement of tricks organized in logical fashion as part of a magic performance). Finally, because an actual magician (i.e. rather than a cartoon or computer simulation) performed all maneuvers, motion features such as timing, duration, etc. could not be exactly equated across all experimental conditions. Future research using computer simulations of the magician\u2019s sleight-of-hand movements should conducted with the goal of replicating and generalizing the current findings to other sleights-of-hand and magic tricks. Blocking or unblocking the magician's face did not affect the observers' perception or oculomotor behavior, suggesting that the \"Cups and balls\" magic trick does not rely on social misdirection (for instance, due to the magician's head or eye position/movements). These results are surprising\u2014the belief among magician's that social misdirection, generated by the face, is one of their most powerful tools, is pervasive\u2014though they agree with those reported by Cui et al (Cui et al. 2011) with a different magic trick. Together they suggest that different magic illusions may differentially be enhanced, unchanged or lessened by social misdirection. Also in agreement with Cui et al's research (Cui et al. 2011), we found significant effects of learning on the perception and gaze behavior or initially naive observers--the more spectators see a trick the less effective the misdirection. Our combined results have implications for how to optimize the performance of the \"Cups and balls\" magic trick, and for the types of hand and object motion that maximize magic misdirection. in \u201cstandard\u201d, the routine performed on the third cup was identical to that performed for the previous : two cups. In \u201cNo ball\u201d, the routine was again the same as in the first two cups, but there was no ball initially placed on top of the cup. In \u201clift\u201d, the ball initially on top of the third cup was lifted to approximately eye level before the cup was loaded. In \u201cTable\u201d, the ball originally on top of the third cup was placed on the table before the cup was loaded. In \u201cDrop\u201d, the ball was dropped out of the screen before the third cup was loaded. In \u201cStuck\u201d, the ball was attached to the top of the third cup but the magician follows it with his gaze as in the \u201cDrop\u201d condition. Pre Prin ts Pre Prin ts Pre Prin ts Pre Prin ts Figure 2 methods : Subjects Seven naive subjects participated in the experiment. All participants had normal or corrected-to-normal vision, and were paid $15 dollars for a single experimental session. The experiment was carried out under the guidelines of the Barrow Neurological Institute\u2019s Internal Review Board (protocol 04BN039), and written informed consent was obtained from each participant. Eye movement recordings During the experiment, subjects rested their head on a chin/forehead-rest 57 cm away from a video monitor (Barco Reference Calibrator V), while free viewing the video clips. Their eye movements were non-invasively recorded with video-based eye tracker (Eyelink 1000, SR Research), at 500 samples per second. From the eye tracker recordings, we identified and removed blink periods as the portions of the recorded data were the pupil information was missing. Furthermore, we removed the 200 ms before and after each identified blink period, to eliminate periods of time in which the pupil is partially occluded. We identified saccades using an objective algorithm (Engbert and Kliegl 2003). To reduce spurious positives due to noise, we analyzed only binocular saccades (i.e. saccades with at least one sample of overlap in both eyes). Furthermore, we ensured that overshoot corrections were not counted as saccades by imposing a minimum intersaccadic interval of 20 ms (Otero-Millan et al. 2011). Experimental design Subjects sat in a dark, quiet room and watched video clips of 10 to 12 seconds each, in which Teller performed different variations of a \u201ccups and balls\u201d magic routine. The videos had a resolution of 720x480 pixels and subtended an area of 28x19 degrees of visual angle inside the visual field. The average luminance of the clips was 23 cd/m2, and their contrast ratio (full on/full off) was 128:1. Areas of the screen not occupied by the video were white. In each clip, Teller performed the manipulation sequentially in each of three different cups, located from left to right on the screen. The manipulation in the first two cups was identical in all the clips (Standard load, see below), whereas the routine used in the third cup varied in each video clip (Figure 1). After the third cup\u2019s sleight was complete, Teller individually lifted all Pre Prin ts Pre Prin ts three cups to show the balls hidden underneath them. Subjects were instructed to report, as fast as possible, the removal and placing of each ball as soon as they were aware of them, by pressing one of two different buttons on a gamepad with their left and right index fingers (button \u201c1\u201d for removals, button \u201d2\u201d for placings, see Figure 2). A removal was defined as the moment each ball stopped touching either the table or a cup, and a placing was defined as when each ball made physical contact with a cup or the table. The different routines tested were: a) Standard (Supplemental Movie S1): the standard maneuver, identical to the one performed in the first two cups b) No ball (Supplemental Movie S2): similar to the \u201cStandard\u201d routine, but there was no ball on top of the third cup c) Lift (Supplemental Movie S3): the ball on top of the third cup was lifted to eye level before loading the cup d) Table (Supplemental Movie S4): the top ball on the third cup was placed on the table before the cup was loaded e) Drop (Supplemental Movie S5): the top ball on the third cup was dropped to the floor before the cup was loaded f) Stuck (Supplemental Movie S6): similar to the \u201cDrop\u201d condition, but the ball was stuck to the cup and therefore it did not fall. For each of these different routines, we tested other variables concerning the magician\u2019s performance. We tested \u201cClear Cups\u201d, in which the cups were transparent, versus \u201cOpaque Cups\u201d, in which the cups were not. We also tested \u201cLoad\u201d versus \u201cNo load\u201d conditions, in which the third cup was either \u201cloaded\u201d (i.e. a ball was surreptitiously placed under it), or not. Finally, we compared a \u201cNo face\u201d condition, in which the magician\u2019s face was occluded by a static black rectangle, versus the unmodified \u201cFace\u201d-visible video clips. This yielded a total of 48 conditions. Each subject saw each condition twice. The order of conditions was blocked and randomized for each subject. Each participant saw all the 48 conditions first in random order, and then the same conditions again in a new random sequence. Data analysis We defined a correct report of ball placement or removal as an appropriate button press in the 2000 ms immediately following the first movie frame in a movie in which the ball had been placed or removed. We also coded correct reports when subject did not indicate a placement after the magician performed a faked load. The reaction time of each report was measured in the conditions in which the cup was loaded. For each placing, gaze distance and saccade rate was calculated as the average distance between the subjects\u2019 gaze and the point where the cup sat on the table, during the 400 ms immediately subsequent to, and following, the first movie frame in which the load occurs (or the equivalent frame in the \u201cNo load\u201d condition). We varied the duration of these two time windows and found that the results were similar. Subjects were allowed to report ball placings during the reveal sequence at the end of each trial, in which the magician lifted the cups to show their contents. We counted the number of reports the subjects made during this period in each trial and considered them \u201clate findings\u201d. Pre Prin ts Pre Prin ts Statistical testing employed a logistic regression fit to correct reports of placings and removals, and a linear regression fit to the reaction times and the gaze distances. The different magic routines, the load or no load of the third cup, the visibility or occlusion of the face, and the clear or opaque cups were factors in the main analyses. The analyses to determine the evolution of responses and gaze positions throughout the experiment used only the trial number as predictor. The statistical models determined main effects and first order interactions, when applicable. Only significant effects are reported in the text. Pairwise comparisons across different routines were tested with the Newman-Keuls post hoc test. example trial : Schematic of a single trial for the standard routine in which the third cup is loaded. The spikes in the time courses show when a load or removal happens (dashed lines) or is reported (solid lines). Blue color symbolizes the removals of balls, and red color indicates the loadings. Pre Prin ts Pre Prin ts Pre Prin ts Pre Prin ts Figure 3 Summary of subjects\u2019 performance across the different conditions a) Subjects performance in reporting the load of the third cup across the different routines, for the conditions with clear and opaque cups. Performance is uniform across the different routines for the opaque cups, and worse than in the conditions with clear cups (logistic regression, p < 10-7). When the cups are clear, and the loading or no loading of the cup is therefore visible, performance is worse for the \u201cTable\u201d and \u201cStuck\u201d routines (logistic regression, p < 0.05). b) Performance was similar regardless of the face being visible or not. c) Performance was better for the \u201cNo load\u201d condition with opaque cups (logistic regression, p < 0.001). Dashed lines show the expected chance performance level. Error bars indicate the standard error from the mean across subjects.Pre Prin ts Pre Prin ts Pre Prin ts Pre Prin ts Figure 4 Gaze displacement from the bottom of the cup at the time the cup is being loaded a) Gaze distance across the different routines. The \u201cLift\u201d routine causes the biggest displacement from the bottom of the cup (linear regression, p <0.0001), while the \u201cNo ball\u201d routine produces the smallest one (linear regression, p < 0.0001). b) Gaze displacement was similar for the \u201cClear Cups\u201d and \u201cOpaque Cups\u201d conditions. c) Gaze displacement was similar for the \u201cFace\u201d and \u201cNo face\u201d conditions. Distance is reported in units of degrees of visual angle, and error bars indicate the standard error from the mean across subjects. Pre Prin ts Pre Prin ts Pre Prin ts Pre Prin ts Figure 5 Effects of learning in perceptual reports and gaze distance a)The number of late findings (placings reported after the magician shows what is under the cups) goes down with the trial number in the conditions with the opaque cups, while is very low during the experiment for the conditions with clear cups. The correlation between the trial number and the number of late findings is statistically significant in the conditions with opaque cups (logistic regression, p < 0.01). b) Probability of subjects reporting correctly the loading of the third cup in the conditions with clear cups as a function of the trial number. The relationship is statistically significant (logistic regression, p < 0.001). c) The reaction times of the subjects reporting the loading of the third cup (in the conditions with clear cups) decreased with the number of trials (linear regression, p < 0.05). d) Similarly, the gaze distance (in degrees of visual angle) to the bottom of the cup decreased with the trial number (linear regression, p < 0.01). Error bars indicate standard error from the mean across subjects. Pre Prin ts Pre Prin ts Pre Prin ts Pre Prin ts",
+    "v3_text": "methods : Subjects Seven naive subjects participated in the experiment. All participants had normal or corrected-to-normal vision, and were paid $15 dollars for a single experimental session. The experiment was carried out under the guidelines of the Barrow Neurological Institute\u2019s Internal Review Board (protocol 04BN039), and written informed consent was obtained from each participant. Eye movement recordings During the experiment, subjects rested their head on a chin/forehead-rest 57 cm away from a video monitor (Barco Reference Calibrator V), while their eye movements were non-invasively recorded with video-based eye tracker (Eyelink 1000, SR Research), at 500 samples per second. From the eye tracker recordings, we identified and removed blink periods as the portions of the recorded data were the pupil information was missing. Furthermore, we removed the 200 ms before and after each identified blink period, to eliminate periods of time in which the pupil is partially occluded. Experimental design Subjects were presented with video clips in which a magician performed different variations of a \u201ccups and balls\u201d routine. In each clip, the magician performed the manipulation sequentially in each of three different cups, located from left to right on the screen. The manipulation in the first two cups was identical in all the clips (the Standard load, see below), whereas the routine used in the third cup varied in each video clip (Figure 1). After the third cup\u2019s sleight was complete, the magician individually lifted all three of the cups to show the balls hidden underneath them. Subjects reported the removal and placing of each ball as soon as they were aware of them, by pressing one of two different buttons on a gamepad (button \u201c1\u201d for removals, button \u201d2\u201d for placings). A removal was defined as the moment each ball stopped touching either the table or a cup, and a placing was defined as when each ball made physical contact with a cup or the table. The different routines tested were: a) Standard: the standard maneuver, identical to the one performed in the first two cups b) No ball: similar to the \u201cStandard\u201d routine, but there was no ball on top of the third cup c) Lift: the ball on top of the third cup was lifted to eye level before loading the cup d) Table: the top ball on the third cup was placed on the table before the cup was loaded e) Drop: the top ball on the third cup was dropped to the floor before the cup was loaded Pre Prin ts Pre Prin ts f) Stuck: similar to the \u201cDrop\u201d condition, but the ball was stuck to the cup and therefore it did not fall. For each of these different routines, we tested other variables concerning the magician\u2019s performance. We tested \u201cClear Cups\u201d , in which the cups were transparent, versus \u201cOpaque Cups\u201d, in which the cups were not. We also tested \u201cLoad\u201d versus \u201cNo load\u201d conditions, in which the third cup was either \u201cloaded\u201d (i.e. a ball was surreptitiously placed under it), or not. Finally, we compared a \u201cNo face\u201d condition, in which the magician\u2019s face was occluded by a static black rectangle, versus the unmodified \u201cFace\u201d-visible video clips. This yielded a total of 48 conditions. Each subject saw each condition twice. Data analysis We defined a correct report of ball placement or removal as an appropriate button press in the 2000 ms immediately following the first movie frame in a movie in which the ball had been placed our removed. We also coded correct reports when subject did not indicate a placement after the magician performed a faked load. The reaction time of each report was measured in the conditions in which the cup was loaded. For each placing, the gaze distance was calculated as the average distance between the subjects\u2019 gaze and the point where the cup sat on the table, during the 400 ms immediately subsequent to, and following, the first movie frame in which the load occurs (or the equivalent frame in the \u201cNo load\u201d condition). We varied the duration of these two time windows and found that the results were similar. Subject reports were considered late if they occurred after the magician lifted the three cups in the reveal sequence at the end of the trick. Statistical testing employed a logistic regression fit to correct reports of placings and removals, and a linear regression fit to the reaction times and the gaze distances. The statistical models determined main effects and first order interactions, when applicable. Pairwise comparisons across different routines were tested using the Newman-Keuls post hoc test. results : Perceptual reports Subjects reported the place and removal of balls: they pressed \u201c1\u201d whenever a ball was removed and \u201c2 whenever a new ball was placed on the table or under a cup (Figure 2; see Methods for details). Performance was significantly better for transparent than for opaque cups (logistic regression, p < 10-7) (Figure 2A), as well as for simulated rather than real loads in the opaque cups (logistic regression, p<10-6) (Figure 2C). From the various sleight-of-hand maneuvers tested, performance was significantly worse for the \u201cTable\u201d (logistic regression, p<0.05) and the \u201cStuck\u201d (logistic regression, p<0.05) conditions than for the \u201cStandard\u201d condition (Figure 2A). Subjects\u2019 performance was equivalent when the magician\u2019s face was visible and when it was blocked (Figure 2B). Subjects\u2019 reaction times were comparable for all three cups, across the six different sleight-of-hand manipulations (for each individual condition and for the six conditions together as a whole), and for visible vs. blocked faces (linear regression, p > 0.05). Pre Prin ts Pre Prin ts Gaze dynamics We studied the subjects\u2019 gaze dynamics during the viewing of each video clip (Figure 3; see Methods for details). Gaze distance to the third cup was highest for the \u201cLift\u201d condition (linear regression, p < 0.0001) and lowest for the \u201cNo ball\u201d condition (linear regression, p < 0.0001), suggesting that the \u201cLift\u201d manipulation causes the largest gaze displacement (i.e. overt misdirection (Macknik et al. 2008)), whereas the \u201cNo ball\u201d manipulation produced the smallest gaze displacement/misdirection (possibly because in the absence of a ball, subjects may allocate stronger attention to the cup) (Figure 3A). Increased gaze distance resulted in higher reaction times (linear regression, p<0.001), with a significant effect of sleight-of-hand manipulation after controlling for the effect of gaze distance (linear regression, p<0.05). Learning effects Subjects\u2019 performance improved over the course of the experiment (Figure 4). In the opaque cups conditions, the number of \u201clate findings\u201d (i.e. ball placement reports after the magician showed the contents of the cups) decreased with trial number (logistic regression, p < 0.01). In the transparent cups conditions, there were few \u201clate findings\u201d, even in the initial trials (Figure 4A). In the clear cups conditions, correct loading reports for the third cup increased (logistic regression, p < 0.001) (Figure 4B) and reaction times decreased (linear regression, p < 0.05) (Figure 4C) with trial number. The effects of learning on reaction times and correct loading reports were less clear in the opaque cups conditions. Gaze distance to the bottom of the third cup decreased with trial number for transparent and opaque cups (linear regression, p < 0.01). discussion : Here we investigated the potential participation of several perceptual elements in professional magicians Penn & Teller's version of the classic \"Cups and balls\" magic trick. We measured the perceptual performance and gaze behavior of naive observers as magician Teller surreptitiously introduced balls inside opaque and transparent upside down plastic cups. Contrary to the magician's intuition, a gravity-driven drop of a ball into his hand (or to the floor) caused less misdirection, both in terms of gaze displacement and impaired perception, than alternative manipulations such as lifting the ball, or placing it on the table. Thus, perception of (the effects of gravity on) falling objects does not enhance magic misdirection, at least in the performance of this particular sleight-of-hand trick. Two experimental factors potentially limit the power of these conclusions. First, the three consecutive sleight-of-hand manipulations (actual or simulated loads) were presented in isolation, rather than as part of a complete \"Cups and balls\" magic routine (an arrangement of tricks organized in logical fashion as part of a magic performance). Second, because an actual magician (i.e. rather than a cartoon or computer simulation) performed all maneuvers, motion features such as timing, duration, etc. could not be exactly equated across all experimental conditions. Future research using computer simulations of the magician\u2019s sleight-of-hand movements should conducted with the goal of replicating and generalizing the current findings to other sleights-of-hand and magic tricks. Pre Prin ts Pre Prin ts Blocking or unblocking the magician's face did not affect the observers' perception or oculomotor behavior, suggesting that the \"Cups and balls\" magic trick does not rely on social misdirection (for instance, due to the magician's head or eye position/movements). These results are surprising--the belief among magician's that social misdirection, generated by the face, is one of their most powerful tools, is pervasive--though they agree with those reported by Cui et al (Cui et al. 2011) with a different magic trick. Together they suggest that different magic illusions may differentially be enhanced, unchanged or lessened by social misdirection. Also in agreement with Cui et al's research (Cui et al. 2011), we found significant effects of learning on the perception and gaze behavior or initially naive observers--the more spectators see a trick the less effective the misdirection. Our combined results have implications for how to optimize the performance of the \"Cups and balls\" magic trick, and for the types of hand and object motion that maximize magic misdirection. acknowledgements : We thank the NOVA production crew for providing the filming of the videos used in the experiment, Penn & Teller from providing us with their theater and Teller for performing the magic tricks we used. in \u201cstandard\u201d, the routine performed on the third cup was identical to that performed for the previous : two cups. In \u201cNo ball\u201d, the routine was again the same as in the first two cups, but there was no ball initially placed on top of the cup. In \u201clift\u201d, the ball initially on top of the third cup was lifted to approximately eye level before the cup was loaded. In \u201cTable\u201d, the ball originally on top of the third cup was placed on the table before the cup was loaded. In \u201cDrop\u201d, the ball was dropped out of the screen before the third cup was loaded. In \u201cStuck\u201d, the ball was attached to the top of the third cup but the magician follows it with his gaze as in the \u201cDrop\u201d condition. Pre Prin ts Pre Prin ts Pre Prin ts Pre Prin ts Figure 2 Summary of the different routines tested in the experiment Summary of subjects\u2019 performance across the different conditions two cups. In \u201cNo ball\u201d, the routine was again the same as in the first two cups, but there was no ball initially placed on top of the cup. In \u201clift\u201d, the ball initially on top of the third cup was lifted to approximately eye level before the cup was loaded. In \u201cTable\u201d, the ball originally on top of the third cup was placed on the table before the cup was loaded. In \u201cDrop\u201d, the ball was dropped out of the screen before the third cup was loaded. In \u201cStuck\u201d, the ball was attached to the top of the third cup but the magician follows it with his gaze as in the \u201cDrop\u201d condition. a) Subjects performance across the different routines, for the conditions with clear and opaque cups. Performance is uniform across the different routines for the opaque cups, and worse than in the conditions with clear cups (logistic regression, p < 10-7). When the cups are clear, and the loading or no loading of the cup is therefore visible, performance is worse for the \u201cTable\u201d and \u201cStuck\u201d routines (logistic regression, p < 0.05). b) Performance was similar regardless of the face being visible or not. c) Performance was better for the \u201cNo load\u201d condition with opaque cups (logistic regression, p < 0.001). Error bars indicate 95% confidence intervals. Pre Prin ts Pre Prin ts Pre Prin ts Pre Prin ts Figure 3 Gaze displacement from the bottom of the cup at the time the cup is being loaded a) Gaze distance across the different routines. The \u201cLift\u201d routine causes the biggest displacement from the bottom of the cup (linear regression, p <0.0001), while the \u201cNo ball\u201d routine produces the smallest one (linear regression, p < 0.0001). b) Gaze displacement was similar for the \u201cClear Cups\u201d and \u201cOpaque Cups\u201d conditions. c) Gaze displacement was similar for the \u201cFace\u201d and \u201cNo face\u201d conditions. Error bars indicate the standard error from the mean across subjects. Pre Prin ts Pre Prin ts Pre Prin ts Pre Prin ts Figure 4 Effects of learning in perceptual reports and gaze distance a) The number of late findings (placings reported after the magician shows what is under the cups) goes down with the trial number in the conditions with the opaque cups, while is very low during the experiment for the conditions with clear cups. The correlation between the trial number and the number of late findings is statistically significant in the conditions with opaque cups (logistic regression, p < 0.01). b) Probability of subjects reporting correctly the loading of the third cup in the conditions with clear cups as a function of the trial number. The relationship is statistically significant (logistic regression, p < 0.001). c) The reaction times of the subjects reporting the loading of the third cup (in the conditions with clear cups) decreased with the number of trials (linear regression, p < 0.05). d) Similarly, the gaze distance to the bottom of the cup decreased with the trial number (linear regression, p < 0.01). Error bars indicate standard error from the mean across subjects. Pre Prin ts Pre Prin ts Pre Prin ts Pre Prin ts",
+    "url": "https://peerj.com/articles/20/reviews/",
+    "review_1": "Linsheng Song \u00b7 Jan 8, 2013 \u00b7 Academic Editor\nACCEPT\nThe manuscript is well revised and can be accepted for PeerJ. Thanks.",
+    "review_2": "Linsheng Song \u00b7 Dec 20, 2012 \u00b7 Academic Editor\nMINOR REVISIONS\nThe manuscript is well written and orgnized. The experiments were conducted in an appropriate manner and data were moderately well presented. It is potentially interesting for researches on iodine nutrition and toxicity in cod larvae. As the reviewers suggested, some tables and figures in the manuscript should be improved.",
+    "review_3": "Karin PITTMAN \u00b7 Dec 19, 2012\nBasic reporting\nThis is a well-written and timely article based on a robust experimental design and thorough analyses. Unfortunately the manuscript mentions two Tables that were not included in the files I received. The figures are all relevant although the figure legends could have been more specific about how many separate figures were meant to be embraced by the figure legend (please insert something like \"Fig. 2 A-G\" at the beginning of the legend). Otherwise the manuscript is of very high intellectual and technical standards.\n\nMinor point: line 226-228 uses the word sample 4 times.\nExperimental design\nThe experimental design was clearly stated and reproducible by those wishing to expend the energy. It was clearly suitable for the question. All standards and codes of conduct have been adhered to.\n\nThe description of the radiography and especially measurement of the neck angle is too brief to be useful (1.5 lines), although the results of this analysis are mentioned in Discussion.\n\nI also had a small question about how the number of follicles was counted (lookup sections?).\n\nOnly one other question arose in my mind, regarding the admixture of rotifers from Control and High iodine enrichments to create a Medium Iodine diet. I can see no suggestion of selectivity by the cod larvae, which would have resulted in high variations for that group, so this may not be important at all.\nValidity of the findings\nThe findings are comprehensively given and analysed. The mortality of one tank changed some of the statistical analyses but this is a minor point. The findings are justified and are put into several relevant contexts: that of dietary requirements and interactions, that of source of water used to rear the larvae (recirculation, wellwater or seawater) and that of longterm effects.\n\nIt might be interesting to have a line or two about synthesis of thyroid hormones in other organs than the thyroid follicles (it happens in fish) and what the \"gross signs\" of iodine deficiency would be in larvae (line 377).\nAdditional comments\nNice work. I like the clear statement about iodine requirements and the codicil about nutrient interactions.\n\nIt would be useful to have the two missing tables, therefore I am recommending \"Minor revisions\".\nCite this review as\nPITTMAN K (2013) Peer Review #1 of \"Iodine nutrition and toxicity in Atlantic cod (Gadus morhua) larvae (v0.1)\". PeerJ https://doi.org/10.7287/peerj.20v0.1/reviews/1",
+    "review_4": "Reviewer 2 \u00b7 Dec 13, 2012\nBasic reporting\nthe manuscript is well written and organized properly\nExperimental design\nyes\nValidity of the findings\nyes\nAdditional comments\nThe manuscript describes effect of Iodine on Atlantic cod (Gadus morhua) larvae. Overall, the manuscript is well written and organized properly, I think the manuscript is suitable for the journal of \u201cThe PeerJ\u201d at current status. Herein, the comments and suggestions are given below in an attempt to improve the paper.\n1) In the abstract section, the sentence \u201clarval Atlantic cod (Gadus morhua) were fed rotifers enriched to intermediate (26 mg I kg -1 dry weight; MI group) or copepod (129 mg I kg -1 DW; HI group) I levels and compared to cod larvae fed control rotifers (0.6mg I kg -1 DW).\u201d should be revised. I could not find any evidence you used copepod for feed trial in Material and method section.\n2) the manuscript contained so many Figures, I recommend the author should be integrated some of them together.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"Iodine nutrition and toxicity in Atlantic cod (Gadus morhua) larvae (v0.1)\". PeerJ https://doi.org/10.7287/peerj.20v0.1/reviews/2",
+    "pdf_1": "https://peerj.com/articles/20v0.2/submission",
+    "pdf_2": "https://peerj.com/articles/20v0.1/submission",
+    "all_reviews": "Review 1: Linsheng Song \u00b7 Jan 8, 2013 \u00b7 Academic Editor\nACCEPT\nThe manuscript is well revised and can be accepted for PeerJ. Thanks.\nReview 2: Linsheng Song \u00b7 Dec 20, 2012 \u00b7 Academic Editor\nMINOR REVISIONS\nThe manuscript is well written and orgnized. The experiments were conducted in an appropriate manner and data were moderately well presented. It is potentially interesting for researches on iodine nutrition and toxicity in cod larvae. As the reviewers suggested, some tables and figures in the manuscript should be improved.\nReview 3: Karin PITTMAN \u00b7 Dec 19, 2012\nBasic reporting\nThis is a well-written and timely article based on a robust experimental design and thorough analyses. Unfortunately the manuscript mentions two Tables that were not included in the files I received. The figures are all relevant although the figure legends could have been more specific about how many separate figures were meant to be embraced by the figure legend (please insert something like \"Fig. 2 A-G\" at the beginning of the legend). Otherwise the manuscript is of very high intellectual and technical standards.\n\nMinor point: line 226-228 uses the word sample 4 times.\nExperimental design\nThe experimental design was clearly stated and reproducible by those wishing to expend the energy. It was clearly suitable for the question. All standards and codes of conduct have been adhered to.\n\nThe description of the radiography and especially measurement of the neck angle is too brief to be useful (1.5 lines), although the results of this analysis are mentioned in Discussion.\n\nI also had a small question about how the number of follicles was counted (lookup sections?).\n\nOnly one other question arose in my mind, regarding the admixture of rotifers from Control and High iodine enrichments to create a Medium Iodine diet. I can see no suggestion of selectivity by the cod larvae, which would have resulted in high variations for that group, so this may not be important at all.\nValidity of the findings\nThe findings are comprehensively given and analysed. The mortality of one tank changed some of the statistical analyses but this is a minor point. The findings are justified and are put into several relevant contexts: that of dietary requirements and interactions, that of source of water used to rear the larvae (recirculation, wellwater or seawater) and that of longterm effects.\n\nIt might be interesting to have a line or two about synthesis of thyroid hormones in other organs than the thyroid follicles (it happens in fish) and what the \"gross signs\" of iodine deficiency would be in larvae (line 377).\nAdditional comments\nNice work. I like the clear statement about iodine requirements and the codicil about nutrient interactions.\n\nIt would be useful to have the two missing tables, therefore I am recommending \"Minor revisions\".\nCite this review as\nPITTMAN K (2013) Peer Review #1 of \"Iodine nutrition and toxicity in Atlantic cod (Gadus morhua) larvae (v0.1)\". PeerJ https://doi.org/10.7287/peerj.20v0.1/reviews/1\nReview 4: Reviewer 2 \u00b7 Dec 13, 2012\nBasic reporting\nthe manuscript is well written and organized properly\nExperimental design\nyes\nValidity of the findings\nyes\nAdditional comments\nThe manuscript describes effect of Iodine on Atlantic cod (Gadus morhua) larvae. Overall, the manuscript is well written and organized properly, I think the manuscript is suitable for the journal of \u201cThe PeerJ\u201d at current status. Herein, the comments and suggestions are given below in an attempt to improve the paper.\n1) In the abstract section, the sentence \u201clarval Atlantic cod (Gadus morhua) were fed rotifers enriched to intermediate (26 mg I kg -1 dry weight; MI group) or copepod (129 mg I kg -1 DW; HI group) I levels and compared to cod larvae fed control rotifers (0.6mg I kg -1 DW).\u201d should be revised. I could not find any evidence you used copepod for feed trial in Material and method section.\n2) the manuscript contained so many Figures, I recommend the author should be integrated some of them together.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"Iodine nutrition and toxicity in Atlantic cod (Gadus morhua) larvae (v0.1)\". PeerJ https://doi.org/10.7287/peerj.20v0.1/reviews/2\nReview 5: nan\nReview 6: \n",
+    "LLM_extracted_review": "['Review 1: The manuscript is well revised and can be accepted for PeerJ. Thanks.\\n\\nReview 2: The manuscript is well written and organized. The experiments were conducted in an appropriate manner and data were moderately well presented. It is potentially interesting for researches on iodine nutrition and toxicity in cod larvae. Some tables and figures in the manuscript should be improved.\\n\\nReview 3: This is a well-written and timely article based on a robust experimental design and thorough analyses. Unfortunately, the manuscript mentions two Tables that were not included in the files. The figures are all relevant, although the figure legends could have been more specific. The description of the radiography and measurement of the neck angle is too brief to be useful. There is a small question about how the number of follicles was counted. The admixture of rotifers from Control and High iodine enrichments may not be important. The findings are comprehensively given and analyzed, but the mortality of one tank changed some of the statistical analyses. It might be interesting to have a line or two about synthesis of thyroid hormones in other organs and the gross signs of iodine deficiency in larvae. It would be useful to have the two missing tables.\\n\\nReview 4: The manuscript is well written and organized properly. The manuscript describes the effect of Iodine on Atlantic cod larvae. The comments and suggestions include revising a sentence in the abstract regarding copepod feed trials, and integrating some of the figures together.']"
+}

peerj_json_files/PeerJ_Json_110.json ADDED Viewed

The diff for this file is too large to render. See raw diff

peerj_json_files/PeerJ_Json_111.json ADDED Viewed

	@@ -0,0 +1,15 @@

+{
+    "v1_Abstract": "The effects of tides on littoral marine habitats are so ubiquitous that shorelines are commonly described as \u2018intertidal\u2019, whereas waves are considered a secondary factor that simply modifies the intertidal habitat. However mean significant wave height exceeds tidal range at many locations worldwide. Here we construct a simple sinusoidal model of coastal water level based on both tidal range and wave height. From the patterns of emergence and submergence predicted by the model, we derive four vertical shoreline benchmarks which bracket up to three novel, spatially distinct, and physically defined zones. The (1) emergent tidal zone is characterized by tidally driven emergence in air; the (2) wave zone is characterized by constant (not periodic) wave wash; and the (3) submergent tidal zone is characterized by tidally driven submergence. The decoupling of tidally driven emergence and submergence made possible by wave action is a critical prediction of the model. On wavedominated shores (wave height >> tidal range), all three zones are predicted to exist separately, but on tide-dominated shores (tidal range >> wave height) the wave zone is absent and the emergent and submergent tidal zones overlap substantially, forming the traditional \u201cintertidal zone\u201d. We conclude by incorporating time and space in the model to illustrate variability in the physical conditions and zonation on littoral shores. The wave:tide physical zonation model is a unifying framework that can facilitate our understanding of physical conditions on littoral shores whether tropical or temperate, marine or lentic.",
+    "v2_Abstract": "The effects of tides on littoral marine habitats are so ubiquitous that shorelines are commonly described as \u2018intertidal\u2019, whereas waves are considered a secondary factor that simply modifies the intertidal habitat. However mean significant wave height exceeds tidal range at many locations worldwide. Here we construct a simple sinusoidal model of coastal water level based on both tidal range and wave height. From the patterns of emergence and submergence predicted by the model, we derive four vertical shoreline benchmarks which bracket up to three novel, spatially distinct, and physically defined zones. The (1) emergent tidal zone is characterized by tidally driven emergence in air; the (2) wave zone is characterized by constant (not periodic) wave wash; and the (3) submergent tidal zone is characterized by tidally driven submergence. The decoupling of tidally driven emergence and submergence made possible by wave action is a critical prediction of the model. On wave-dominated shores (wave height >> tidal range), all three zones are predicted to exist separately, but on tide-dominated shores (tidal range >> wave height) the wave zone is absent and the emergent and submergent tidal zones overlap substantially, forming the traditional \u201cintertidal zone\u201d. We conclude by incorporating time and space in the model to illustrate variability in the physical conditions and zonation on littoral shores. The wave:tide physical zonation model is a unifying framework that can facilitate our understanding of physical conditions on littoral shores whether tropical or temperate, marine or lentic.",
+    "url": "https://peerj.com/articles/155/reviews/",
+    "review_1": "Mikhail Gelfand \u00b7 Aug 18, 2013 \u00b7 Academic Editor\nACCEPT\nWhile I still think that some estimate of statistical significance could improve the paper, I do not insist.",
+    "review_2": "Mikhail Gelfand \u00b7 Aug 10, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nThe revised manuscript still does not address the issue of statistical significance.\nEditorial: the sentence \u201cWhile the information about the speed of processing that EEG exceeds that of fMRI, information about the spatial distribution of function across brain regions pales comparatively\u201d seems to be incomplete.\nUnfortunately, the revised manuscript did not contain tracked changes, as it has been suggested in the decision letter.",
+    "review_3": "Mikhail Gelfand \u00b7 Jul 14, 2013 \u00b7 Academic Editor\nMAJOR REVISIONS\nIn addition to the reviewers' comments, the following points should be taken into account.\nOne major issue not addressed in the text is disambiguation of homonymous author names. Chimeric \u201cauthors\u201d assigned with papers (MeSH terms etc) by different people may (or may not) severely confound the analysis. This needs to be addressed or at least discussed.\nStatements such as \u201c(something is) climbing linearly\u201d need to be supplied with a calculation of statistical significance or re-phrased.\nThe Methods session is not sufficiently detailed, contains many implicit assumptions and undefined terms, does not provide motivation for specific analyses. E.g. why calculation of node centrality (ca. line 98) is relevant in the context of the study? why pruning is sufficient to counteract the Matthew effect (footnote note 1 to line 90)?\nThe Introduction lacks references to many relevant papers on similar bibliometric studies. In fact, the manuscript shows almost no connection to previous work.\nThe term \u201clow-level research\u201d (line 150) sounds denigrating, although probably not intentionally.\nHow is correction for years since publication applied (if any)? \u2013 this is relevant when, e.g., most frequently cited papers are discussed (line 155).\nTrends for NiMCS papers are not compared with trends in similar or generic areas: hence, no comparison can be made and no conclusions formed.",
+    "review_4": "Ahmad Aziz \u00b7 Jul 8, 2013\nBasic reporting\n- In order to aid the reader in understanding the full scope of the article please provide a short description of the techniqual terms such as 'ForceAtlas2 algorithm' and Louvain modularity.\nExperimental design\n- Do the authors have also searched PubMed in 'free text' mode? This is important as not all relevant/recent articles are MeSh indexed.\n- Choosing to include only open acces articles may introduce a bias as the 'top gear'\njournals are unfortunately usually not open access. However, I understand very well\nthat this method is likely dictated by practical considerations. The authors should\ntherefore at least mention this point in the discussion section of the article.\n- The authors note that in the field of NiMCH the average number of authors is gradually increasing. Although this is true, it is not an exception as the average\nnumber of authors on biomedical papers in general is gradually increasing as more\nand more researchers begin to collaborate.\nValidity of the findings\n- The authors should include a paragraph in the discussion section of their manuscript elaborating on the potential limitations of their method. In particular it is important to stress that bibliometric analysis is only a very indirect measure of the impact and direction of a field, for example, as it is necessarily based on past results, In other words, the most recent ideas, conceps and results are simply waiting to be published.\nCite this review as\nAziz A (2013) Peer Review #1 of \"Publication trends in neuroimaging of minimally conscious states (v0.1)\". PeerJ https://doi.org/10.7287/peerj.155v0.1/reviews/1",
+    "pdf_1": "https://peerj.com/articles/155v0.3/submission",
+    "pdf_2": "https://peerj.com/articles/155v0.2/submission",
+    "review_5": "Reviewer 2 \u00b7 Jul 3, 2013\nBasic reporting\nExcellent-- a scholarly presentation\nExperimental design\nStudy used appropriate, and widely accepted bibliometric methods in their design\nValidity of the findings\nFindings are rather straightforward and corroborate prior research\nAdditional comments\nI wish that the author(s) had expanded in their discussion , in more detail, regarding the specific topical areas in Neuro-Imaging or MCS that were highly ranked in the analysis. The reasons why some topics are emphasized (& others not) should be explored.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"Publication trends in neuroimaging of minimally conscious states (v0.1)\". PeerJ https://doi.org/10.7287/peerj.155v0.1/reviews/2",
+    "pdf_3": "https://peerj.com/articles/155v0.1/submission",
+    "all_reviews": "Review 1: Mikhail Gelfand \u00b7 Aug 18, 2013 \u00b7 Academic Editor\nACCEPT\nWhile I still think that some estimate of statistical significance could improve the paper, I do not insist.\nReview 2: Mikhail Gelfand \u00b7 Aug 10, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nThe revised manuscript still does not address the issue of statistical significance.\nEditorial: the sentence \u201cWhile the information about the speed of processing that EEG exceeds that of fMRI, information about the spatial distribution of function across brain regions pales comparatively\u201d seems to be incomplete.\nUnfortunately, the revised manuscript did not contain tracked changes, as it has been suggested in the decision letter.\nReview 3: Mikhail Gelfand \u00b7 Jul 14, 2013 \u00b7 Academic Editor\nMAJOR REVISIONS\nIn addition to the reviewers' comments, the following points should be taken into account.\nOne major issue not addressed in the text is disambiguation of homonymous author names. Chimeric \u201cauthors\u201d assigned with papers (MeSH terms etc) by different people may (or may not) severely confound the analysis. This needs to be addressed or at least discussed.\nStatements such as \u201c(something is) climbing linearly\u201d need to be supplied with a calculation of statistical significance or re-phrased.\nThe Methods session is not sufficiently detailed, contains many implicit assumptions and undefined terms, does not provide motivation for specific analyses. E.g. why calculation of node centrality (ca. line 98) is relevant in the context of the study? why pruning is sufficient to counteract the Matthew effect (footnote note 1 to line 90)?\nThe Introduction lacks references to many relevant papers on similar bibliometric studies. In fact, the manuscript shows almost no connection to previous work.\nThe term \u201clow-level research\u201d (line 150) sounds denigrating, although probably not intentionally.\nHow is correction for years since publication applied (if any)? \u2013 this is relevant when, e.g., most frequently cited papers are discussed (line 155).\nTrends for NiMCS papers are not compared with trends in similar or generic areas: hence, no comparison can be made and no conclusions formed.\nReview 4: Ahmad Aziz \u00b7 Jul 8, 2013\nBasic reporting\n- In order to aid the reader in understanding the full scope of the article please provide a short description of the techniqual terms such as 'ForceAtlas2 algorithm' and Louvain modularity.\nExperimental design\n- Do the authors have also searched PubMed in 'free text' mode? This is important as not all relevant/recent articles are MeSh indexed.\n- Choosing to include only open acces articles may introduce a bias as the 'top gear'\njournals are unfortunately usually not open access. However, I understand very well\nthat this method is likely dictated by practical considerations. The authors should\ntherefore at least mention this point in the discussion section of the article.\n- The authors note that in the field of NiMCH the average number of authors is gradually increasing. Although this is true, it is not an exception as the average\nnumber of authors on biomedical papers in general is gradually increasing as more\nand more researchers begin to collaborate.\nValidity of the findings\n- The authors should include a paragraph in the discussion section of their manuscript elaborating on the potential limitations of their method. In particular it is important to stress that bibliometric analysis is only a very indirect measure of the impact and direction of a field, for example, as it is necessarily based on past results, In other words, the most recent ideas, conceps and results are simply waiting to be published.\nCite this review as\nAziz A (2013) Peer Review #1 of \"Publication trends in neuroimaging of minimally conscious states (v0.1)\". PeerJ https://doi.org/10.7287/peerj.155v0.1/reviews/1\nReview 5: Reviewer 2 \u00b7 Jul 3, 2013\nBasic reporting\nExcellent-- a scholarly presentation\nExperimental design\nStudy used appropriate, and widely accepted bibliometric methods in their design\nValidity of the findings\nFindings are rather straightforward and corroborate prior research\nAdditional comments\nI wish that the author(s) had expanded in their discussion , in more detail, regarding the specific topical areas in Neuro-Imaging or MCS that were highly ranked in the analysis. The reasons why some topics are emphasized (& others not) should be explored.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"Publication trends in neuroimaging of minimally conscious states (v0.1)\". PeerJ https://doi.org/10.7287/peerj.155v0.1/reviews/2\nReview 6: \n",
+    "LLM_extracted_review": "[\"Review 1: While I still think that some estimate of statistical significance could improve the paper, I do not insist.\\n\\nReview 2: The revised manuscript still does not address the issue of statistical significance. The sentence \u201cWhile the information about the speed of processing that EEG exceeds that of fMRI, information about the spatial distribution of function across brain regions pales comparatively\u201d seems to be incomplete. Unfortunately, the revised manuscript did not contain tracked changes, as it has been suggested in the decision letter.\\n\\nReview 3: One major issue not addressed in the text is disambiguation of homonymous author names. Chimeric \u201cauthors\u201d assigned with papers (MeSH terms etc) by different people may (or may not) severely confound the analysis. Statements such as \u201c(something is) climbing linearly\u201d need to be supplied with a calculation of statistical significance or re-phrased. The Methods session is not sufficiently detailed, contains many implicit assumptions and undefined terms, does not provide motivation for specific analyses. The Introduction lacks references to many relevant papers on similar bibliometric studies. The term \u201clow-level research\u201d sounds denigrating, although probably not intentionally. How is correction for years since publication applied (if any)? Trends for NiMCS papers are not compared with trends in similar or generic areas.\\n\\nReview 4: Please provide a short description of the technical terms such as 'ForceAtlas2 algorithm' and Louvain modularity. Do the authors have also searched PubMed in 'free text' mode? Choosing to include only open access articles may introduce a bias as the 'top gear' journals are usually not open access. The authors should mention this point in the discussion section. The authors note that in the field of NiMCH the average number of authors is gradually increasing, but this is not an exception as it is true for biomedical papers in general. The authors should include a paragraph in the discussion section elaborating on the potential limitations of their method, stressing that bibliometric analysis is only a very indirect measure of the impact and direction of a field.\\n\\nReview 5: Findings are rather straightforward and corroborate prior research. I wish that the author(s) had expanded in their discussion regarding the specific topical areas in Neuro-Imaging or MCS that were highly ranked in the analysis. The reasons why some topics are emphasized (& others not) should be explored.\"]"
+}

peerj_json_files/PeerJ_Json_112.json ADDED Viewed

The diff for this file is too large to render. See raw diff

peerj_json_files/PeerJ_Json_113.json ADDED Viewed

	@@ -0,0 +1,16 @@

+{
+    "v1_Abstract": "Background. The affective profile model categorizes individuals as self-fulfilling (high positive affect, low negative affect), high affective (high positive affect, high negative affect), low affective (low positive affect, low negative affect), and self-destructive (low positive affect, high negative affect). The model has been used extensively among Swedes to discern differences between profiles regarding happiness, depression, and also life satisfaction. The aim of the present study was to investigate such differences in a sample of residents of the USA. The study also investigated differences between profiles with regard to happiness-increasing strategies. Methods. In Study I, 900 participants reported affect (Positive Affect Negative Affect Schedule; PANAS) and happiness (Happiness-Depression Scale). In Study II, 500 participants self-reported affect (PANAS), life satisfaction (Satisfaction With Life Scale), and how often they used specific strategies to increase their own happiness (Happiness-Increasing Strategies Scales). Results. The results showed that, compared to the other profiles, self-fulfilling individuals were less depressed, happier, and more satisfied with their lives. Nevertheless, self-destructive were more depressed, unhappier, and less satisfied that all other profiles. The self-fulfilling individuals tended to use strategies related to agentic (e.g., instrumental goal-pursuit), communal (e.g., social affiliation), and spiritual (e.g., religion) values when pursuing happiness. Conclusion. These differences suggest that promoting positive emotions can positively influence a depressive-to-happy state as well as increasing life satisfaction. Moreover, the present study shows that pursuing happiness through strategies guided by agency, communion, and spirituality is related to a self-fulfilling experience described as high positive affect and low negative affect.",
+    "v1_col_limitations": "limitations and future research : One limitation of the present set of studies is that the results are based on MTurk workers\u2019 selfreports. Nevertheless, consistent with earlier research suggesting MTurk as a valid tool for collecting data using personality scales (Buhrmester et al., 2011), other researchers have found that health measures using MTurk data shows satisfactory internal reliability and test-retest reliability (Shapiro, Chandler & Mueller, 2013). Furthermore, the prevalence of depression among MTurk workers matches the prevalence of this illness in the general population; which makes MTurk a valid tool even for clinical research (Shapiro et al., 2013). The measures used here are validated and reliable measures of happiness, depression, life satisfaction, and affect; however, there are other established measures that could have been used for the measurement of depression (e.g., The Patient Health Questionnaire; Kroenke, Spitzer & Williams, 2001). The Short Depression-Happiness Scale (Joseph et al., 2004), used in Study I, was found appropriate firstly because it was developed as a short easy-to-distribute scale based on the increasing awareness of the therapeutic potential of the positive psychological perspective (e.g. Cloninger, 2006, Joseph & Linley, 2004; Keyes & Lopez, 2002). This scale has shown good psychometric\n342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366\nPeerJ reviewing PDF | (v2013:07:649:1:1:NEW 15 Aug 2013)\nR ev ie w in g M an\nus cr ip t\n17\nproperties of internal consistency reliability (Cronbach\u2019s \u03b1 between .77-92), test-retest reliability (r = .68 in a 2-week interval), and convergent and discriminant validity with measures of depression (Beck\u2019s Depression Inventory), happiness (Oxford\u2019s Happiness Inventory) and personality (NEO Five Factor Inventory) (Joseph et al., 2004).\nThe lack of studies in adult populations using the affective profiles model and positive\nmeasures of well-being did not permit comparison of the results presented to other than earlier research among adolescents and young adults, thus, showing the need for further studies on adults regarding these factors. The reliability coefficients for some of the happiness-increasing strategies were low (e.g., Direct Attempts showed an Cronbach\u2019s alpha = .56). In studies among Swedes this scales have been modified through factor analyses (Nima et al., 2013). Although most of the scales in the present study showed alphas above .63, further studies focusing in the validation of these scales are needed. Furthermore, specific emotions vary widely across the lifespan. Findings among men and women in the US, for example, show that as people age they become less stressed and angry, although worry seems to persist as a negative emotion in peoples lives during middle age (Stone, Schwartz, Broderick & Deaton, 2010). Positive emotions such as happiness and enjoyment along with negative emotions such as sadness, however, show very limited change with age (Stone et al., 2010). Although the present study did not aim to investigate variations in specific emotions with respect to age, further studies exploring increases/decreases in PA and NA are needed.\nFinally, since median splits distort the meaning of high and low, it is plausible to criticize\nthe validity of the procedure used here to create the different affective profiles scores just-above\nand just-below the median become high and low by fiat, not by reality (Sch\u00fctz, Archer & Garcia, 2013). Nevertheless, a recent study (MacDonald & Kormi-Nouri, 2013) used k-means cluster analysis to test if the affective profiles model emerged as theorized by Archer and colleagues. The\n367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386\n387\n388 389 390\nPeerJ reviewing PDF | (v2013:07:649:1:1:NEW 15 Aug 2013)\nR ev ie w in g M an\nus cr ip t\n18\naffective profile model was replicated using the k-means cluster analysis and the four affective profiles emerged as the combinations of high vs. low affectivity. The procedure used by these researchers is useful for person-oriented analyses (see Bergman, Magnusson et al., 2003), thus, suggesting the original procedure by Archer as valid.",
+    "v2_Abstract": "Background. The affective profile model categorizes individuals as self-fulfilling ( high positive affect, low negative affect), high affective (high positive affect, high negative affect), low affective (low positive affect, low negative affect), and self-destructive (low positive affect, high negative affect). The model has been used extensively among Swedes to discern differences between profiles regarding happiness, depression, and also life satisfaction. The aim of the present study was to investigate such differences in a sample of residents of the USA. The study also investigated differences between profiles with regard to happiness-increasing strategies. Methods. In Study 1, 900 participants reported affect (Positive Affect Negative Affect Schedule; PANAS) and happiness (Happiness-Depression Scale). In Study 2, 500 participants self-reported affect (PANAS), life satisfaction (Satisfaction With Life Scale), and how often they used specific strategies to increase their own happiness (Happiness-Increasing Strategies Scales) Results. The results showed that, compared to the other profiles, self-fulfilling individuals were less depressed, happier, and more satisfied with their lives. Nevertheless, self-destructive were more depressed, unhappier, and less satisfied that all other profiles. The self-fulfilling individuals tended to use strategies related to agentic (e.g., instrumental goal-pursuit), communal (e.g., social affiliation), and spiritual (e.g., religion) values when pursuing happiness. Conclusion. These differences suggest that promoting positive emotions can positively influence a depressive-to-happy state as well as increasing life satisfaction. Moreover, the present study shows that pursuing happiness through strategies guided by agency, communion, and spirituality is related to a self-fulfilling experience described as high positive affect and low negative affect.",
+    "v1_text": "results and discussion : A Multiple Analysis of Variance (MANOVA) indicated a significant effect for gender (F(4, 889) = 4.32; p = .002, Eta2 = 0.02, power = 0.93) as well as for affective profile (F(12, 2673) = 162.19; p < .001, Eta2 = 0.42, power = 1.00). The interaction of gender and affective profile was not significant (p = .236). A between-subjects ANOVA showed an significant gender effects for happiness (F(1, 892) = 7.60; p = 0.006), whereby the female participants expressed a higher level of happiness (M = 9.66, SD = 2.13) than the male participants (M =9.35, SD = 2.33). A between-subject ANOVA indicated significant affective profile effects for PA (F(3, 892) = 513.78; p < .001), NA (F(3, 892) = 503.58; p < .001), happiness (F(3, 892) = 68.20; p < .001), and depression (F(3, 892) = 71.50; p < .001). A Bonferroni correction to the alpha level of .01 showed that the self-destructive group had significantly higher scores in NA and depression as well as lower scores in happiness in comparison to the other affective profiles. The self-fulfilling group differed significantly from the self-destructive profiles in all measured variables; PA, NA, happiness and depression. As expected, the high affective ones differed significantly from the self-fulfilling group in all variables except PA and the low affective ones differed significantly from the self-fulfilling group in all variables except NA. Which is not so strange since both the self-fulfilling group and the high affective group are characterized as high in PA and the same goes for self-fulfilling individuals and low affective individuals who are characterized by low NA. For further details, see table 1. Table 1 here Study II First a MANOVA (3 x 2 factorial design) was applied with affective profiles and gender as independent variables and with PA, NA and life satisfaction as dependent variables. The analysis did not indicate any significant interaction effect (p = 0.14), but did indicate a significant effect for gender (F(3, 490) = 4.91; p < 0.01, Eta2 = 0.03, power = 0.91) as well as for affective profiles (F(9, 1476) = 119.15; p < 0.001, Eta2 = 0.42, power = 1.00). Secondly, a MANOVA (1 x 2 factorial design) was applied with affective profiles and gender as independent variables and with happiness-increasing strategies as dependent variables. The analysis did not indicate any significant interaction effect (p = 0.93), but did indicate a significant effect for gender (F(8, 485) = 5.85; p < 0.001, Eta2 = 0.09, power = 1.00) as well as for affective profiles (F(24, 1461) = 8.64; p < 0.001, Eta2 = 0.12, power = 1.00). A between-subjects ANOVA was conducted in order to test gender differences in PA, NA and life satisfaction. The result indicated significant gender effects for: NA (F(1, 492) = 10.89; p 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 PeerJ reviewing PDF | (v2013:07:649:1:1:NEW 15 Aug 2013) R ev ie w in g M an us cr ip t 12 <0.01), whereby the female participants expressed a higher level of NA (M = 1.94, SD = 0.83) than the male participants (M =1.72, SD = 0.67). This specific result stands in contrast to the results from Study I, which showed that females reported higher happiness than males. Nevertheless, this is a well-known paradox in the literature females seem to experience positive and negative emotions equally intensive, explaining why female often report both experiencing more negative moods and depressive symptoms and also higher levels of happiness than males (Fujita, Diener & Sandvik, 1991). A between-subjects ANOVA was conducted to investigate gender differences in happiness-increasing strategies. The result indicated significant gender effects for: Social Affiliation (F(1, 492) = 17.67; p <0.001), whereby the female participants expressed a higher level of Social Affiliation (M = 3.43, SD = 0.56) than the male participants (M =3.27, SD = 0.65); Instrumental Goal Pursuit (F(1, 492) = 6.60; p <0.01), whereby the female participants expressed a higher level of Instrumental Goal Pursuit (M = 3.33, SD = 0.81) than the male participants (M =3.19, SD = 0.82); Religion (F(1, 492) = 23.18; p <0.001), whereby the female participants expressed a higher Religion (M = 3.08, SD = 1.13) than the male participants (M =2.63, SD = 1.04); Passive Leisure (F(1, 492) = 9.25; p <0.01), whereby the female participants expressed a higher level of Passive Leisure (M = 3.30, SD = 0.55) than the male participants (M =3.16, SD = 0.60); Direct Attempts (F(1, 492) = 4.06; p <0.05), whereby the female participants expressed a higher level of Direct Attempts (M = 3.66, SD = 0.58) than the male participants (M =3.60, SD = 0.64). The differences presented here are a replication of the original study conducted by Tkach and Lyubomirsky (2006): females focus on behaviour such as maintaining relationships (i.e., Social Affiliation), pursuing career goals (i.e., Instrumental Goal Pursuit), performing religious activities (i.e., Religion), and watching TV (i.e., Passive Leisure) more frequently than males when they try to increase their happiness. As suggested by Tkach and Lyubomirsky (2006, pp. 214), the gender differences replicated here \u201care consistent with the 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 PeerJ reviewing PDF | (v2013:07:649:1:1:NEW 15 Aug 2013) R ev ie w in g M an us cr ip t 13 gender differences reported for behaviors used to combat bad moods (Thayer et al., 1994)\u201d. In order to test differences in life satisfaction for each of the four affective profiles a between-subject ANOVA was conducted The result indicated significant effects for life satisfaction (F(3, 492) = 49.26; p <0.001). Further, a between-subject ANOVA was conducted in order to test differences in happiness-increasing strategies for each of the four affective profiles. The mean scores of life satisfaction as well as for happiness-increasing strategies for all four affective profiles are presented in Table 2. Table 2 here A Bonferroni test, with alpha level set to .01, was conducted to compare the mean differences in life satisfaction as well as for happiness-increasing strategies between affective profiles. The results showed, replicating earlier findings, among Swedes, that that the selfdestructive group had lower scores in life satisfaction compared to all the other affective profiles. The self-fulfilling group had higher scores in life satisfaction compared to all the other affective profiles. Regarding happiness-increasing strategies the results showed that that the selfdestructive group had lower scores in all happiness-increasing strategies except for Mental Control. For further details, see Table 3. Table 3 here General discussion The aim of this set of studies was to examine the connections between the four types of affective profiles (self-fulfilling, high affective, low affective, self-destructive) to happiness and depression (Study I), satisfaction with life and happiness-increasing strategies (Study II) in US-residents. The results showed that the self-fulfilling group reported a significantly higher level of happiness and a significantly lower level of depression than all the three other groups (high affective, low affective, self-destructive). Furthermore, the self-destructive group reported a significantly higher 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 PeerJ reviewing PDF | (v2013:07:649:1:1:NEW 15 Aug 2013) R ev ie w in g M an us cr ip t 14 level of depression and lower level of happiness than all the other three groups (self-fulfilling, high and low affective). The results also show that the high affective and low affective reported higher level of happiness and lower level of depression than the self-destructive group. But at the same time these groups (high and low affective) also showed significantly lower levels of happiness and significantly higher levels depression than the self-fulfilling group. As suggested by Garcia (2011), low PA among low affectives seems to influence happiness negatively as high NA influences happiness negatively among high affectives. The results presented here are corresponding to the results found in research with Swedish populations showing that high PA is related to less stress, depression, and anxiety (e.g., Garcia et al., 2012; Lindahl & Archer, 2013; Nima, Rosenberg, Archer & Garcia, 2013). Moreover, self-fulfilling, high affective and low affective participants all have higher life satisfaction compared with self-destructive participants. This result also replicates findings among Swedish pupils where self-fulfilling, high and low affective participants showed higher level of life satisfaction compared with self-destructives (e.g., Garcia & Archer, 2012). As suggested by Lindahl and Archer (2013; see also Archer & Kostrzewa, 2013; Archer, Oscar-Berman, Blum & Gold, 2013), positive affect might serve as an anti-depressive factor and, as suggested here, also as protective factor for happiness and life satisfaction. The self-fulfilling participants showed significantly higher results than all other profiles on the direct attempts strategy. Suggesting that in order to increase their happiness the selffulfilling individuals are more prone to directly attempt to smile, get them selves in a happy mood, improve their social skills, and work on their self-control. Indeed, Garcia (2012a) showed that self-fulfilling score higher in personality traits related to agentic values (i.e., autonomy, responsibility, self-acceptance, intern locus of control, self-control) as measured by the Temperament and Character Inventory (Cloninger, Svrakic & Przybeck, 1993). Moreover, selffulfilling individuals scored lower than high NA individuals (high affectives and self- 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 PeerJ reviewing PDF | (v2013:07:649:1:1:NEW 15 Aug 2013) R ev ie w in g M an us cr ip t 15 destructives) in the strategy of mental control. The mental control scale has been defined as ambivalent behavior, that is, the individual using this happiness-increasing strategy make efforts to avoid negative experiences by suppressing negative thoughts and feelings but also ruminating about negative aspects of life (Tkach and Lyubomirsky, 2006). These tendencies may not only prolong unhappiness, suppressing negative thoughts actually may end up in maintaining these thoughts and thereby aggravate negative affect (Tkach & Lyubomirsky, 2006), which may explain why these tendencies are more frequent among high affective and self-destructive than self-fulfilling individuals. Compared to low PA individuals (i.e., low affectives and self-destructives), the self- fulfilling individuals also reported using more often three of the other happiness-increasing strategies: social affiliation, instrumental goal pursuit, active leisure. Social affiliations activities comprise communal (i.e., cooperation) values to guide behavior such as: supporting and encouraging friends, helping others, trying to improve one self, interacting with friends, and receiving help from friends (Tkach & Lyubomirsky, 2006). Instrumental goal pursuit includes activities directed to achieving goals by trying to reach one\u2019s full potential, studying, organizing one\u2019s life and goals, and striving for the accomplishment of tasks (Tkach & Lyubomirsky, 2006). Finally, the use of active leisure comprises a proness to wellness through fitness and flow, that is, exercising and working on hobbies or activities in which the individual uses her/his strengths and becomes absorbed by the activity itself (Tkach & Lyubomirsky, 2006). In other words, both instrumental goal pursuit and active leisure comprises agentic (i.e., autonomous, self-directed) values guiding behavior in order to approach well-being. Indeed, among Swedes (Nima et al., 2012, 2013), these three strategies (social affiliation, instrumental goal pursuit, and active leisure) have been found to be positively related to subjective well-being. Agency and cooperation are also related to mental health, dysfunction and suffering (Cloninger & Zohar, 2011; Garcia, Anckars\u00e4ter & Lundstr\u00f6m, 2013; Garcia, Lundstrom, Brandstrom, Rastam, Cloninger, et al., \u0308 \u0308 \u0308 \u030a 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 PeerJ reviewing PDF | (v2013:07:649:1:1:NEW 15 Aug 2013) R ev ie w in g M an us cr ip t 16 2013; Garcia, Nima & Archer, 2013) and are suggested to help the individual to become happier and healthier (Cloninger, 2013; see also Johansson Lyssarides, Andersson & Rousseau, 2013, who showed that increases in agency and cooperation are associated to improvement in depression). Moreover, compared to the self-destructives, the self-fulfilling individuals reported more frequently seeking support from faith, performing religious activities, praying, and drinking less alcohol (i.e., the religion happiness-increasing strategy). Indeed, Cloninger (2013) has suggested that while agency and cooperation might lead to happiness and health, spiritual values might be needed for becoming a self-fulfilled individual that lives in harmony with the changing world. See Figure 1 for a summary of the results. Figure 1 should be here statistical treatment : We used participants\u2019 self-reported affect measured by the PANAS from both Study I and 2 (N = 1,400) in order to classify participants in the four affective profiles. Participants\u2019 PA and NA scores were divided into high and low (cut-off points: low PA = 3.0 or less; high PA = 3.1 or above; low NA = 1.8 or less; and high NA = 1.9 or above). For Study I, the two independent variables of the study were gender and affective profile: self-fulfilling (n = 241; 153 males, 88 females), low affective (n = 236; 137 males, 99 females), high affective (n = 180; 115 males, 65 females), and self-destructive (n = 243; 145 males, 98 females). The dependent variables were PA, NA, happiness, and depression. 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 PeerJ reviewing PDF | (v2013:07:649:1:1:NEW 15 Aug 2013) R ev ie w in g M an us cr ip t 9 As detailed in Study I, both samples were used in the classification of the four affective profiles. 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 PeerJ reviewing PDF | (v2013:07:649:1:1:NEW 15 Aug 2013) R ev ie w in g M an us cr ip t 11 The number of participants in each profile for Study II were as follows: 158 self-fulfilling (75 males, 83 females), 92 low affective (42 males, 50 females), 123 high affective (54 males, 69 females), and 127 self-destructive (46 males, 81 females). The affective profiles and gender were the independent variables, PA, NA, life satisfaction, and the happiness-increasing strategies were the dependent variables. An important observation here is the gender distribution between profiles. For example, in Study I there were more self-destructive males than females, while in Study II there were more self-destructive females than males. This difference might mirror the gender distribution across Study I (550 males and 350 females) and Study II (217 male and 283 females). Across both samples of females, the prevalence of the self-destructive profile was 28%, while among men was 25%. The prevalence of this profile reported here among males and females is the same that was observed among Swedes (Sch\u00fctz, Garcia & Archer, 2013). participants and procedure : The participants (N = 900, age mean = 28.72 sd. = 19.10, 550 males and 350 females) were USresidents recruited through Amazons\u2019 Mechanical Turk (MTurk; https://www.mturk.com/mturk/welcome). MTurk allows data collectors to recruit participants (workers) online for completing different tasks in change for wages. This method for data collection online has become more common during recent years and it is an empirical tested valid tool for conducting research in the social sciences (see Buhrmester, Kwang & Gosling, 2011). Participants were recruited by the following criteria: US-resident and to both speak and write fluent in English. Participants were paid a wage of two American dollars for completing the task 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 PeerJ reviewing PDF | (v2013:07:649:1:1:NEW 15 Aug 2013) R ev ie w in g M an us cr ip t 8 and informed that the study was confidential and voluntary. The participants were presented with a battery of self-reports comprising the affect and happiness measures, as well as questions pertaining age and gender. As in Study I, participants (N = 500, age mean = 34.08 sd. =12.55; 217 male and 283 female) 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 PeerJ reviewing PDF | (v2013:07:649:1:1:NEW 15 Aug 2013) R ev ie w in g M an us cr ip t 10 were recruited from MTurk by the following criteria: resident of the USA and to both speak and write fluent in English. Participants were paid a wage of two American dollars for completing the task and informed that the study was confidential and voluntary. The participants were presented with a battery of self-reports comprising the affect, life satisfaction, and happiness-increasing strategies measures, as well as questions pertaining age and gender. ethics statement : This research protocol was approved by the Ethics Committee of the University of Gothenburg and written informed consent was obtained from all the study participants. conclusion : The present set of studies expands earlier results among Swedes to a relative large sample of USresidents. The results suggest that the affective profile model distinguish important differences in happiness, depression, and life satisfaction between individuals. These differences suggest that promoting positive emotions can positively influence a depressive-to-happy state as well as increasing life satisfaction. Moreover, the present study describes further how affective profiles differ with regard to happiness-increasing strategies. These specific results suggest that the pursue of happiness through agentic, communal, and spiritual values leads to a self-fulfilling experience defined as frequently experiencing positive emotions and infrequently experiencing negative emotions. \u201cIt was right then that I started thinking about legends and captions : Table 1. Mean scores in PA, NA, happiness, and depression for each affective profile in Study I. Table 2. Means in life satisfaction and happiness-increasing strategies among affective profiles in Study II. Table 3. Mean differences in life satisfaction and happiness-increasing strategies between affective profiles. Figure 1. Summary of the results from Study I and II showing the differences between affective profiles in happiness, depression, life satisfaction, and the happiness-increasing strategies. 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 PeerJ reviewing PDF | (v2013:07:649:1:1:NEW 15 Aug 2013) R ev ie w in g M an us cr ip t 20 happiness-increasing strategies : In order to intentionally pursue happiness, people seem to use different strategies. Tkach and Lyubomirsky (2006) have identified, using first an open-ended survey, 53 happiness-increasing strategies used by residents of the USA (for studies using this scale among Swedes see, Garcia, 2012b; Nima, Archer and Garcia, 2012, 2013). Tkach and Lyubomirsky (2006) found, using 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 PeerJ reviewing PDF | (v2013:07:649:1:1:NEW 15 Aug 2013) R ev ie w in g M an us cr ip t 6 factor analysis, eight factors: Social Affiliation (e.g. \u2018\u2018Support and encourage friends\u2019\u2019), Partying and Clubbing (e.g. \u2018\u2018Drink alcohol\u2019\u2019), Mental Control (e.g. \u2018\u2018Try not to think obout being unhappy\u2019\u2019), Instrumental Goal Pursuit (e.g. \u2018\u2018Study\u2019\u2019), Passive Leisure (e.g. \u2018\u2018Surf the internet\u2019\u2019), Active Leisure (e.g. \u2018\u2018Exercise\u2019\u2019), Religion (e.g. \u2018\u2018Seek support from faith\u2019\u2019) and Direct Attempts (e.g. \u2018\u2018Act happy/smile, etc.\u2019\u2019). Results have shown that these happiness-increasing strategies accounted for 52% of the variance in happiness, while the Big Five personality traits, which traditionally have been linked to happiness, accounted for 46%. Further, even after controlling for the contribution of personality, the happiness-increasing strategies accounted for 16% of the variance in happiness. However, these relationships might not be a direct one. For example, Extraversion, which is strongly related to high PA (Larsen & Ketelaar, 1991), is related to the use of the Social Affiliation strategy, which, in turn, is related to happiness. Tkach and Lyubomirsky (2006) suggested that the efficacy of the happiness-increasing strategies is also likely to vary to some extent. However, the strategy that was the most robust predictor of low levels of happiness was Mental Control, which was closely related to Neuroticism. This strategy is defined as ambivalent intentional efforts aimed, on one side and avoidance of negative thoughts and feelings as well as proneness towards contemplation of negative aspects of life on the other. Regarding the affective profiles, if the profiles differ in the way they pursue happiness (i.e., approaching happy experiences versus preventing unhappy experiences), then it could be expected that the profiles differ in the use of the strategies described here. For example, it could be expected that high PA profiles should score higher in strategies such as Social Affiliation and Active Leisure due to the close positive relationship between Extraversion and PA. High NA profiles could be expected to score higher in strategies such as Mental Control, because the positive relationship between Neuroticism and NA. The present study 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 PeerJ reviewing PDF | (v2013:07:649:1:1:NEW 15 Aug 2013) R ev ie w in g M an us cr ip t 7 To the best of our knowledge, the affective profiles model has been mostly used among Swedish participants. Some cultures explain the world as good and controllable and others emphasize negative emotions as normal (Myers & Diener, 1995; Diener, Suh and Oishi, 1997). In this context, it is interesting noticing that the right to pursuit individual happiness is listed as an absolute right in the United States of America\u2019s declaration of independence (Tkach and Lyubomirsky, 2006). The model, however, has shown identical results in the few studies using other populations (for three studies using Dutch, Indonesian, respectively Iranian participants see Kunst, 2011; Adrianson, Djamaludin, Neila and Archer, 2013; Garcia and Moradi, 2013). The aim of the present study was to investigate differences in happiness, depression, life satisfaction and use of strategies to increase happiness among affective profiles in residents of the United States of America (US-residents). Study I method : instruments : Positive Affect and Negative Affect Schedule (PANAS; Watson et al., 1988). The PANAS instructs participants to rate to what extent they generally have experienced 20 different feelings or emotions (10 PA and 10 NA) during the last weeks, using a 5-point Likert scale (1 = very slightly, 5 = extremely). The 10\u2013item PA scale includes adjectives such as strong, proud, and interested. The 10\u2013 item NA scale includes adjectives such as afraid, ashamed and nervous. Cronbach\u2019s \u03b1 were .87 for PA and .89 for NA in the present study. The Short Depression-Happiness Scale (Joseph et al., 2004). This instrument consists of six items, three items measuring happiness (e.g., \u201cI felt happy\u201d) and three reverse coded items measuring depressive states (e.g., \u201cI felt my life was meaningless\u201d). Participants rate how frequently they feel the way described in the item on a four-point scale: \u201cnever\u201d, \u201crarely\u201d, \u201csometimes\u201d, \u201coften\u201d. In the present study, Cronbach\u2019s \u03b1 was .85 for the happiness scale and .76 for the depression scale. The same instrument as in Study I was used in Study II to measure PA and NA (i.e., the PANAS). Cronbach\u2019s \u03b1 were .88 for PA and .90 for NA in Study II. Satisfaction with Life Scale (Diener, Emmons, Larsen and Griffin, 1985). The instrument consists of 5 statements (e.g., \u201cIn most of my ways my life is close to my ideal\u201d) for which participants are asked to indicate degree of agreement in a 7-point Likert scale (1 = strongly disagree, 7 = strongly agree). The life satisfaction score was established by summarizing the 5 statements for each participant. Cronbach\u2019s \u03b1 were .90 in the present study. Happiness-Increasing Strategies Scales (Tkach and Lyubomirsky, 2006). In the present study, participants were asked to rate (1 = never, 7 = all the time) how often they used the strategies identified by Tkach and Lyubomirsky (2006). The happiness-increasing strategies are organized in eight clusters: Social Affiliation (e.g., \u2018\u2018Support and encourage friends\u2019\u2019; Cronbach\u2019s \u03b1 = 0.79), Partying and Clubbing (e.g., \u2018\u2018Drink alcohol\u2019\u2019; Cronbach\u2019s \u03b1 = 0.74), Mental Control (e.g., \u2018\u2018Try not to think about being unhappy\u2019\u2019; Cronbach\u2019s \u03b1 = 0.43), Instrumental Goal Pursuit (e.g. \u2018\u2018Study\u2019\u2019; Cronbach\u2019s \u03b1 = 0.76), Passive Leisure (e.g. \u2018\u2018Surf the internet\u2019\u2019; Cronbach\u2019s \u03b1 = 0.63), Active Leisure (e.g. \u2018\u2018Exercise\u2019\u2019; Cronbach\u2019s \u03b1 = 0.65), Religion (e.g. \u2018\u2018Seek support from faith\u2019\u2019; Cronbach\u2019s \u03b1 = 0.70), and Direct Attempts (e.g. \u2018\u2018Act happy/smile, etc.\u2019\u2019; Cronbach\u2019s \u03b1 = 0.56). thomas jefferson on the declaration of independence : and the part about our right to life, liberty, and the pursuit of happiness. And I remember thinking how did he know to put the pursuit part in there?\u201d Will Smith as Christopher Gardner in The Pursuit of Happyness 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 PeerJ reviewing PDF | (v2013:07:649:1:1:NEW 15 Aug 2013) R ev ie w in g M an us cr ip t 19 affective profiles self-fulfilling : N = 158 High affective N = 123 Low affective N = 92 Self-destructive N = 127 self-fulfilling : Positive Affect 0.17* 1.32* 1.46* Negative Affect -1.12* 0.05ns -1.28* Life satisfaction 1.05* 0.75* 2.01* Social Affiliation 0.05ns 0.28* 0.54* Partying and Clubbing -0.16ns -0.06ns 0.12ns Mental Control -0.31* -0.09ns -0.47* Instrumental Goal Pursuit -0.04ns 0.39* 0.54* Religion 0.17ns 0.23ns 0.54* Passive Leisure -0.16ns 0.05ns 0.05ns Active Leisure 0.11ns 0.29* 0.49* Direct Attempts 0.24* 0.31* 0.64* High affective Positive Affect -0.17* 1.15* 1.29* Negative Affect 1.11* -1.17* -0.16ns Life satisfaction -1.05* -0.31ns 0.96* Social Affiliation -0.05ns 0.24ns 0.50* Partying and Clubbing 0.16ns 0.11ns 0.29* Mental Control 0.31* 0.23* -0.16ns Instrumental Goal Pursuit 0.04ns 0.43* 0.58* Religion -0.17ns 0.05ns 0.36ns Passive Leisure 0.16ns 0.21ns 0.20ns Active Leisure -0.11ns 0.18ns 0.38* Direct Attempts -0.23* 0.07ns 0.40ns Low affective Positive Affect -1.32* -1.15* 0.14ns Negative Affect -0.05 -1.17* -1.32* 1 2 PeerJ reviewing PDF | (v2013:07:649:1:1:NEW 15 Aug 2013) R ev ie w in g M an us cr ip t 34 Life satisfaction -0.75* 0.31ns 1.26* Social Affiliation -0.28* -0.24ns 0.26* Partying and Clubbing 0.06ns -0.11ns 0.18ns Mental Control 0.09ns -0.23* -0.40* Instrumental Goal Pursuit -0.39* -0.43* 0.15ns Religion -0.23ns -0.05ns 0.31ns Passive Leisure -0.05ns -0.21ns -0.00ns Active Leisure -0.29* -0.18ns 0.20ns Direct Attempts -0.31* -0.07ns 0.33* Self-destructive Positive Affect -1.46* -1.29* -0.14* Negative Affect 1.28* 0.16ns 1.33* Life satisfaction -2.01* -0.96* -1.26* Social Affiliation -0.54* -0.50* -0.26* Partying and Clubbing -0.12ns -0.29* -0.18ns Mental Control 0.47* 0.16ns 0.39* Instrumental Goal Pursuit -0.54* -0.58* -0.15ns Religion -0.54* -0.36ns -0.31ns Passive Leisure -0.05ns -0.20ns 0.00ns Active Leisure -0.49* -0.38* -0.20ns Direct Attempts -0.64* -0.40* -0.33* ns = non significant, * p < 0.01 with Bonferroni Correction.3 PeerJ reviewing PDF | (v2013:07:649:1:1:NEW 15 Aug 2013) R ev ie w in g M an us cr ip t 35 Figure 1 Figure 1 Summary of the results from Study I and II showing the differences between affective profiles in happiness, depression, life satisfaction, and the happiness-increasing strategies. PeerJ reviewing PDF | (v2013:07:649:1:1:NEW 15 Aug 2013) R ev ie w in g M an us cr ip t 36 PeerJ reviewing PDF | (v2013:07:649:1:1:NEW 15 Aug 2013) R ev ie w in g M an us cr ip t",
+    "v2_text": "results and discussion : First a MANOVA (3 x 2 factorial design) was applied with affective profiles and gender as independent variables and with PA, NA and life satisfaction as dependent variables. The analysis did not indicate any significant interaction effect (p = 0.14), but did indicate a significant effect for gender (F(3, 490) = 4.91; p < 0.01, Eta2 = 0.03, power = 0.91) as well as for affective profiles (F(9, 1476) = 119.15; p < 0.001, Eta2 = 0.42, power = 1.00). Secondly, a MANOVA (1 x 2 factorial design) was applied with affective profiles and gender as independent variables and with happiness-increasing strategies as dependent variables. The analysis did not indicate any significant interaction effect (p = 0.93), but did indicate a significant effect for gender (F(8, 485) = 5.85; p < 0.001, Eta2 = 0.09, power = 1.00) as well as for affective profiles (F(24, 1461) = 8.64; p < 0.001, Eta2 = 0.12, power = 1.00). A between-subjects ANOVA was conducted in order to test gender differences in PA, NA and life satisfaction. The result indicated significant gender effects for: NA (F(1, 492) = 10.89; p <0.01), whereby the female participants expressed a higher level of NA (M = 1.94, SD = 0.83) than the male participants (M =1.72, SD = 0.67). This specific result stands in contrast to the results from Study I which showed that females reported higher happiness than males. Nevertheless, this is a well-known paradox in the literaturefemales seem to experience positive and negative emotions equally intensive, explaining why female often report both experiencing more negative moods and depressive symptoms and also higher levels of happiness than males (Fujita, Diener & Sandvik, 1991). A between-subjects ANOVA was conducted to investigate gender differences in happiness-increasing strategies. The result indicated significant gender effects for: Social Affiliation (F(1, 492) = 17.67; p <0.001), whereby the female participants expressed a higher level of Social Affiliation (M = 3.43, SD = PeerJ reviewing PDF | (v2013:07:649:0:0:NEW 12 Jul 2013) R ev ie w in g M an us cr ip t 0.56) than the male participants (M =3.27, SD = 0.65); Instrumental Goal Pursuit (F(1, 492) = 6.60; p <0.01), whereby the female participants expressed a higher level of Instrumental Goal Pursuit (M = 3.33, SD = 0.81) than the male participants (M =3.19, SD = 0.82); Religion (F(1, 492) = 23.18; p <0.001), whereby the female participants expressed a higher Religion (M = 3.08, SD = 1.13) than the male participants (M =2.63, SD = 1.04); Passive Leisure (F(1, 492) = 9.25; p <0.01), whereby the female participants expressed a higher level of Passive Leisure (M = 3.30, SD = 0.55) than the male participants (M =3.16, SD = 0.60); Direct Attempts (F(1, 492) = 4.06; p <0.05), whereby the female participants expressed a higher level of Direct Attempts (M = 3.66, SD = 0.58) than the male participants (M =3.60, SD = 0.64). The differences presented here are a replication of the original study conducted by Tkach and Lyubomirsky (2006). In order to test differences in life satisfaction for each of the four affective profiles a between-subject ANOVA was conducted The result indicated significant effects for life satisfaction (F(3, 492) = 49.26; p <0.001). Further, a between-subject ANOVA was conducted in order to test differences in happiness-increasing strategies for each of the four affective profiles. The mean scores of life satisfaction as well as for happiness-increasing strategies for all four affective profiles are presented in Table 2. Table 2 here A Bonferroni test, with alpha level set to .01, was conducted to compare the mean differences in life satisfaction as well as for happiness-increasing strategies between affective profiles. The results showed, replicating earlier findings, among Swedes, that that the self-destructive group had lower scores in life satisfaction compared to all the other affective profiles. The self-fulfilling group had higher scores PeerJ reviewing PDF | (v2013:07:649:0:0:NEW 12 Jul 2013) R ev ie w in g M an us cr ip t in life satisfaction compared to all the other affective profiles. Regarding happiness-increasing strategies the results showed that that the self-destructive group had lower scores in all happiness-increasing strategies except for Mental Control. For further details, see Table 3. Table 3 here General discussion The aim of this set of studies was to examine the connections between the four types of affective profiles (self-fulfilling, high affective, low affective, self-destructive) to happiness and depression (Study I), satisfaction with life and happiness-increasing strategies (Study II) in US-residents. The results showed that the self-fulfilling group reported a significantly higher level of happiness and a significantly lower level of depression than all the three other groups (high affective, low affective, self-destructive). Furthermore, the self-destructive group reported a significantly higher level of depression and lower level of happiness than all the other three groups (self-fulfilling, high and low affective). The results also show that the high affective and low affective reported higher level of happiness and lower level of depression than the self-destructive group. But at the same time these groups (high and low affective) also showed significantly lower levels of happiness and significantly higher levels depression than the self-fulfilling group. As suggested by Garcia (2011), low PA among low affectives seems to influence happiness negatively as high NA influences happiness negatively among high affectives. The results presented here are corresponding to the results found in research with Swedish populations showing that high PA is related to less stress, depression, and anxiety (e.g., Garcia et al., 2012; Lindahl & Archer, 2013). Moreover, self-fulfilling, high affective and low affective participants all have higher life satisfaction compared with self-destructive PeerJ reviewing PDF | (v2013:07:649:0:0:NEW 12 Jul 2013) R ev ie w in g M an us cr ip t participants. This result also replicates findings among Swedish pupils where self-fulfilling, high and low affective participants showed higher level of life satisfaction compared with self-destructives (e.g., Garcia & Archer, 2012). As suggested by Lindahl and Archer (2013; see also Archer & Kostrzewa, 2013; Archer, Oscar-Berman, Blum & Gold, 2013), positive affect might serve as an anti-depressive factor and, as suggested here, also as protective factor for happiness and life satisfaction. The self-fulfilling participants showed significantly higher results than all other profiles on the direct attempts strategy. Suggesting that in order to increase their happiness the self-fulfilling individuals are more prone to directly attempt to smile, get them selves in a happy mood, improve their social skills, and work on their self-control. Indeed, Garcia (2012a) showed that self-fulfilling score higher in personality traits related to agentic values (i.e., autonomy, responsibility, self-acceptance, intern locus of control, self-control) as measured by the Temperament and Character Inventory (Cloninger, Svrakic & Przybeck, 1993). Moreover, self-fulfilling individuals scored lower than high NA individuals (high affectives and self-destructives) in the strategy of mental control. The mental control scale has been defined as ambivalent behavior, that is, the individual using this happiness-increasing strategy make efforts to avoid negative experiences by suppressing negative thoughts and feelings but also ruminating about negative aspects of life (Tkach and Lyubomirsky, 2006). These tendencies may not only prolong unhappiness, suppressing negative thoughts actually may end up in maintaining these thoughts and thereby aggravate negative affect (Tkach & Lyubomirsky, 2006), which may explain why these tendencies are more frequent among high affective and self-destructive than self-fulfilling individuals. PeerJ reviewing PDF | (v2013:07:649:0:0:NEW 12 Jul 2013) R ev ie w in g M an us cr ip t Compared to low PA individuals (i.e., low affectives and self-destructives), the self-fulfilling individuals also reported using more often three of the other happiness-increasing strategies: social affiliation, instrumental goal pursuit, active leisure. Social affiliations activities comprise communal (i.e., cooperation) values to guide behavior such as: supporting and encouraging friends, helping others, trying to improve one self, interacting with friends, and receiving help from friends (Tkach & Lyubomirsky, 2006). Instrumental goal pursuit includes activities directed to achieving goals by trying to reach one\u2019s full potential, studying, organizing one\u2019s life and goals, and striving for the accomplishment of tasks (Tkach & Lyubomirsky, 2006). Finally, the use of active leisure comprises a proness to wellness through fitness and flow, that is, exercising and working on hobbies or activities in which the individual uses her/his strengths and becomes absorbed by the activity itself (Tkach & Lyubomirsky, 2006). In other words, both instrumental goal pursuit and active leisure comprises agentic (i.e., autonomous, self-directed) values guiding behavior in order to approach well-being. Indeed, among Swedes (Nima et al., 2012, 2013), these three strategies (social affiliation, instrumental goal pursuit, and active leisure) have been found to be positively related to subjective well-being. Agency and cooperation are also related to mental health, dysfunction and suffering (Cloninger & Zohar, 2011; Garcia, Anckars\u00e4ter & Lundstr\u00f6m, 2013; Garcia, Lundstrom, Brandstrom, Rastam, \u0308 \u0308 \u0308 \u030a Cloninger, et al., 2013) and are suggested to help the individual to become happier and healthier (Cloninger, 2013; see also Johansson Lyssarides, Andersson & Rousseau, 2013, who showed that increases in agency and cooperation are associated to improvement in depression). Moreover, compared to the self-destructives, the self-fulfilling individuals reported more frequently seeking support from faith, performing religious activities, praying, and drinking less alcohol (i.e., the religion PeerJ reviewing PDF | (v2013:07:649:0:0:NEW 12 Jul 2013) R ev ie w in g M an us cr ip t happiness-increasing strategy). Indeed, Cloninger (2013) has suggested that while agency and cooperation might lead to happiness and health, spiritual values might be needed for becoming a self-fulfilled individual that lives in harmony with the changing world. See Figure 1 for a summary of the results. Figure 1 should be here Limitations and future research One major limitation is that the present set of studies was conducted using self-reports. Nevertheless, the measures used here are validated and reliable measures of happiness, depression, life satisfaction, and affect. Moreover, the lack of studies in adult populations using the affective profiles model and measures of well-being did not permit comparison of the results presented to other than earlier research among adolescents and young adults, thus, showing the need for further studies on adults regarding these factors. The reliability coefficients for some of the happiness-increasing strategies were low (e.g., Direct Attempts showed an Cronbach\u2019s alpha = .56). In studies among Swedes this scale has been modified through factor analyses (Nima et al., 2013). Although most of the scales in the present study showed alphas above .63, further studies focusing in the validation of these scales are needed. Finally, since median splits distort the meaning of high and low, it is plausible to criticize the validity of the procedure used here to create the different affective profilesscores just-above and just-below the median become high and low by fiat, not by reality (Sch\u00fctz, Archer & Garcia, 2013). Nevertheless, a recent study (MacDonald & Kormi-Nouri, 2013) used k-means cluster analysis to test if the affective profiles model emerged as theorized by Archer and colleagues. The affective profile model was replicated using the k-means cluster analysis and the four affective PeerJ reviewing PDF | (v2013:07:649:0:0:NEW 12 Jul 2013) R ev ie w in g M an us cr ip t profiles emerged as the combinations of high vs. low affectivity. The procedure used by these researchers is useful for person-oriented analyses (see Bergman, Magnusson et al., 2003), thus, suggesting the original procedure by Archer as valid. Conclusion The present set of studies expands earlier results among Swedes to a relative large sample of US-residents. The results suggest that the affective profile model distinguish important differences in happiness, depression, and life satisfaction between individuals. These differences suggest that promoting positive emotions can positively influence a depressive-to-happy state as well as increasing life satisfaction. Moreover, the present study describes further how affective profiles differ with regard to happiness-increasing strategies. Showing that agentic, communal, and spiritual values guide behaviour when self-fulfilling individuals pursue happiness. \u201cIt was right then that I started thinking about Thomas Jefferson on the Declaration of Independence and the part about our right to life, liberty, and the pursuit of happiness. And I remember thinking how did he know to put the pursuit part in there?\u201d Will Smith as Christopher Gardner in The Pursuit of Happyness PeerJ reviewing PDF | (v2013:07:649:0:0:NEW 12 Jul 2013) R ev ie w in g M an us cr ip t participants and procedure : The participants (N = 900, age mean = 28.72 sd. = 19.10, 550 males and 350 females) were US-residents recruited through Amazons\u2019 Mechanical Turk (MTurk; https://www.mturk.com/mturk/welcome). MTurk allows data collectors to recruit participants (workers) online for completing different tasks in change for wages. This method for data collection online has become more common during recent years and it is an empirical tested valid tool for conducting research in the social sciences (see Buhrmester, Kwang & Gosling, 2011). Participants were recruited by the following criteria: US-resident and to both speak and write fluent in English. Participants were paid a wage of two American dollars for completing the task and informed that the study was confidential and voluntary. The participants were presented with a battery of self-reports comprising the affect and happiness measures, as well as questions pertaining age and gender. Instruments Positive Affect and Negative Affect Schedule (PANAS; Watson et al., 1988). The PANAS instructs participants to rate to what extent they generally have experienced 20 different feelings or emotions (10 PA and 10 NA) during the last weeks, using a 5-point Likert scale (1 = very slightly, 5 = extremely). The 10\u2013item PA scale includes adjectives such as strong, proud, and interested. The 10\u2013 item NA scale includes adjectives such as afraid, ashamed and nervous. Cronbach\u2019s \u03b1 were .87 for PA and .89 for NA in the present study. The Short Depression-Happiness Scale (Joseph Linley, Harwood, Lewis & McCollam, 2004). This instrument consists of six items, three items measuring happiness (e.g., \u201cI felt happy\u201d) and three reverse coded items measuring depressive PeerJ reviewing PDF | (v2013:07:649:0:0:NEW 12 Jul 2013) R ev ie w in g M an us cr ip t states (e.g., \u201cI felt my life was meaningless\u201d). Participants rate how frequently they feel the way described in the item on a four-point scale: \u201cnever\u201d, \u201crarely\u201d, \u201csometimes\u201d, \u201coften\u201d. In the present study, Cronbach\u2019s \u03b1 was .85 for the happiness scale and .76 for the depression scale. Statistical treatment We used participants\u2019 self-reported affect measured by the PANAS from both Study 1 and 2 (N = 1,400) in order to classify participants in the four affective profiles. Participants\u2019 PA and NA scores were divided into high and low (cut-off points: low PA = 3.0 or less; high PA = 3.1 or above; low NA = 1.8 or less; and high NA = 1.9 or above). For study 1, the two independent variables of the study were gender and affective profile: self-fulfilling (n = 241; 153 males, 88 females), low affective (n = 236; 137 males, 99 females), high affective (n = 180; 115 males, 65 females), and self-destructive (n = 243; 145 males, 98 females). The dependent variables were PA, NA, happiness, and depression. Results and discussion A Multiple Analysis of Variance (MANOVA) indicated a significant effect for gender (F(4, 889) = 4.32; p = .002, Eta2 = 0.02, power = 0.93) as well as for affective profile (F(12, 2673) = 162.19; p < .001, Eta2 = 0.42, power = 1.00). The interaction of gender and affective profile was not significant (p = .236). A between-subjects ANOVA showed an significant gender effects for happiness (F(1, 892) = 7.60; p = 0.006), whereby the female participants expressed a higher level of happiness (M = 9.66, SD = 2.13) than the male participants (M =9.35, SD = 2.33). A between-subject ANOVA indicated significant affective profile effects for PA (F(3, 892) = 513.78; p < .001), NA (F(3, 892) = 503.58; p < .001), happiness (F(3, PeerJ reviewing PDF | (v2013:07:649:0:0:NEW 12 Jul 2013) R ev ie w in g M an us cr ip t 892) = 68.20; p < .001), and depression (F(3, 892) = 71.50; p < .001). A Bonferroni correction to the alpha level of .01 showed that the self-destructive group had significantly higher scores in NA and depression as well as lower scores in happiness in comparison to the other affective profiles. The self-fulfilling group differed significantly from the self-destructive profiles in all measured variables; PA, NA, happiness and depression. As expected, the high affective ones differed significantly from the self-fulfilling group in all variables except PA and the low affective ones differed significantly from the self-fulfilling group in all variables except NA. Which is not so strange since both the self-fulfilling group and the high affective group are characterized as high in PA and the same goes for self-fulfilling individuals and low affective individuals who are characterized by low NA. For further details, see table 1. Table 1 here Study II As in Study I, participants (N = 500, age mean = 34.08 sd. =12.55; 217 male and 283 female) were recruited from MTurk by the following criteria: resident of the USA and to both speak and write fluent in English. Participants were paid a wage of two American dollars for completing the task and informed that the study was confidential and voluntary. The participants were presented with a battery of self-reports comprising the affect, life satisfaction, and happiness-increasing strategies measures, as well as questions pertaining age and gender. Instruments The same instrument as in Study I was used in Study II to measure PA and NA (i.e., the PANAS). Cronbach\u2019s \u03b1 were .88 for PA and .90 for NA in Study II. PeerJ reviewing PDF | (v2013:07:649:0:0:NEW 12 Jul 2013) R ev ie w in g M an us cr ip t Satisfaction with Life Scale (Diener, Emmons, Larsen and Griffin, 1985). The instrument consists of 5 statements (e.g., \u201cIn most of my ways my life is close to my ideal\u201d) for which participants are asked to indicate degree of agreement in a 7-point Likert scale (1 = strongly disagree, 7 = strongly agree). The life satisfaction score was established by summarizing the 5 statements for each participant. Cronbach\u2019s \u03b1 were . 90 in the present study. Happiness-Increasing Strategies Scales (Tkach and Lyubomirsky, 2006). In the present study, participants were asked to rate (1 = never, 7 = all the time) how often they used the strategies identified by Tkach and Lyubomirsky (2006). The happiness-increasing strategies are organized in eight clusters: Social Affiliation (e.g., \u2018\u2018Support and encourage friends\u2019\u2019; Cronbach\u2019s \u03b1 = 0.79), Partying and Clubbing (e.g., \u2018\u2018Drink alcohol\u2019\u2019; Cronbach\u2019s \u03b1 = 0.74), Mental Control (e.g., \u2018\u2018Try not to think about being unhappy\u2019\u2019; Cronbach\u2019s \u03b1 = 0.43), Instrumental Goal Pursuit (e.g. \u2018\u2018Study\u2019\u2019; Cronbach\u2019s \u03b1 = 0.76), Passive Leisure (e.g. \u2018\u2018Surf the internet\u2019\u2019; Cronbach\u2019s \u03b1 = 0.63), Active Leisure (e.g. \u2018\u2018Exercise\u2019\u2019; Cronbach\u2019s \u03b1 = 0.65), Religion (e.g. \u2018\u2018Seek support from faith\u2019\u2019; Cronbach\u2019s \u03b1 = 0.70), and Direct Attempts (e.g. \u2018\u2018Act happy/smile, etc.\u2019\u2019; Cronbach\u2019s \u03b1 = 0.56). Statistical treatment As detailed in Study I, both samples were used in the classification of the four affective profiles. The number of participants in each profile for Study II were as follows: 158 self-fulfilling (75 males, 83 females), 92 low affective (42 males, 50 females), 123 high affective (54 males, 69 females), and 127 self-destructive (46 males, 81 females). The affective profiles and gender were the independent variables, PA, NA, life satisfaction, and the happiness-increasing strategies were the dependent variables. PeerJ reviewing PDF | (v2013:07:649:0:0:NEW 12 Jul 2013) R ev ie w in g M an us cr ip t ethics statement : This research protocol was approved by the Ethics Committee of the University of Gothenburg and written informed consent was obtained from all the study participants. PeerJ reviewing PDF | (v2013:07:649:0:0:NEW 12 Jul 2013) R ev ie w in g M an us cr ip t legends and captions : Table 1. Mean scores in PA, NA, happiness and depression for each affective profile in Study I. Table 2. Means in life satisfaction and happiness-increasing strategies among affective profiles in Study II. Table 3. Mean differences, in life satisfaction and happiness-increasing strategies between affective temperaments. Figure 1. Summary of the differences between affective profiles in happiness, depression, life satisfaction, and the happiness-increasing strategies. PeerJ reviewing PDF | (v2013:07:649:0:0:NEW 12 Jul 2013) R ev ie w in g M an us cr ip t method : affective profiles self-fulfilling : N = 158 High affective N = 123 Low affective N = 92 Self-destructive N = 127 self-fulfilling : Positive Affect 0.17* 1.32* 1.46* Negative Affect -1.12* 0.05ns -1.28* Life satisfaction 1.05* 0.75* 2.01* Social Affiliation 0.05ns 0.28* 0.54* Partying and Clubbing -0.16ns -0.06ns 0.12ns Mental Control -0.31* -0.09ns -0.47* Instrumental Goal Pursuit -0.04ns 0.39* 0.54* Religion 0.17ns 0.23ns 0.54* Passive Leisure -0.16ns 0.05ns 0.05ns Active Leisure 0.11ns 0.29* 0.49* Direct Attempts 0.24* 0.31* 0.64* High affective Positive Affect -0.17* 1.15* 1.29* Negative Affect 1.11* -1.17* -0.16ns Life satisfaction -1.05* -0.31ns 0.96* Social Affiliation -0.05ns 0.24ns 0.50* Partying and Clubbing 0.16ns 0.11ns 0.29* Mental Control 0.31* 0.23* -0.16ns Instrumental Goal Pursuit 0.04ns 0.43* 0.58* Religion -0.17ns 0.05ns 0.36ns Passive Leisure 0.16ns 0.21ns 0.20ns Active Leisure -0.11ns 0.18ns 0.38* Direct Attempts -0.23* 0.07ns 0.40ns Low affective Positive Affect -1.32* -1.15* 0.14ns Negative Affect -0.05 -1.17* -1.32* Life satisfaction -0.75* 0.31ns 1.26* PeerJ reviewing PDF | (v2013:07:649:0:0:NEW 12 Jul 2013) R ev ie w in g M an us cr ip t Social Affiliation -0.28* -0.24ns 0.26* Partying and Clubbing 0.06ns -0.11ns 0.18ns Mental Control 0.09ns -0.23* -0.40* Instrumental Goal Pursuit -0.39* -0.43* 0.15ns Religion -0.23ns -0.05ns 0.31ns Passive Leisure -0.05ns -0.21ns -0.00ns Active Leisure -0.29* -0.18ns 0.20ns Direct Attempts -0.31* -0.07ns 0.33* Self-destructive Positive Affect -1.46* -1.29* -0.14* Negative Affect 1.28* 0.16ns 1.33* Life satisfaction -2.01* -0.96* -1.26* Social Affiliation -0.54* -0.50* -0.26* Partying and Clubbing -0.12ns -0.29* -0.18ns Mental Control 0.47* 0.16ns 0.39* Instrumental Goal Pursuit -0.54* -0.58* -0.15ns Religion -0.54* -0.36ns -0.31ns Passive Leisure -0.05ns -0.20ns 0.00ns Active Leisure -0.49* -0.38* -0.20ns Direct Attempts -0.64* -0.40* -0.33* ns = non significant, * p < 0.01 with Bonferroni Correction. PeerJ reviewing PDF | (v2013:07:649:0:0:NEW 12 Jul 2013) R ev ie w in g M an us cr ip t Figure 1 Figure 1. Summary of the differences between affective profiles in happiness, depression, life satisfaction, and the happiness-increasing strategies. PeerJ reviewing PDF | (v2013:07:649:0:0:NEW 12 Jul 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:07:649:0:0:NEW 12 Jul 2013) R ev ie w in g M an us cr ip t",
+    "url": "https://peerj.com/articles/157/reviews/",
+    "review_1": "Ludmila Chistoserdova \u00b7 Aug 20, 2013 \u00b7 Academic Editor\nACCEPT\nI find modifications to the original manuscript satisfactory.",
+    "review_2": "Ludmila Chistoserdova \u00b7 Jun 1, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nI look forward to the revised manuscript.",
+    "review_3": "Anders Lanz\u00e9n \u00b7 Jun 1, 2013\nBasic reporting\nThe submitted manuscript appears to adhere to all PeerJ policies and is written in very in a very clear and concise English. The Introduction provides a clear overview and presents the scope of the manuscript in relation to previous work.\nExperimental design\nIn this manuscript, McCoy and Matsen provide a well-motivated measure of phylogenetic alpha-diversity (PD), able to take into interpolate between abundance-weighted (Barker 2002, as interpreted by Vellend) and classic -unweigheted PD . Further, they convincingly demonstrate that this measure is able to distinguish between different communities using a selection of three microbiome datasets from healthy and diseased individuals. The work is meaningful, rigorous and the methods described clearly.\n\nHowever, filtering, clustering and other methods to remove PCR- and sequencing-induced artefacts were not addressed properly by the original authors in the two larger sequence datasets (vaginal and oral microbiomes). The first of these used filtering based only on quality scores and the second the RDP pipeline. It has been demonstrated that this method is not sufficient to remove such artefacts (see e.g. Quince et al 2011 (BMC Bioinformatics 12:38.), Schloss et al 2011 (PLoS ONE 6:e27310), Kunin et al 2009 (Environ Microbiol 12:118-123) and will result in inflated OTU richness estimates and skewed OTU-abundance relationships. PD should be less sensitive to such artefacts and this can very well be the main reason for why the estimate suggested here performed better than OTU-based ones. I think that this fact deserves to be mentioned in the Discussion. Better still, the authors could include a dataset in the comparison which has been handled using e.g. AmpliconNoise or DeNoiser and where chimeric sequences were properly addressed, or re-analyse the raw sequence data using such methods, resulting in a new OTU table. Failing this, I think that this common pitfuall of OTU-based analysis deserves to be mentioned in the Discussion as the richness estimates in the original articles do appear very high.\n\nIf the authors are interested in repeating a test with noise-cleaned sequence data and a more robust OTU table, I happily volunteer to help out with data-treatment (sequence cleaning)\nValidity of the findings\nSee Experimental design above. Apart from that no comments.\nAdditional comments\nSpecific comments:\n\nIn the abstract, Simpson diversity is called a \"count-only\" measure. This is unclear and leads this reader to think more of OTU richness. In Table 1, Af and Pf are mentioned in the caption but in the headers \"Ac\" and \"Pc\" are used.\nCite this review as\nLanz\u00e9n A (2013) Peer Review #1 of \"Abundance-weighted phylogenetic diversity measures distinguish microbial community states and are robust to sampling depth (v0.1)\". PeerJ https://doi.org/10.7287/peerj.157v0.1/reviews/1",
+    "review_4": "Catherine Lozupone \u00b7 May 16, 2013\nBasic reporting\nmeets standards\nExperimental design\ngood\nValidity of the findings\ngood\nAdditional comments\nIn this paper, McCoy and Matsen introduce a novel family of phylogenetic alpha diversity measures that interpolates between classical phylogenetic diversity (PD), which does not account for the abundance of phylogenetic lineages in a sample, and an abundance-weighted extension of PD. Using 3 published studies of the human gut microbiota, they evaluate how their new measure and other alpha diversity measures compare in their ability to differentiate samples in different categories, including healthy versus bacterial vaginosis, periodontitis and controls, and skin samples at different developmental stages. Overall, this paper highlights that phylogenetic alpha diversity measures can be more powerful than commonly used \u201cdiscrete\u201d measures that rely on OTU assignments. It also highlights that abundance-weighted phylogenetic alpha diversity measures can be more powerful than phylogenetic alpha diversity measures that do not account for abundances (i.e. PD). The authors point out that abundance-weighted phylogenetic diversity measures are not used commonly in studies of the microbiota and not implemented in commonly used analysis tools such as QIIME and mother, and argue based on their results that they should be. Overall, I agree with them, and thus think that this paper is a valuable contribution to the field. I did, however, think that there were some ways in which the paper could be improved.\n\n1) I was confused by the treatment of rarefaction in this paper. My understanding is that for e.g. the data presented in Table 1, for the OTU and family based \u201cdiscrete\u201d measures (i.e. Shannon, Simpson, Chao1, and ACE), they rarefied the data, and then for the phylogenetic measures they did not rarefy, with the exception of PD, for which they present both the rarefied and unrarefied results. Why do it this way (i.e. rarefy one class of measures and not the other and then for one (PD) do it both ways?).\n\nIn general, I think that you should always rarefy. The variability in the number of sequences per sample has no real meaning (i.e. just an artifact of sequencing since equal amounts of DNA from each sample is added to the sequencer), and this variability has the potential to affect alpha diversity estimates. I realize that later in the paper (in Figure 3 and mathematically) they show that abundance weighted measures are not very sensitive to sampling depth, although I am not convinced that there will not be more sensitivity for environments that are very undersampled.\n\nSo, by making the point that abundance-weighted measures are not sensitive to sampling depth and then not rarefying the data in their analyses, are they trying to say that because of this lack of sensitivity that we should not rarefy when using these measures? What would be the advantage? The only thing that I can think is that you can potentially more accurately estimate alpha diversity with more data, but as this has not been demonstrated here, I still think that it is good practice to rarefy.\n\nWith regard to this, in Figure 3 and Fig. S4, they show that 0.25D(T) and BWPD0.25 are sensitive to sampling depth, and yet as far as I can tell they still do not used rarefied data in their analyses of all three of the microbiota datasets with these measures. On page 10, lines 276-278, the authors note \u201cclassical phylogenetic diversity was among the worst predictors; rarefaction did help\u2026\u201d. Why look at unrarefied at all when they show in Figure 3 themselves that sequencing depth matters and conceptually it of course makes sense that if you have not sequenced fully, the more sequences you look at with PD the more diversity you will see.\n\nFor all three studies, I would really like to see all of the measurements made on rarefied data.\n\n2) The authors evaluate many different alpha diversity measures in this paper. One thing that would be really helpful is a table that described and classifies them all. Perhaps a columns that 1) designated discrete versus phylogenetic 2) designate abundance-weighted/non/in-between, 3) show an equation where appropriate, 4) briefly describes the measure with words, and 5) shows the info in paragraph 1 of the introduction of which measures are phylogenetic \u201cversions\u201d of particular discrete measures and 6) gives a reference.\n\n3) The information given on page 3 on the example datasets should have better consistency on the types of information provided for each one. The description of the skin microbiome is particularly sparse. Information provided for the other samples, such as the range of sequences per sample, how these sequences were generated, quality filtered etc. should be provided.\n\n4) It is kinda interesting that the discrete measures applied at the family level often appear more powerful than those at the OTU level. Any ideas on why this may be the case?\n\nMinor comments:\n1) In the sentence in the abstract \u201cIn all three of the datasets considered, an abundance-weighted measure is the best differentiator between community states.\u201d, the authors should say \u201ca phylogenetic abundance-weighted measure is the best differentiator\u201d as stated could be a discrete abundance-weighted measure which didn't do so good.\n2) Lines 141, 159: I am not sure what the authors mean exactly when they say that they \u201cassigned the root taxonomically\u201d\n3) The meaning of \u201cshallow\u201d and \u201cdeep\u201d in Fig 4 are not defined anywhere.\n4) typos/grammatical\na. line 79: Fix Oh et al reference formatting\nb. line 304: should Tab. 5 be Table 2?\nCite this review as\nLozupone C (2013) Peer Review #2 of \"Abundance-weighted phylogenetic diversity measures distinguish microbial community states and are robust to sampling depth (v0.1)\". PeerJ https://doi.org/10.7287/peerj.157v0.1/reviews/2",
+    "pdf_1": "https://peerj.com/articles/157v0.2/submission",
+    "pdf_2": "https://peerj.com/articles/157v0.1/submission",
+    "all_reviews": "Review 1: Ludmila Chistoserdova \u00b7 Aug 20, 2013 \u00b7 Academic Editor\nACCEPT\nI find modifications to the original manuscript satisfactory.\nReview 2: Ludmila Chistoserdova \u00b7 Jun 1, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nI look forward to the revised manuscript.\nReview 3: Anders Lanz\u00e9n \u00b7 Jun 1, 2013\nBasic reporting\nThe submitted manuscript appears to adhere to all PeerJ policies and is written in very in a very clear and concise English. The Introduction provides a clear overview and presents the scope of the manuscript in relation to previous work.\nExperimental design\nIn this manuscript, McCoy and Matsen provide a well-motivated measure of phylogenetic alpha-diversity (PD), able to take into interpolate between abundance-weighted (Barker 2002, as interpreted by Vellend) and classic -unweigheted PD . Further, they convincingly demonstrate that this measure is able to distinguish between different communities using a selection of three microbiome datasets from healthy and diseased individuals. The work is meaningful, rigorous and the methods described clearly.\n\nHowever, filtering, clustering and other methods to remove PCR- and sequencing-induced artefacts were not addressed properly by the original authors in the two larger sequence datasets (vaginal and oral microbiomes). The first of these used filtering based only on quality scores and the second the RDP pipeline. It has been demonstrated that this method is not sufficient to remove such artefacts (see e.g. Quince et al 2011 (BMC Bioinformatics 12:38.), Schloss et al 2011 (PLoS ONE 6:e27310), Kunin et al 2009 (Environ Microbiol 12:118-123) and will result in inflated OTU richness estimates and skewed OTU-abundance relationships. PD should be less sensitive to such artefacts and this can very well be the main reason for why the estimate suggested here performed better than OTU-based ones. I think that this fact deserves to be mentioned in the Discussion. Better still, the authors could include a dataset in the comparison which has been handled using e.g. AmpliconNoise or DeNoiser and where chimeric sequences were properly addressed, or re-analyse the raw sequence data using such methods, resulting in a new OTU table. Failing this, I think that this common pitfuall of OTU-based analysis deserves to be mentioned in the Discussion as the richness estimates in the original articles do appear very high.\n\nIf the authors are interested in repeating a test with noise-cleaned sequence data and a more robust OTU table, I happily volunteer to help out with data-treatment (sequence cleaning)\nValidity of the findings\nSee Experimental design above. Apart from that no comments.\nAdditional comments\nSpecific comments:\n\nIn the abstract, Simpson diversity is called a \"count-only\" measure. This is unclear and leads this reader to think more of OTU richness. In Table 1, Af and Pf are mentioned in the caption but in the headers \"Ac\" and \"Pc\" are used.\nCite this review as\nLanz\u00e9n A (2013) Peer Review #1 of \"Abundance-weighted phylogenetic diversity measures distinguish microbial community states and are robust to sampling depth (v0.1)\". PeerJ https://doi.org/10.7287/peerj.157v0.1/reviews/1\nReview 4: Catherine Lozupone \u00b7 May 16, 2013\nBasic reporting\nmeets standards\nExperimental design\ngood\nValidity of the findings\ngood\nAdditional comments\nIn this paper, McCoy and Matsen introduce a novel family of phylogenetic alpha diversity measures that interpolates between classical phylogenetic diversity (PD), which does not account for the abundance of phylogenetic lineages in a sample, and an abundance-weighted extension of PD. Using 3 published studies of the human gut microbiota, they evaluate how their new measure and other alpha diversity measures compare in their ability to differentiate samples in different categories, including healthy versus bacterial vaginosis, periodontitis and controls, and skin samples at different developmental stages. Overall, this paper highlights that phylogenetic alpha diversity measures can be more powerful than commonly used \u201cdiscrete\u201d measures that rely on OTU assignments. It also highlights that abundance-weighted phylogenetic alpha diversity measures can be more powerful than phylogenetic alpha diversity measures that do not account for abundances (i.e. PD). The authors point out that abundance-weighted phylogenetic diversity measures are not used commonly in studies of the microbiota and not implemented in commonly used analysis tools such as QIIME and mother, and argue based on their results that they should be. Overall, I agree with them, and thus think that this paper is a valuable contribution to the field. I did, however, think that there were some ways in which the paper could be improved.\n\n1) I was confused by the treatment of rarefaction in this paper. My understanding is that for e.g. the data presented in Table 1, for the OTU and family based \u201cdiscrete\u201d measures (i.e. Shannon, Simpson, Chao1, and ACE), they rarefied the data, and then for the phylogenetic measures they did not rarefy, with the exception of PD, for which they present both the rarefied and unrarefied results. Why do it this way (i.e. rarefy one class of measures and not the other and then for one (PD) do it both ways?).\n\nIn general, I think that you should always rarefy. The variability in the number of sequences per sample has no real meaning (i.e. just an artifact of sequencing since equal amounts of DNA from each sample is added to the sequencer), and this variability has the potential to affect alpha diversity estimates. I realize that later in the paper (in Figure 3 and mathematically) they show that abundance weighted measures are not very sensitive to sampling depth, although I am not convinced that there will not be more sensitivity for environments that are very undersampled.\n\nSo, by making the point that abundance-weighted measures are not sensitive to sampling depth and then not rarefying the data in their analyses, are they trying to say that because of this lack of sensitivity that we should not rarefy when using these measures? What would be the advantage? The only thing that I can think is that you can potentially more accurately estimate alpha diversity with more data, but as this has not been demonstrated here, I still think that it is good practice to rarefy.\n\nWith regard to this, in Figure 3 and Fig. S4, they show that 0.25D(T) and BWPD0.25 are sensitive to sampling depth, and yet as far as I can tell they still do not used rarefied data in their analyses of all three of the microbiota datasets with these measures. On page 10, lines 276-278, the authors note \u201cclassical phylogenetic diversity was among the worst predictors; rarefaction did help\u2026\u201d. Why look at unrarefied at all when they show in Figure 3 themselves that sequencing depth matters and conceptually it of course makes sense that if you have not sequenced fully, the more sequences you look at with PD the more diversity you will see.\n\nFor all three studies, I would really like to see all of the measurements made on rarefied data.\n\n2) The authors evaluate many different alpha diversity measures in this paper. One thing that would be really helpful is a table that described and classifies them all. Perhaps a columns that 1) designated discrete versus phylogenetic 2) designate abundance-weighted/non/in-between, 3) show an equation where appropriate, 4) briefly describes the measure with words, and 5) shows the info in paragraph 1 of the introduction of which measures are phylogenetic \u201cversions\u201d of particular discrete measures and 6) gives a reference.\n\n3) The information given on page 3 on the example datasets should have better consistency on the types of information provided for each one. The description of the skin microbiome is particularly sparse. Information provided for the other samples, such as the range of sequences per sample, how these sequences were generated, quality filtered etc. should be provided.\n\n4) It is kinda interesting that the discrete measures applied at the family level often appear more powerful than those at the OTU level. Any ideas on why this may be the case?\n\nMinor comments:\n1) In the sentence in the abstract \u201cIn all three of the datasets considered, an abundance-weighted measure is the best differentiator between community states.\u201d, the authors should say \u201ca phylogenetic abundance-weighted measure is the best differentiator\u201d as stated could be a discrete abundance-weighted measure which didn't do so good.\n2) Lines 141, 159: I am not sure what the authors mean exactly when they say that they \u201cassigned the root taxonomically\u201d\n3) The meaning of \u201cshallow\u201d and \u201cdeep\u201d in Fig 4 are not defined anywhere.\n4) typos/grammatical\na. line 79: Fix Oh et al reference formatting\nb. line 304: should Tab. 5 be Table 2?\nCite this review as\nLozupone C (2013) Peer Review #2 of \"Abundance-weighted phylogenetic diversity measures distinguish microbial community states and are robust to sampling depth (v0.1)\". PeerJ https://doi.org/10.7287/peerj.157v0.1/reviews/2\nReview 5: nan\nReview 6: \n",
+    "LLM_extracted_review": "['Review 1: I find modifications to the original manuscript satisfactory.\\n\\nReview 2: I look forward to the revised manuscript.\\n\\nReview 3: The submitted manuscript appears to adhere to all PeerJ policies and is written in very clear and concise English. The Introduction provides a clear overview and presents the scope of the manuscript in relation to previous work. However, filtering, clustering and other methods to remove PCR- and sequencing-induced artefacts were not addressed properly by the original authors in the two larger sequence datasets (vaginal and oral microbiomes). The first of these used filtering based only on quality scores and the second the RDP pipeline. It has been demonstrated that this method is not sufficient to remove such artefacts and will result in inflated OTU richness estimates and skewed OTU-abundance relationships. PD should be less sensitive to such artefacts and this can very well be the main reason for why the estimate suggested here performed better than OTU-based ones. I think that this fact deserves to be mentioned in the Discussion. Better still, the authors could include a dataset in the comparison which has been handled using e.g. AmpliconNoise or DeNoiser and where chimeric sequences were properly addressed, or re-analyse the raw sequence data using such methods, resulting in a new OTU table. Failing this, I think that this common pitfall of OTU-based analysis deserves to be mentioned in the Discussion as the richness estimates in the original articles do appear very high. If the authors are interested in repeating a test with noise-cleaned sequence data and a more robust OTU table, I happily volunteer to help out with data-treatment (sequence cleaning). Apart from that no comments. Specific comments: In the abstract, Simpson diversity is called a \"count-only\" measure. This is unclear and leads this reader to think more of OTU richness. In Table 1, Af and Pf are mentioned in the caption but in the headers \"Ac\" and \"Pc\" are used.\\n\\nReview 4: Overall, this paper highlights that phylogenetic alpha diversity measures can be more powerful than commonly used \u201cdiscrete\u201d measures that rely on OTU assignments. It also highlights that abundance-weighted phylogenetic alpha diversity measures can be more powerful than phylogenetic alpha diversity measures that do not account for abundances (i.e. PD). The authors point out that abundance-weighted phylogenetic diversity measures are not used commonly in studies of the microbiota and not implemented in commonly used analysis tools such as QIIME and mother, and argue based on their results that they should be. I did, however, think that there were some ways in which the paper could be improved. I was confused by the treatment of rarefaction in this paper. My understanding is that for the data presented in Table 1, for the OTU and family based \u201cdiscrete\u201d measures, they rarefied the data, and then for the phylogenetic measures they did not rarefy, with the exception of PD, for which they present both the rarefied and unrarefied results. Why do it this way? In general, I think that you should always rarefy. The variability in the number of sequences per sample has no real meaning, and this variability has the potential to affect alpha diversity estimates. I realize that later in the paper they show that abundance weighted measures are not very sensitive to sampling depth, although I am not convinced that there will not be more sensitivity for environments that are very undersampled. So, by making the point that abundance-weighted measures are not sensitive to sampling depth and then not rarefying the data in their analyses, are they trying to say that because of this lack of sensitivity that we should not rarefy when using these measures? What would be the advantage? The only thing that I can think is that you can potentially more accurately estimate alpha diversity with more data, but as this has not been demonstrated here, I still think that it is good practice to rarefy. With regard to this, they show that 0.25D(T) and BWPD0.25 are sensitive to sampling depth, and yet as far as I can tell they still do not use rarefied data in their analyses of all three of the microbiota datasets with these measures. On page 10, lines 276-278, the authors note \u201cclassical phylogenetic diversity was among the worst predictors; rarefaction did help\u2026\u201d. Why look at unrarefied at all when they show in Figure 3 themselves that sequencing depth matters? For all three studies, I would really like to see all of the measurements made on rarefied data. The authors evaluate many different alpha diversity measures in this paper. One thing that would be really helpful is a table that described and classified them all. The information given on page 3 on the example datasets should have better consistency on the types of information provided for each one. The description of the skin microbiome is particularly sparse. Minor comments: In the sentence in the abstract \u201cIn all three of the datasets considered, an abundance-weighted measure is the best differentiator between community states.\u201d, the authors should say \u201ca phylogenetic abundance-weighted measure is the best differentiator\u201d. Lines 141, 159: I am not sure what the authors mean exactly when they say that they \u201cassigned the root taxonomically\u201d. The meaning of \u201cshallow\u201d and \u201cdeep\u201d in Fig 4 are not defined anywhere. Typos/grammatical issues noted.\\n\\nReview 5: nan\\n\\nReview 6: ']"
+}

peerj_json_files/PeerJ_Json_114.json ADDED Viewed

	@@ -0,0 +1,17 @@

+{
+    "v1_Abstract": "In microbial ecology studies, the most commonly used ways of investigating alpha (withinsample) diversity are either to apply non-phylogenetic measures such as Simpson's index to Operational Taxonomic Unit (OTU) groupings, or to use classical phylogenetic diversity (PD), which is not abundance-weighted. Although alpha diversity measures that use abundance information in a phylogenetic framework do exist, but are not widely used within the microbial ecology community. The performance of abundance-weighted phylogenetic diversity measures compared to classical discrete measures has not been explored, and the behavior of these measures under rarefaction (sub-sampling) is not yet clear. In this paper we compare the ability of various alpha diversity measures to distinguish between different community states in the human microbiome for three different data sets. We also present and compare a novel one-parameter family of alpha diversity measures, \\ (\\operatorname{BWPD}_\\theta\\), that interpolates between classical phylogenetic diversity (PD) and an abundance-weighted extension of PD. Additionally, we examine the sensitivity of these phylogenetic diversity measures to sampling, via computational experiments and by deriving a closed form solution for the expectation of phylogenetic quadratic entropy under re-sampling. In all three of the datasets considered, an abundance-weighted measure is the best differentiator between community states. OTU-based measures, on the other hand, are less effective in distinguishing community types. In addition, abundance-weighted phylogenetic diversity measures are less sensitive to differing sampling intensity than their unweighted counterparts. Based on these results we encourage the use of abundanceweighted phylogenetic diversity measures, especially for cases such as microbial ecology where species delimitation is difficult.",
+    "v2_Abstract": "In microbial ecology studies, the most commonly used ways of investigating alpha (within-sample) diversity are either to apply count-only measures such as Simpson's index to Operational Taxonomic Unit (OTU) groupings, or to use classical phylogenetic diversity (PD), which is not abundance-weighted. Although alpha diversity measures that use abundance information in a phylogenetic framework do exist, but are not widely used within the microbial ecology community. The performance of abundance-weighted phylogenetic diversity measures compared to classical discrete measures has not been explored, and the behavior of these measures under rarefaction (sub-sampling) is not yet clear. In this paper we compare the ability of various alpha diversity measures to distinguish between different community states in the human microbiome for three different data sets. We also present and compare a novel one-parameter family of alpha diversity measures, BWPD \u03b8 \\operatorname{BWPD}_\\theta , that interpolates between classical phylogenetic diversity (PD) and an abundance-weighted extension of PD. Additionally, we examine the sensitivity of these phylogenetic diversity measures to sampling, via computational experiments and by deriving a closed form solution for the expectation of phylogenetic quadratic entropy under re-sampling. In all three of the datasets considered, an abundance-weighted measure is the best differentiator between community states. OTU-based measures, on the other hand, are less effective in distinguishing community types. In addition, abundance-weighted phylogenetic diversity measures are less sensitive to differing sampling intensity than their unweighted counterparts. Based on these results we encourage the use of abundance-weighted phylogenetic diversity measures, especially for cases such as microbial ecology where species delimitation is difficult.",
+    "v1_text": "in microbial ecology studies, the most commonly used ways of investigating alpha (within- : sample) diversity are either to apply non-phylogenetic measures such as Simpson's index to Operational Taxonomic Unit (OTU) groupings, or to use classical phylogenetic diversity (PD), which is not abundance-weighted. Although alpha diversity measures that use abundance information in a phylogenetic framework do exist, but are not widely used within the microbial ecology community. The performance of abundance-weighted phylogenetic diversity measures compared to classical discrete measures has not been explored, and the behavior of these measures under rarefaction (sub-sampling) is not yet clear. In this paper we compare the ability of various alpha diversity measures to distinguish between different community states in the human microbiome for three different data sets. We also present and compare a novel one-parameter family of alpha diversity measures, \\ (\\operatorname{BWPD}_\\theta\\), that interpolates between classical phylogenetic diversity (PD) and an abundance-weighted extension of PD. Additionally, we examine the sensitivity of these phylogenetic diversity measures to sampling, via computational experiments and by deriving a closed form solution for the expectation of phylogenetic quadratic entropy under re-sampling. In all three of the datasets considered, an abundance-weighted measure is the best differentiator between community states. OTU-based measures, on the other hand, are less effective in distinguishing community types. In addition, abundance-weighted phylogenetic diversity measures are less sensitive to differing sampling intensity than their unweighted counterparts. Based on these results we encourage the use of abundance- weighted phylogenetic diversity measures, especially for cases such as microbial ecology where species delimitation is difficult. PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t Connor O. McCoy and Frederick A. Matsen IV\u2217 Fred Hutchinson Cancer Research Center 1100 Fairvew Ave. N Seattle, WA 98109 \u2217Corresponding author: matsen@fhcrc.org 1 PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t 2 1. INTRODUCTION It is now well accepted that incorporating phylogenetic information into alpha (single-sample) and beta (between-sample) diversity measures can be useful in a variety of ecological contexts. Phylogenetic equivalents of all of major alpha diversity measures have been developed (Table 1). Starting with Faith\u2019s original definition of phylogenetic diversity (Faith, 1992), which generalizes species count, there are now phylogenetic generalizations of the Simpson index to Rao\u2019s quadratic entropy (Rao, 1982; Warwick and Clarke, 1995), the Shannon index to phylogenetic entropy (Allen et al., 2009), and the Hill numbers to qD(T) (Chao et al., 2010). Phylogenetic diversity itself has been extended to incorporate taxon counts (Barker, 2002) and proportional abundance (Vellend et al., 2011). There have also been abundance-weighted measures that explicitly measure phylogenetic community structure (Fine and Kembel, 2011), or an \u201ceffective number of species\u201d (Chao et al., 2010). Many diversity measures can be tidily expressed in the framework of Leinster and Cobbold (2012), although the expression of phylogenetic diversity measures for non-ultrametric trees is complex. In this paper we use three example human microbiome datasets to demonstrate the utility of abundance-weighted phylogenetic diversity measures. We also introduce a one-parameter family interpolating between classical PD and an abundance-weighted generalization. We call the parameter \u03b8 and denote the one-parameter family BWPD\u03b8; BWPD0 is classical PD, whereas BWPD1 is balanceweighted phylogenetic diversity, effectively PDaw of Vellend et al. (2011). Intermediate values of \u03b8 allow a partially-abundance-weighted compromise. Such a compromise has recently been shown to be useful for measuring beta diversity, with the introduction of a oneparameter family of \u201cgeneralized UniFrac\u201d measures (Chen et al., 2012). We use the name Balance Weighted Phylogenetic Diversity as described below because there are a variety of abundance weighted phylogenetic diversity measures. We compare the behavior of PD measures, including BWPD\u03b8, under various levels of sampling using theory and example data sets. 2. MATERIALS AND METHODS 2.1. Datasets. We apply the methods described below to three previously described 16S rRNA surveys of the human microbiome. The PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t 3 first two datasets are composed of samples from \u201cnormal\u201d and dysbiotic microbial communities, where previous studies have associated changes in diversity with changes in health. The third dataset investigates the changes of the skin microbiome through time. 2.1.1. Bacterial vaginosis. First, we reanalyze a pyrosequencing dataset describing bacterial communities from women being monitored in a sexually transmitted disease clinic for bacterial vaginosis (BV). BV has previously been shown to be associated with increased community diversity (Fredricks et al., 2005). For this study, swabs were taken from 242 women from the Public Health, Seattle and King County Sexually Transmitted Diseases Clinic between September 2006 and June 2010 of which 220 samples resulted in enough material to analyze (Srinivasan et al., 2012). Selection of reference sequences and sequence preprocessing were performed using the methods described in (Srinivasan et al., 2012). 452,358 reads passed quality filtering, with a median of 1,779 reads per sample (range: 523\u20132,366). 2.1.2. Oral periodontitis. We also utilize sequence data from a study of subgingival communities in 29 subjects with periodontitis, along with an equal number of healthy controls (Griffen et al., 2011a). The publication analyzing this dataset showed increased community diversity in samples from dysbiotic patients compared to healthy controls. Raw sequences were filtered, retaining only those reads with: a mean quality score of at least 25, no ambiguous bases, at least 150 base pairs in length, and an exact match to the sequencing primer and barcode. A total of 759,423 reads passed quality filtering, with a median of 8,320 reads per sample (range: 4,096\u201314,319). As the phylogenetic placement method used below to calculate our measures requires a reference tree and alignment, we created a tree with FastTree 2.1.4 (Price et al., 2010) using the alignment and accompanying taxonomic annotation from the curated CORE database of oral microbiota (Griffen et al., 2011b). 2.1.3. Skin microbiome through time. Our third data set is a study of skin microbial diversity through adolescence (Oh et al., 2012). Aligned sequences were obtained courtesy of the authors, although sequence data is available under the accession numbers [GQ000001] to [GQ116391] and can be accessed through BioProject ID 46333. A total of 90,142 Sanger sequences were available, with a median of 693 sequences per sample (range: 317\u20132884). PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t 4 2.2. Balance-weighted phylogenetic diversity. In this section we introduce BWPD\u03b8, our one-parameter family interpolating between classical PD and fully balance-weighted phylogenetic diversity. We will primarily consider so-called unrooted (Pardi and Goldman, 2007) phylogenetic diversity, which does not necessarily include the root. The case of rooted phylogenetic diversity can be calculated in a similar though simpler way as described below. Although we will primarily be working in an unrooted sense, it will be useful to use terminology that corresponds to the rooted case. For this reason, if the tree is not already rooted, assume an arbitrary root has been chosen; let the proximal side of a given edge be the side that contains the root and distal be the other. We will describe BWPD\u03b8 in terms of a phylogenetic tree T with leaves L, and a contingency table describing the number of observations of the organisms at the leaves in various samples. The contingency table has rows labeled with the leaves of T , and columns labeled by samples. In microbial ecology this is frequently known as an OTU table. The entry corresponding to a given leaf and a given sample is the number of times that leaf was observed in that sample. The classical (unrooted) phylogenetic diversity of a given sample in this context is the total branch length of the tree subtended by the leaves in that sample. The path to generalizing PD is to note that this can be expressed as a sum of branch lengths multiplied by a step function. Let f(x) be the function that is one for x > 0 and zero otherwise. Let g(x) = min(f(x), f(1\u2212x)) and Ds(i) be the fraction of reads in sample s that are in leaves on the distal side of edge i. Phylogenetic diversity can be then expressed as (1) PDu(s) = \ufffd i \ufffdi g(Ds(i)) That is, the sum of edge lengths in T which have reads from s on both the distal and proximal side. Note that the step function g is the limit of a one-parameter family of functions (Fig. 1). Indeed, defining (2) g\u03b8(x) = [2min(x, 1\u2212 x)]\u03b8 , g is the pointwise limit of the g\u03b8 on the closed unit interval as \u03b8 goes to zero. Thus our one-parameter generalization is (3) BWPD\u03b8(s) = \ufffd i \ufffdi g\u03b8(Ds(i)). PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t 5 Note that when \u03b8 = 0 this is PD and when \u03b8 = 1 this is an abundanceweighted version of PD equivalent to executing the \u0394nPD recipe of Barker (2002) up to a multiplicative factor. The rooted equivalent of (3) is (4) RBWPD\u03b8(s) = \ufffd i \ufffdi (Ds(i)) \u03b8, which interpolates between rooted PD and an abundance-weighted version. Vellend et al. (2011) describe a measure, PDaw, which is equal to RBWPD1 multiplied by the total number of branches in T . We call BWPD1 balance-weighted phylogenetic diversity because it weights edges according to the balance of read fractions on either side of an edge\u2013 edges with even amount of mass on either side are up-weighted, while edges with an uneven balance of mass are down-weighted. Indeed, if |x \u2212 (1 \u2212 x)| is thought of as the imbalance of read fraction on either side of an edge, then 1\u2212 |x\u2212 (1\u2212x)| is a measure of balance; note that on the unit interval, 2min(x, 1\u2212 x) = 1\u2212 |x\u2212 (1\u2212 x)|. Because a small x or an x close to 1 gives a small coefficient in the summation, small collections of reads or small perturbations of the read distribution will not change the value of BWPD1 appreciably. 2.3. Calculation of PD measures in example applications. Reads from the vaginal and oral studies were placed on a tree created from a curated set of taxonomically annotated reference sequences. As phylogenetic entropy and qD(T) operate on a rooted phylogeny, reference trees were assigned a root taxonomically (Matsen and Gallagher, 2012) meaning that a root was found that best separated highlevel taxonomic groupings. pplacer was run in posterior probability mode (using the -p and --informative-prior flags), which defines an informative prior for pendant branch lengths with a mean derived from the average distances from the edge in question to the leaves of the tree. The resulting set of placements were classified at the family rank using a hybrid classifier implemented in the guppy tool from the pplacer suite. The hybrid classifier assigns taxonomic annotations to sequences using the combination of a na\u0131\u0308ve Bayes classifier (Wang et al., 2007) with a phylogenetic classifier (Matsen et al., unpublished results). Any reads that could not be confidently classified to the family rank were not used in measures based on classification. Full-length 16S sequences were available for the skin data, and so a more traditional tree-building approach was used. Representative PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t 6 OTUs were chosen for each site by clustering at 97% identity using USEARCH 5.1 (Edgar, 2010), with trees built on OTU centroids using FastTree (Price et al., 2010). To conform with methods used in Oh et al. (2012), the na\u0131\u0308ve Bayes classifier (Wang et al., 2007) was used to infer genus-level classifications to taxonomically root the tree; in our case we used the RDP classifier v2.5. The contingency (OTU) tables generated by clustering were made available to our tools via the BIOM (McDonald et al., 2012) format. PDu (unrooted PD), phylogenetic quadratic entropy (Rao, 1982), phylogenetic entropy (Allen et al., 2009), and qD(T) (Chao et al., 2010) were implemented for phylogenetic placements in the freelyavailable pplacer suite of tools (Matsen et al., 2010) (http://matsen. fhcrc.org/pplacer) in the subcommand guppy fpd. Prior to diversity estimation, either phylogenetic placements were rarefied to the read count of the specimen in the dataset with the fewest sequences using guppy rarefy, or the corresponding rarefaction on fulllength sequences was performed with the QIIME (Caporaso et al., 2010) rarefaction tool single rarefaction.py. The mean value of each statistic over 100 such rarefactions was used for analysis. Discrete measures of alpha diversity and richness were calculated on contingency tables obtained from clustering and taxonomic classification. Sequences were clustered into Operational Taxonomic Units (OTUs) at a 97% identity threshold using USEARCH 5.1 (Edgar, 2010). Similar results were observed when clustering at 95% identity (results not shown). OTU counts and family-level taxon counts were then rarefied as above in R 3.0.1 (R Development Core Team, 2012) using the vegan package (Oksanen et al., 2012). We obtained values for the Simpson (1949) and Shannon (1948) diversity indices, as well as the Chao1 (Chao, 1984) and ACE (Chao and Lee, 1992) measures of species richness using vegan functions diversity and estimateR. 2.4. Comparative analysis of alpha diversity measures. To investigate the relation between various measures of alpha diversity, we calculated Pearson\u2019s r between all pairs of measures using the function rcorr from the R package Hmisc (Harrell Jr., 2012). We then performed hierarchical clustering with the R function hclust, using d = 1\u2212 r as the distance between two measures. Association of each measure with clinical criteria for the first two data sets was evaluated by examining the accuracy of a logistic regression using the measure as the sole predictor of whether the sample came from a \u201cnormal\u201d or dysbiotic subject. In the vaginal dataset, we assessed each measure\u2019s ability to predict whether a sample was PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t 7 from a subject positive for BV by Amsel\u2019s criteria, a clinical diagnostic method (Amsel et al., 1983). In the oral dataset, we assessed each measure\u2019s ability to predict whether a sample was from a healthy control, or a subject with periodontitis. Accuracy in predicting sample community state was assessed by leave-one-out cross-validation using the R package boot (Davison and Hinkley, 1997; Canty and Ripley, 2012). For the vaginal dataset, we also calculated R2 values using each measure individually as a predictor for sample Nugent score in a linear regression. The Nugent score provides a diagnostic score for BV which ranges from 0 (BV-negative) to 10 (BV-positive) based on presence and absence of bacterial morphotypes as viewed under a microscope (Nugent et al., 1991). We calculated p-values to compare within- and between-stratification variability using R\u2019s built-in t.test function for the vaginal data, which had a binary stratification, and the aov function for the oral and skin data sets. The vaginal dataset data was stratified by Amsel\u2019s criterion, the oral dataset by condition and sampling site, and the skin microbiome dataset by Tanner scale of physical development (Oh et al., 2012). Note that we are not presenting these uncorrected pvalues as evidence that there is an interesting relationship between the microbiome and a given stratification, but rather are using pvalues as a way of measuring within-stratum heterogeneity compared to between-stratum heterogeneity for the various measures. 3. RESULTS 3.1. Application to the human microbiome. 3.1.1. Vaginal microbiome. Like Srinivasan et al. (2012) and many others in the field, we observe greater diversity in BV positive specimens using a variety of diversity and richness measures (Fig. S1). In particular, this is true for BWPD\u03b8 for a variety of values of \u03b8 (Fig. S2). In the vaginal data, phylogenetic measures of alpha diversity have better cross-validation accuracy for the Amsel classification and better correlation with the Nugent score than discrete OTU-based measures (Table 2). All measures were somewhat accurate in identifying community state, with even the worst performers classifying almost 70% of samples correctly. BWPD0.25, BWPD0.5, PDu, and phylogenetic entropy perform well predicting BV status. Correlation with Nugent score varies from 0.19 using Simpson (OTU) to 0.74 using PDu. OTU-based measures rank in the bottom half of the measures tested, and below all phylogenetic measures. PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t 8 In the hierarchical clustering of alpha measures on the vaginal data set, phylogenetic methods are separated from OTU-based methods (Fig. 2). BWPD\u03b8 is similar to different extant phylogenetic alpha diversity measures for different \u03b8. The Simpson and Shannon diversity measures cluster together, as do the ACE and Chao1 richness measures. Fig. 3 shows values of BWPD\u03b8 calculated before (x-axis) and after (y-axis) a single rarefaction to 523 sequences per sample. Samples for which the BWPD\u03b8 value changes little lie close to the blue line, which shows the case of no difference between original and rarefied samples. Increasing \u03b8, which corresponds to increased use of abundance information, reduces the change in BWPD\u03b8 induced by rarefaction. Phylogenetic quadratic entropy and phylogenetic entropy both show behavior similar to BWPD1, with rarefaction introducing little effect. It might be possible to formalize a statement to this effect by computing the expectation of these alpha measures under rarefaction. However, computing the expectation for BWPD\u03b8 under rarefaction does not appear to be straightforward: the methods of Dremin (1994) might be applicable in this setting, however, even the integer moments of the hypergeometric distribution are complicated and the non-integer moments are bound to be very complex. We have, however, shown in the Appendix that the expectation of phylogenetic quadratic entropy under rarefaction to k sequences assigned to the tips of a phylogenetic tree is E[PQEk] = k \u2212 1 kn(n\u2212 1) \ufffd i \ufffdidi(n\u2212 di) where di is the number of sequences falling below edge i and \ufffdi is the length of edge i. This is almost identical to the unrarefied value of phylogenetic quadratic entropy, i.e. PQE = 1 n2 \ufffd i \ufffdidi(n\u2212 di). Thus it is not surprising to see that the expectation of PQE under rarefaction is very close to the original value (Fig. S3) for reasonably large k and n. 3.1.2. Oral microbiome. As previously observed by Griffen et al. (2011a), we find generally higher diversity in samples from dysbiotic patients (Fig. 4). We evaluated the ability of each alpha diversity measure to PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t 9 predict whether a sample came from an individual with periodontitis, regardless of sample collection site, using the above methods. In the oral dataset, phylogenetic alpha diversity measures incorporating abundance gave the best predictions of community state (Table 3, Fig. 4). In contrast, classical phylogenetic diversity performed less well. These results were almost identical in terms of rank order after applying additional quality filtering steps to correct sequencing errors and remove potentially chimeric sequences (Table S1). OTU-based methods and phylogenetic methods are not as separated in a hierarchical clustering as for the vaginal dataset (Fig. S6). However, many of the same pairings are present in both clusterings: BWPD0.5 with phylogenetic entropy, BWPD1 with quadratic entropy, Simpson with Shannon, and ACE with Chao1. Interestingly, PDu, BWPD0.25, and the qD(T) measures all cluster with the discrete richness measures ACE and Chao1. Like the vaginal dataset, incorporating abundance information decreases the effect of rarefaction on BWPD\u03b8 values (Figs. S4, S5). 3.1.3. Skin microbiome. To further assess resolution and robustness of phylogenetic diversity measures, we considered skin microbiome data from a study by Oh et al. (2012). This study tracked the changes of the skin microbiome through \u201cTanner\u201d developmental stages of adolescence(Tanner and Whitehouse, 1976). Because there are five Tanner stages, and they do not have a monotonic relationship with skin microbiome diversity (Oh et al., 2012), we focused on ANOVA p-values to see if the diversity measurements had small within-stage heterogeneity compared to between-stage heterogeneity. To compare the ANOVA p-values associated with the diversity measurements across the various data sets, we ranked the p-value of the diversity measures from lowest to highest for each data set individually. We averaged these ranks to gain an overall measure of performance. The results again show phylogenetic measures generally performing better than OTU-based measures (Table 4). This finding holds true even after removing potential chimeras (Table S1). In this case, a light weighting or no weighting of phylogenetic diversity by abundance performed better than full abundance-weighting. 3.1.4. Applications summary. In all three of the data sets investigated, abundance-weighted phylogenetic diversity measures showed good performance to distinguish between community states: between \u201cnormal\u201d and dysbiotic samples in the oral and vaginal microbiomes, and between developmental stages in the skin microbiome. Notably, PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t 10 the best distinguishing measure in each dataset was phylogenetic; in addition BWPD0.25 and BWPD0.5 were the only measures that were in the top four for all data sets. The result that partial abundance weighting performs well corresponds to analogous results for beta diversity, where an intermediate exponent for \u201cgeneralized UniFrac\u201d was the most powerful (Chen et al., 2012). 4. DISCUSSION Phylogenetic alpha diversity measures were more closely related to community state than were discrete measures based on OTU clustering for the data sets investigated here. This result is especially interesting given that the Simpson index, the Shannon index, or counting applied to OTU tables are very common ways of characterizing microbial diversity (Fierer et al., 2007; Grice et al., 2009; Hill et al., 2003; Dethlefsen and Relman, 2011). As also noted by Aagaard et al. (2012), we find that measurements of diversity using taxonomic classification can be useful in describing communities, and in fact perform much better than the same measurements of diversity applied to OTU counts; however, this approach requires a taxonomically well characterized environment. Our results can be viewed as an experimental confirmation of the notion that incorporating similarity between species is important to get sensible measures of diversity, which has been advocated by many, including most recently by Leinster and Cobbold (2012). We find that classical phylogenetic diversity is sensitive to sampling depth, underestimating the true value in small samples. Biases have also been described for diversity measures using OTU tables (Gihring et al., 2012). In contrast, we observe that some abundanceweighted phylogenetic measures are relatively robust to varying levels of sampling. These results did not appear to be the result of of sequencing issues. In principle, OTU methods could have performed badly because of error-prone and chimeric sequences inflating the number of OTUs. Although this is a real danger for OTU quantification, in this study its impact appears to be limited\u2013 similar results were obtained with the oral data after de-noising and chimera removal and the skin data (which used Sanger sequencing) after chimera removal. We note that on our data, non-phylogenetic measures applied to family level taxonomic groupings are generally more discriminating than the corresponding measures applied to OTUs. This may be PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t 11 because our sequences are from the human microbiome, and taxonomic classification is especially well-developed in that setting; in particular the taxonomic names may already be defined in a way that corresponds to dysbiosis. Thus this particular difference may not continue to be true in a taxonomically less-well-characterized environment. As of the publication of this paper, no abundance-weighted phylogenetic alpha diversity measures are implemented in either mothur (Schloss et al., 2009) or QIIME (Caporaso et al., 2010), two of the most popular tools for analysis of microbial ecology data. Although the fact that abundance-weighted phylogenetic diversity measures performed very well for the three data sets investigated here does not imply that they are best in general, we suggest that abundanceweighted phylogenetic measures be given greater consideration for microbial ecology studies. For this to happen, implementations in commonly used microbial ecology software packages will be needed, in addition to our implementation and that of the picante R package (Kembel et al., 2010). 5. ACKNOWLEDGEMENTS The authors would like to thank Steven Kembel for encouragement and guidance, Steven N. Evans for probability consultation, and David Nipperess for an interesting dialog concerning phylogenetic diversity and rarefaction. The Segre lab at the NIH, in particular Sean Conlan and Julia Oh, were very generous and helpful with the skin data. This work would not have been possible without an ongoing collaboration with David Fredricks, Noah Hoffman, Martin Morgan, and Sujatha Srinivasan at the Fred Hutchinson Cancer Research Center. This work was supported in part by NIH R01 HG005966-01. REFERENCES K. Aagaard, K. Riehle, J. Ma, N. Segata, T.A. Mistretta, C. Coarfa, S. Raza, S. Rosenbaum, I. Van den Veyver, A. Milosavljevic, D. Gevers, C. Huttenhower, J. Petrosino, and J. Versalovic. A metagenomic approach to characterization of the vaginal microbiome signature in pregnancy. PLOS ONE, 7(6):e36466, 2012. B. Allen, M. Kon, and Y. Bar-Yam. A new phylogenetic diversity measure generalizing the Shannon index and its application to phyllostomid bats. The American Naturalist, 174(2):236\u2013243, 2009. PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t 12 R. Amsel, P.A. Totten, C.A. Spiegel, K. Chen, D. Eschenbach, and K.K. Holmes. Nonspecific vaginitis: diagnostic criteria and microbial and epidemiologic associations. The American Journal of Medicine, 74(1):14\u201322, 1983. G.M. Barker. Phylogenetic diversity: a quantitative framework for measurement of priority and achievement in biodiversity conservation. Biological Journal of the Linnean Society, 76(2):165\u2013194, 2002. A. Canty and B.D. Ripley. boot: Bootstrap R (S-Plus) Functions, 2012. R package version 1.3-5. J.G. Caporaso, J. Kuczynski, J. Stombaugh, K. Bittinger, F.D. Bushman, E.K. Costello, N. Fierer, A.G. Pen\u0303a, J.K. Goodrich, J.I. Gordon, et al. Qiime allows analysis of high-throughput community sequencing data. Nature Methods, 7(5):335\u2013336, 2010. A. Chao. Nonparametric estimation of the number of classes in a population. Scandinavian Journal of Statistics, 11(4):pp. 265\u2013270, 1984. A. Chao and S.M. Lee. Estimating the number of classes via sample coverage. Journal of the American Statistical Association, pages 210\u2013 217, 1992. A. Chao, C.H. Chiu, and L. Jost. Phylogenetic diversity measures based on Hill numbers. Philosophical Transactions of the Royal Society B: Biological Sciences, 365(1558):3599\u20133609, 2010. J. Chen, K. Bittinger, E.S. Charlson, C. Hoffmann, J. Lewis, G.D. Wu, R.G. Collman, F.D. Bushman, and H. Li. Associating microbiome composition with environmental covariates using generalized UniFrac distances. Bioinformatics, 28(16):2106\u20132113, 2012. A.C. Davison and D.V. Hinkley. Bootstrap Methods and Their Applications. Cambridge University Press, Cambridge, 1997. ISBN 0-521- 57391-2. L. Dethlefsen and D.A. Relman. Incomplete recovery and individualized responses of the human distal gut microbiota to repeated antibiotic perturbation. Proceedings of the National Academy of Sciences, 108(Supplement 1):4554, 2011. I.M. Dremin. Fractional moments of distributions. JETP letters c/c of Pis\u2019ma v zhurnal eksperimental\u2019noi teoreticheskoi fiziki, 59:585\u2013585, 1994. R.C. Edgar. Search and clustering orders of magnitude faster than BLAST. Bioinformatics, 26(19):2460\u20132461, 2010. D.P. Faith. Conservation evaluation and phylogenetic diversity. Biological Conservation, 61(1):1\u201310, 1992. PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t 13 N. Fierer, M. Breitbart, J. Nulton, P. Salamon, C. Lozupone, R. Jones, M. Robeson, R.A. Edwards, B. Felts, S. Rayhawk, et al. Metagenomic and small-subunit rRNA analyses reveal the genetic diversity of bacteria, archaea, fungi, and viruses in soil. Applied and Environmental Microbiology, 73(21):7059\u20137066, 2007. P.V.A. Fine and S.W. Kembel. Phylogenetic community structure and phylogenetic turnover across space and edaphic gradients in western amazonian tree communities. Ecography, 34(4):552\u2013565, 2011. D.N. Fredricks, T.L. Fiedler, and J.M. Marrazzo. Molecular identification of bacteria associated with bacterial vaginosis. New England Journal of Medicine, 353(18):1899\u20131911, 2005. T.M. Gihring, S.J. Green, and C.W. Schadt. Massively parallel rRNA gene sequencing exacerbates the potential for biased community diversity comparisons due to variable library sizes. Environmental Microbiology, 2012. E.A. Grice, H.H. Kong, S. Conlan, C.B. Deming, J. Davis, A.C. Young, G.G. Bouffard, R.W. Blakesley, P.R. Murray, E.D. Green, et al. Topographical and temporal diversity of the human skin microbiome. Science, 324(5931):1190\u20131192, 2009. A.L. Griffen, C.J. Beall, J.H. Campbell, N.D. Firestone, P.S. Kumar, Z.K. Yang, M. Podar, and E.J. Leys. Distinct and complex bacterial profiles in human periodontitis and health revealed by 16S pyrosequencing. The ISME Journal, 2011a. A.L. Griffen, C.J. Beall, N.D. Firestone, E.L. Gross, J.M. DiFranco, J.H. Hardman, B. Vriesendorp, R.A. Faust, D.A. Janies, and E.J. Leys. CORE: a phylogenetically-curated 16S rDNA database of the core oral microbiome. PLOS ONE, 6(4):e19051, 2011b. F.E. Harrell Jr. Hmisc: Harrell Miscellaneous, 2012. URL http: //CRAN.R-project.org/package=Hmisc. R package version 3.9-3. T.C.J. Hill, K.A. Walsh, J.A. Harris, and B.F. Moffett. Using ecological diversity measures with bacterial communities. FEMS Microbiology Ecology, 43(1):1\u201311, 2003. S.W. Kembel, P.D. Cowan, M.R. Helmus, W.K. Cornwell, H. Morlon, D.D. Ackerly, S.P. Blomberg, and C.O. Webb. Picante: R tools for integrating phylogenies and ecology. Bioinformatics, 26(11):1463\u2013 1464, 2010. Tom Leinster and Christina A Cobbold. Measuring diversity: the importance of species similarity. Ecology, 93(3):477\u2013489, 2012. F.A. Matsen and A. Gallagher. Reconciling taxonomy and phylogenetic inference: formalism and algorithms for describing discord and inferring taxonomic roots. Algorithms for Molecular Biology, 7 PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t 14 (1):8, 2012. F.A. Matsen, R.B. Kodner, and E. Armbrust. pplacer: linear time maximum-likelihood and Bayesian phylogenetic placement of sequences onto a fixed reference tree. BMC Bioinformatics, 11(1):538, 2010. D. McDonald, J.C. Clemente, J. Kuczynski, J.R. Rideout, J. Stombaugh, D. Wendel, A. Wilke, S. Huse, J. Hufnagle, F. Meyer, et al. The Biological Observation Matrix (BIOM) format or: how I learned to stop worrying and love the ome-ome. Giga Science, 1(1):1\u20136, 2012. R.P. Nugent, M.A. Krohn, and SL Hillier. Reliability of diagnosing bacterial vaginosis is improved by a standardized method of gram stain interpretation. Journal of Clinical Microbiology, 29(2):297\u2013301, 1991. Julia Oh, Sean Conlan, E Polley, Julia A Segre, Heidi H Kong, et al. Shifts in human skin and nares microbiota of healthy children and adults. Genome medicine, 4(10):1\u201311, 2012. J. Oksanen, F.G. Blanchet, R. Kindt, P. Legendre, R. Minchin, R.B. O\u2019Hara, G.L. Simpson, P. Solymos, M.H.H. Stevens, and H. Wagner. vegan: Community Ecology Package, 2012. URL http://CRAN. R-project.org/package=vegan. R package version 2.0-4. F. Pardi and N. Goldman. Resource-aware taxon selection for maximizing phylogenetic diversity. Systematic Biology, 56(3):431\u2013444, 2007. M.N. Price, P.S. Dehal, and A.P. Arkin. FastTree 2\u2013approximately maximum-likelihood trees for large alignments. PLOS ONE, 5(3): e9490, 2010. R Development Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria, 2012. URL http://www.R-project.org/. C.R. Rao. Diversity and dissimilarity coefficients: a unified approach. Theoretical Population Biology, 21(1):24\u201343, 1982. P.D. Schloss, S.L. Westcott, T. Ryabin, J.R. Hall, M. Hartmann, E.B. Hollister, R.A. Lesniewski, B.B. Oakley, D.H. Parks, C.J. Robinson, et al. Introducing mothur: open-source, platform-independent, community-supported software for describing and comparing microbial communities. Applied and Environmental Microbiology, 75 (23):7537\u20137541, 2009. C.E. Shannon. A mathematical theory of communication. Bell System Technical Journal, 27(1):379\u2013423, 1948. E.H. Simpson. Measurement of diversity. Nature, 163(4148):688, 1949. PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t 15 S. Srinivasan, N.G. Hoffman, M.T. Morgan, F.A. Matsen, T.L. Fiedler, R.W. Hall, F.J. Ross, C.O. McCoy, R. Bumgarner, J.M. Marrazzo, et al. Bacterial communities in women with bacterial vaginosis: high resolution phylogenetic analyses reveal relationships of microbiota to clinical criteria. PLOS ONE, 7(6):e37818, 2012. JM Tanner and RH Whitehouse. Clinical longitudinal standards for height, weight, height velocity, weight velocity, and stages of puberty. Archives of disease in childhood, 51(3):170\u2013179, 1976. M. Vellend, W.K. Cornwell, K. Magnuson-Ford, and A. Mooers. Measuring phylogenetic biodiversity. In Biological Diversity: Frontiers in Measurement and Assessment, A.E. Magurran and B.J. McGill, editors, pages 194\u2013207. Oxford University Press, 2011. Q. Wang, G.M. Garrity, J.M. Tiedje, and J.R. Cole. Na\u0131\u0308ve Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Applied and environmental microbiology, 73(16): 5261\u20135267, 2007. R.M. Warwick and K.R. Clarke. New \u2018biodiversity\u2019 measures reveal a decrease in taxonomic distinctness with increasing stress. Marine Ecology Progress Series, 129(1):301\u2013305, 1995. PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t Figure 1 \\(g_\\theta\\) curves for various \\(\\theta\\) parameters. as \\(\\theta\\) goes to zero, the \\(g_\\theta\\) converge pointwise to \\(g\\), which is 1 on the : interior of the unit interval and 0 on the boundaries. PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t Figure 2 Dendrogram relating alpha diversity measures applied to the vaginal dataset. PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t Figure 3 Comparison of rarefied and unrarefied values of various phylogenetic alpha diversity measures as applied to the vaginal dataset. The value of six alpha measures for each specimen using all available sequences is plotted on the \\(x\\)-axis. The value of the alpha measures for each specimen after a single rarefaction to 523 sequences (the smallest sequence count across specimens) is plotted on the \\(y\\)-axis. The \\(y=x\\) line is shown in blue. PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t Figure 4 Comparison of diversity between samples from healthy controls, healthy sites of dysbiotic patients, and dysbiotic sites of dysbiotic patients on the oral dataset, using various measures of diversity. \"Shallow\" means a shallow pocket between tooth and gum tissue, while \"deep\" means a sample from a deep pocket between gum tissue that has separated from its tooth. Top row: cluster-based methods. Bottom rows: phylogenetic methods. PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t Table 1(on next page) Overview of phylogenetic diversity measures used in the text. PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t phylogenetic diversity (Faith, 1992) phylogenetic generalization of species count phylo. quadratic entropy (Rao, 1982; Warwick and Clarke, 1995) phylogenetic generalization of the Simpson index phylogenetic entropy (Allen et al., 2009) phylogenetic generalization of the Shannon index qD(T) (Chao et al., 2010) phylogenetic generalization of Hill numbers BWPD1 (Barker, 2002; Vellend et al., 2011) abundance-weighted version of phylogenetic diversity BWPD\u03b8 (this paper) one-parameter family interpolating between PD and BWPD1 TABLE 1. Overview of phylogenetic diversity measures used in the text. PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t Table 2(on next page) Correlation and predictive performance of the various alpha diversity measures applied to the vaginal data set. Rows are ordered by increasing mean rank across performance measurements. Nugent \\ (R^2\\): \\(R^2\\) value using the measure as a predictor, and the Nugent score as response in a linear model. Amsel accuracy: proportion of specimens with correct BV classification under a leave-one-out cross-validation. Amsel p-value: p-value from a two-sample \\(t\\)-test on values stratified by BV classification. ``OTU'' designates the measure applied to 97\\% clustering groups, and ``Family'' designates taxonomic classification at the family level. Measures described in main text. PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t Measure Amsel Accuracy Nugent R2 Amsel p-value mean rank PDu 0.834 0.737 1.84E-35 1.3 BWPD0.25 0.836 0.735 1.98E-35 2.0 Simpson (Family) 0.817 0.735 4.11E-33 4.0 BWPD0.5 0.827 0.700 3.33E-33 4.3 Shannon (Family) 0.813 0.724 2.28E-32 5.3 Phylo. entropy 0.831 0.679 1.81E-31 5.7 Chao1 (Family) 0.813 0.704 6.27E-31 7.0 0.5D(T) 0.818 0.658 7.47E-29 8.0 0.25D(T) 0.809 0.682 2.25E-30 8.3 Phylo. quad. entropy 0.813 0.648 7.89E-30 9.0 BWPD1 0.795 0.611 5.38E-28 11.0 Chao1 (OTU) 0.766 0.488 1.64E-23 12.7 ACE (Family) 0.766 0.491 2.82E-11 13.0 ACE (OTU) 0.764 0.469 6.82E-22 13.7 Shannon (OTU) 0.758 0.380 5.27E-16 14.7 Simpson (OTU) 0.697 0.191 1.42E-07 16.0 TABLE 2. Correlation and predictive performance of the various alpha diversity measures applied to the vaginal data set. Rows are ordered by increasing mean rank across performance measurements. Nugent R2: R2 value using the measure as a predictor, and the Nugent score as response in a linear model. Amsel accuracy: proportion of specimens with correct BV classification under a leave-one-out cross-validation. Amsel p-value: p-value from a two-sample t-test on values stratified by BV classification. \u201cOTU\u201d designates the measure applied to 97% clustering groups, and \u201cFamily\u201d designates taxonomic classification at the family level. Measures described in main text. PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t Table 3(on next page) Predictive accuracy of each measure in the oral dataset and p-value from an ANOVA stratified by disease status and sampling site. PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t Measure Diseased status accuracy ANOVA p-value mean rank Phylo. entropy 0.791 4.97E-09 1.0 BWPD0.5 0.782 6.50E-09 2.0 BWPD0.25 0.755 7.16E-08 4.0 Phylo. quad. entropy 0.770 2.47E-07 4.0 Simpson (Family) 0.776 1.45E-06 4.0 0.5D(T) 0.734 4.74E-06 6.5 0.25D(T) 0.735 4.33E-05 7.0 PDu 0.691 6.37E-06 8.5 Shannon (Family) 0.734 5.32E-05 8.5 BWPD1 0.698 3.57E-04 9.5 Chao1 (OTU) 0.685 9.94E-04 11.0 ACE (OTU) 0.682 1.30E-03 12.0 Simpson (OTU) 0.676 2.39E-02 13.5 Shannon (OTU) 0.672 1.31E-03 14.0 Chao1 (Family) 0.674 2.64E-01 15.0 ACE (Family) 0.663 1.82E-01 15.5 TABLE 3. Predictive accuracy of each measure in the oral dataset and p-value from an ANOVA stratified by disease status and sampling site PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t Table 4(on next page) ANOVA p-values for various diversity statistics applied to the skin microbiome data of oh et al. (2013). : Rows are ordered by increasing mean rank across sites. The same site abbreviations are used as in their paper: Ac, antecubital fossa; N, nares; Pc, popliteal fossa; Vf, volar forearm. PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t Ac N Pc Vf mean rank PDu 3.34e-02 4.94e-03 3.03e-03 2.28e-04 4.50 BWPD0.25 4.22e-02 1.32e-03 6.37e-03 5.86e-04 5.25 BWPD0.5 8.54e-02 9.85e-05 3.65e-02 5.85e-03 6.00 Shannon (OTU) 6.61e-02 9.48e-02 9.65e-02 1.05e-05 6.50 Chao1 (OTU) 8.00e-02 6.46e-03 3.98e-03 3.18e-03 6.75 Phylo. quad. entropy 2.52e-01 1.12e-05 4.99e-01 1.67e-01 7.50 Phylo. entropy 1.37e-01 1.15e-03 1.55e-01 2.09e-02 7.75 0.5D(T) 8.91e-01 5.63e-04 3.84e-03 9.09e-01 8.25 0.25D(T) 7.00e-01 2.27e-03 2.35e-03 9.41e-01 8.50 BWPD1 3.09e-01 5.95e-05 6.65e-01 5.41e-01 8.50 0D(T) 4.42e-01 1.05e-02 1.38e-03 9.38e-01 8.75 Simpson (OTU) 9.38e-02 4.01e-01 8.49e-01 1.01e-04 8.75 TABLE 4. ANOVA p-values for various diversity statistics applied to the skin microbiome data of Oh et al. (2012). Rows are ordered by increasing mean rank across sites. The same site abbreviations are used as in their paper: Ac, antecubital fossa; N, nares; Pc, popliteal fossa; Vf, volar forearm. PeerJ reviewing PDF | (v2013:04:390:1:1:NEW 19 Aug 2013) R ev ie w in g M an us cr ip t",
+    "url": "https://peerj.com/articles/158/reviews/",
+    "review_1": "Lesley Rogers \u00b7 Aug 21, 2013 \u00b7 Academic Editor\nACCEPT\nThank you for making the final corrections.",
+    "review_2": "Lesley Rogers \u00b7 Aug 19, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nThe following points still need to be addressed.\n1. The following sentence in the Introduction is not entirely clear to me. \u201cIn her view, basic level categories may be categorized perceptually before superordinate categories, but with regards to conceptual categories, which are based on kind as opposed to perceptual similarity, more global, abstract categories such as animals, foods etc. may emerge first.\u201d. Please can you improve it?\n2. In the new sentence \u201cWe predicted that orangutans may readily category stimuli from both concrete and intermediate level categories, whereas the gorilla might categorize stimuli more readily at the concrete level.\u201d Correct category to categorize.\n2. You use the word \u2018hoofstock\u2019 in the Introduction but it is not exactly clear what is meant by this.\n3. The first sentence of the Results and Discussion section of Experiment 1 states \u201cAs shown in Figure 2 and confirmed by binomial tests, each subject performed significantly greater than chance..\u201d. Please insert \u2018at a level\u2019 after \u2018performed\u2019.\n4. The second sentence under the heading Experiment 2 doubles up on commas and brackets - ,( and ,). Please choose which of these you want to use.\n4.One of the reviewers said, \u201cI found the tables and figures rather confusing. In table two, the caption says \"percentages\", yet the numbers are all less than 1. And you replied, \u201cA: The numbers are less than one because percentage correct was calculated as a score between 0 and 1 (e.g. 55% = .55).\u201d. Tables 2 and 4 remain a problem because they are not percent scores. Please change to the 0 to 100% range.\n5. In response to a reviewer you changed the following sentence: \u201cAlthough the current study was not designed to address developmental changes in concept acquisition, the fact that at least one gorilla showed greater facility with concrete level categories is interesting and should be explored further.\u201d Replace \u2018should be explored\u2019 by \u2018worth exploring\u2019.",
+    "review_3": "Lesley Rogers \u00b7 Aug 15, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nPlease consider all points raised by the reviewers and address them by making appropriate changes to your manuscript.",
+    "review_4": "Lynda Birke \u00b7 Aug 13, 2013\nBasic reporting\nOverall, an interesting paper and generally well written, this study will be of interest to behavioural scientists. I think the paper is publishable, subject to some revisions.\nFirst, at a general level, animals' discriminative skills should be discussed with some reference to perceptual differences. The subjects (great apes) were tested using visual discriminations, based on photos, which is no doubt appropriate for these species. However, the author also makes reference to discriminative skills in other species (eg dogs) whose skills may well have different perceptual bases. The paper should acknowledge that the tests used were specifically testing visual skills, and contextualise discussion accordingly.\nSecond, there should be more discussion of generalisation from the relatively small number of animals used. I am well aware of the difficulties of doing studies with zoo animals - it almost always involves a small N. But here, I felt the paper would benefit from considering how general the demonstrated abilities are (especially for gorillas).\nMore specific points: there should be more detail of how the apes were kept. This is particularly relevant regarding proximity of other species, and the subjects' experience of other primate species. Following on from this, I am sceptical of the implication that specific skills \"pre-existed\" (p. 24) - ie are hard-wired - unless we know much more about these individuals' histories.\nPage 8, paragraph 2 - this was not entirely clear regarding chimps. Please clarify what \"most difficulty with the most abstract distinctions\" means.\nPage 9, lines 4-5 - needs clarification.\nP.12. Please explain why different numbers of trials were used.\np.26. Discussion of how humans classify other species should acknowledge cross-cultural differences in classification.\nI found the tables and figures rather confusing. In table two, the caption says \"percentages\", yet the numbers are all less than 1. The figures do not seem to show any significant improvement over time - please add more information.\nExperimental design\nThe overall design is satisfactory, although I would encourage the author to include more detail of statistics. In particular, the justification of the ANOVA, and the assumptions made in its use, should be clarified (c. page 15)\nValidity of the findings\nSee comments above.\nCite this review as\nBirke L (2013) Peer Review #1 of \"Matching based on biological categories in Orangutans (Pongo abelii) and a Gorilla (Gorilla gorilla gorilla) (v0.1)\". PeerJ https://doi.org/10.7287/peerj.158v0.1/reviews/1",
+    "pdf_1": "https://peerj.com/articles/158v0.3/submission",
+    "pdf_2": "https://peerj.com/articles/158v0.2/submission",
+    "review_5": "Reviewer 2 \u00b7 Aug 6, 2013\nBasic reporting\n(referring to introduction): excellent.\n\nsee general comments also\nExperimental design\nThorough and circumspect and avoiding many potential pitfalls. The researchers took every possible precaution to avoid preferences and ambiguous interpretations of results. Because of the design the experimenter was able to determine that subjects did not rely upon associations they formed during the course of the experiment between particular stimuli pairings and reward in performing this task.\nValidity of the findings\nThe experiments are carefully controlled, the test variables rigorously controlled and the statistical tests appropriate. The results therefore seem robust and the findings very convincing. We are also told the prehistory of experiences of these specific great apes taking part in these experiments reassuring one that these were novel experiences and that it was not prior learning or extensive testing which had produced the current results.\nAdditional comments\nThis is an exciting paper concerned with a possible distinction recognized and known in humans between specific and concrete and global categorizations of objects, object classes and overall classifications without the benefit of language. They used biological classifications of species, family and class. Four orangutans and one gorilla were presented with delayed matching-to-sample tasks and, in two experiments, taken through levels of perceptual matching to abstract matching., i.e. same species group or different families and, finally, fitting individual images into appropriate classes of insect, reptiles, birds or mammals respectively. Impressively, the paper found (for the first experiment) that five of the apes performed above chance levels within the first six sessions. For the second experiment testing the apes\u2019 ability to assign specific photos to classes most apes performed again well above chance, suggesting either that they spontaneously recognised the categories or were able to learn about such categories with relative ease and speed. The authors were cautious and parsimonious in their interpretation and, despite their strong results, resisted the temptation to attribute abstract thinking to the apes but they are saying that biological categorisations do not require specialised taxonomy knowledge but, in this case, a recognition of the basic rules governing the classifications. Their experiments have opened the debate for further studies in this field on how categorisation occurs in pre-linguistic contexts.\nI have very little to add in a critical sense and highly recommend publishing.\nThere are a few minor grammar and other errors in the paper that the author should correct before final submission. One glaring example was the formulation that traits are \u2018shared in common\u2019 \u2013a bad logical error/doubling up: it\u2019s either \u2018have in common\u2019 or \u2018share xyz with someone\u2019\u2026(leaving out \u2018in common\u2019) because when one has certain traits in common with someone else they are, in fact, shared traits (the male stallion problem).\nThere are other small (i.e. readily fixable) issues in presentation: The discussion/conclusion is not as good and tight as the introduction and I would hope that the author could strengthen the discussion by referring back to the introduction in which the categories and distinctions were very well presented indeed and theoretically cogent. Some of the sparkle is gone in the discussion and that is a pity and could be improved simply by using and expanding upon the same sharply defined theoretical concepts and issues as in the beginning of the paper.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"Matching based on biological categories in Orangutans (Pongo abelii) and a Gorilla (Gorilla gorilla gorilla) (v0.1)\". PeerJ https://doi.org/10.7287/peerj.158v0.1/reviews/2",
+    "pdf_3": "https://peerj.com/articles/158v0.1/submission",
+    "review 6": "Reviewer 3 \u00b7 Aug 1, 2013\nBasic reporting\nNo comments\nExperimental design\nNo comments\nValidity of the findings\nNo comments\nAdditional comments\nThis paper investigates the conceptual organization of biological stimuli in perceptual and conceptual categories by non-human primates (gorillas and orangoutangs). In the human literature there has been a debate on the role of language vs. perceptual similarity for the formation of categories, in relation to the level of abstraction implied by a given category. The present study aims at investigating this topic from a comparative perspective, testing non-verbal animals that are phylogenetically close to humans. Categorizing stimuli according to biological taxonomies has hypothesized to be a uniquely human tendency. Contrary to that claim, recent neuroimaging and behavioral data suggested that non-human primates may spontaneously form categories of biological objects in a similar way to what humans do. The current paper investigated the ability to make explicit classifications of natural class distinctions, to determine whether exemplars of more closely related groupings are more readily categorized together compared to more distantly related members of the same class. In the present study, overall 5 subjects were required to match images based on biological classifications at the level of species, family or class. The results of the first experiment indicate that the subjects can form categories at the concrete (species) level, even when confronted with images of unfamiliar primate species. This task however can be solved on the basis of perceptual similarity and does not imply sophisticate abstraction abilities. In Experiment 2 subjects proved able to solve the task also based on a more abstract level of categorization, i.e. using the intermediate categories of taxonomic classes. Moreover, at least orangoutangs proved able to learn intermediate-level categories as readily (or even more readily) as concrete categories, in line with the human developmental literature. On the contrary the gorillas seemed to find it easier to form categories at the concrete level. However, the small sample size prevents from making meaningful comparisons between the performance of the two species (orangoutangs and gorillas). Unfortunately the present study cannot provide any information on the role of experience with different living creatures in the relative facilitation for forming intermediate-level categories displayed by some of the subjects (for example see page 23, first paragraph). This most interesting issue can be assessed only by investigating the performance of subjects reared in strictly controlled environment and/or tested at a very early age after birth. Authors should be careful in their claims on this point (e.g., at page 24). The introduction contains some repeated information, which should be avoided (e.g. the work of Autier-D\u00e9rian et al., 2013 is described two times for no apparent reason). Moreover, the structure of the introduction itself is somehow unclear, some topics seem to assed repeatedly in different parts of the text. It would be helpful to make the introduction more dense and to provide a clearer structure for the reader. It could be also helpful to add, at the end of the introduction, a short paragraph highlighting the main conclusions and open issues originated from the existing literature. Similar problems are present also in the general discussion. The most problematic issue with the present paper is that it is not clear which are the conclusions that can be drawn from the results obtained. It is necessary to highlight the open questions that the study wanted to asses, and how (if) the results obtained provide novel information to answer these questions. Minor comments In the discussion session evidence is reported that categorization of own species images tends to occur at a more concrete level than categorization of distantly related species (page 24 and 25). This is likely to be due to the special status of own-species, which could be perceived as a category per se and elicit more detailed elaboration. It could be also appropriate to discuss this evidence in relation to the \"other species\" effect observed during face perception in humans. The abstract should mention which are the conclusions obtained from the present study. It would be very helpful to include images of the stimuli, since their perceptual appearance is crucial to the interpretation of the results.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #3 of \"Matching based on biological categories in Orangutans (Pongo abelii) and a Gorilla (Gorilla gorilla gorilla) (v0.1)\". PeerJ https://doi.org/10.7287/peerj.158v0.1/reviews/3",
+    "all_reviews": "Review 1: Lesley Rogers \u00b7 Aug 21, 2013 \u00b7 Academic Editor\nACCEPT\nThank you for making the final corrections.\nReview 2: Lesley Rogers \u00b7 Aug 19, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nThe following points still need to be addressed.\n1. The following sentence in the Introduction is not entirely clear to me. \u201cIn her view, basic level categories may be categorized perceptually before superordinate categories, but with regards to conceptual categories, which are based on kind as opposed to perceptual similarity, more global, abstract categories such as animals, foods etc. may emerge first.\u201d. Please can you improve it?\n2. In the new sentence \u201cWe predicted that orangutans may readily category stimuli from both concrete and intermediate level categories, whereas the gorilla might categorize stimuli more readily at the concrete level.\u201d Correct category to categorize.\n2. You use the word \u2018hoofstock\u2019 in the Introduction but it is not exactly clear what is meant by this.\n3. The first sentence of the Results and Discussion section of Experiment 1 states \u201cAs shown in Figure 2 and confirmed by binomial tests, each subject performed significantly greater than chance..\u201d. Please insert \u2018at a level\u2019 after \u2018performed\u2019.\n4. The second sentence under the heading Experiment 2 doubles up on commas and brackets - ,( and ,). Please choose which of these you want to use.\n4.One of the reviewers said, \u201cI found the tables and figures rather confusing. In table two, the caption says \"percentages\", yet the numbers are all less than 1. And you replied, \u201cA: The numbers are less than one because percentage correct was calculated as a score between 0 and 1 (e.g. 55% = .55).\u201d. Tables 2 and 4 remain a problem because they are not percent scores. Please change to the 0 to 100% range.\n5. In response to a reviewer you changed the following sentence: \u201cAlthough the current study was not designed to address developmental changes in concept acquisition, the fact that at least one gorilla showed greater facility with concrete level categories is interesting and should be explored further.\u201d Replace \u2018should be explored\u2019 by \u2018worth exploring\u2019.\nReview 3: Lesley Rogers \u00b7 Aug 15, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nPlease consider all points raised by the reviewers and address them by making appropriate changes to your manuscript.\nReview 4: Lynda Birke \u00b7 Aug 13, 2013\nBasic reporting\nOverall, an interesting paper and generally well written, this study will be of interest to behavioural scientists. I think the paper is publishable, subject to some revisions.\nFirst, at a general level, animals' discriminative skills should be discussed with some reference to perceptual differences. The subjects (great apes) were tested using visual discriminations, based on photos, which is no doubt appropriate for these species. However, the author also makes reference to discriminative skills in other species (eg dogs) whose skills may well have different perceptual bases. The paper should acknowledge that the tests used were specifically testing visual skills, and contextualise discussion accordingly.\nSecond, there should be more discussion of generalisation from the relatively small number of animals used. I am well aware of the difficulties of doing studies with zoo animals - it almost always involves a small N. But here, I felt the paper would benefit from considering how general the demonstrated abilities are (especially for gorillas).\nMore specific points: there should be more detail of how the apes were kept. This is particularly relevant regarding proximity of other species, and the subjects' experience of other primate species. Following on from this, I am sceptical of the implication that specific skills \"pre-existed\" (p. 24) - ie are hard-wired - unless we know much more about these individuals' histories.\nPage 8, paragraph 2 - this was not entirely clear regarding chimps. Please clarify what \"most difficulty with the most abstract distinctions\" means.\nPage 9, lines 4-5 - needs clarification.\nP.12. Please explain why different numbers of trials were used.\np.26. Discussion of how humans classify other species should acknowledge cross-cultural differences in classification.\nI found the tables and figures rather confusing. In table two, the caption says \"percentages\", yet the numbers are all less than 1. The figures do not seem to show any significant improvement over time - please add more information.\nExperimental design\nThe overall design is satisfactory, although I would encourage the author to include more detail of statistics. In particular, the justification of the ANOVA, and the assumptions made in its use, should be clarified (c. page 15)\nValidity of the findings\nSee comments above.\nCite this review as\nBirke L (2013) Peer Review #1 of \"Matching based on biological categories in Orangutans (Pongo abelii) and a Gorilla (Gorilla gorilla gorilla) (v0.1)\". PeerJ https://doi.org/10.7287/peerj.158v0.1/reviews/1\nReview 5: Reviewer 2 \u00b7 Aug 6, 2013\nBasic reporting\n(referring to introduction): excellent.\n\nsee general comments also\nExperimental design\nThorough and circumspect and avoiding many potential pitfalls. The researchers took every possible precaution to avoid preferences and ambiguous interpretations of results. Because of the design the experimenter was able to determine that subjects did not rely upon associations they formed during the course of the experiment between particular stimuli pairings and reward in performing this task.\nValidity of the findings\nThe experiments are carefully controlled, the test variables rigorously controlled and the statistical tests appropriate. The results therefore seem robust and the findings very convincing. We are also told the prehistory of experiences of these specific great apes taking part in these experiments reassuring one that these were novel experiences and that it was not prior learning or extensive testing which had produced the current results.\nAdditional comments\nThis is an exciting paper concerned with a possible distinction recognized and known in humans between specific and concrete and global categorizations of objects, object classes and overall classifications without the benefit of language. They used biological classifications of species, family and class. Four orangutans and one gorilla were presented with delayed matching-to-sample tasks and, in two experiments, taken through levels of perceptual matching to abstract matching., i.e. same species group or different families and, finally, fitting individual images into appropriate classes of insect, reptiles, birds or mammals respectively. Impressively, the paper found (for the first experiment) that five of the apes performed above chance levels within the first six sessions. For the second experiment testing the apes\u2019 ability to assign specific photos to classes most apes performed again well above chance, suggesting either that they spontaneously recognised the categories or were able to learn about such categories with relative ease and speed. The authors were cautious and parsimonious in their interpretation and, despite their strong results, resisted the temptation to attribute abstract thinking to the apes but they are saying that biological categorisations do not require specialised taxonomy knowledge but, in this case, a recognition of the basic rules governing the classifications. Their experiments have opened the debate for further studies in this field on how categorisation occurs in pre-linguistic contexts.\nI have very little to add in a critical sense and highly recommend publishing.\nThere are a few minor grammar and other errors in the paper that the author should correct before final submission. One glaring example was the formulation that traits are \u2018shared in common\u2019 \u2013a bad logical error/doubling up: it\u2019s either \u2018have in common\u2019 or \u2018share xyz with someone\u2019\u2026(leaving out \u2018in common\u2019) because when one has certain traits in common with someone else they are, in fact, shared traits (the male stallion problem).\nThere are other small (i.e. readily fixable) issues in presentation: The discussion/conclusion is not as good and tight as the introduction and I would hope that the author could strengthen the discussion by referring back to the introduction in which the categories and distinctions were very well presented indeed and theoretically cogent. Some of the sparkle is gone in the discussion and that is a pity and could be improved simply by using and expanding upon the same sharply defined theoretical concepts and issues as in the beginning of the paper.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"Matching based on biological categories in Orangutans (Pongo abelii) and a Gorilla (Gorilla gorilla gorilla) (v0.1)\". PeerJ https://doi.org/10.7287/peerj.158v0.1/reviews/2\nReview 6: Reviewer 3 \u00b7 Aug 1, 2013\nBasic reporting\nNo comments\nExperimental design\nNo comments\nValidity of the findings\nNo comments\nAdditional comments\nThis paper investigates the conceptual organization of biological stimuli in perceptual and conceptual categories by non-human primates (gorillas and orangoutangs). In the human literature there has been a debate on the role of language vs. perceptual similarity for the formation of categories, in relation to the level of abstraction implied by a given category. The present study aims at investigating this topic from a comparative perspective, testing non-verbal animals that are phylogenetically close to humans. Categorizing stimuli according to biological taxonomies has hypothesized to be a uniquely human tendency. Contrary to that claim, recent neuroimaging and behavioral data suggested that non-human primates may spontaneously form categories of biological objects in a similar way to what humans do. The current paper investigated the ability to make explicit classifications of natural class distinctions, to determine whether exemplars of more closely related groupings are more readily categorized together compared to more distantly related members of the same class. In the present study, overall 5 subjects were required to match images based on biological classifications at the level of species, family or class. The results of the first experiment indicate that the subjects can form categories at the concrete (species) level, even when confronted with images of unfamiliar primate species. This task however can be solved on the basis of perceptual similarity and does not imply sophisticate abstraction abilities. In Experiment 2 subjects proved able to solve the task also based on a more abstract level of categorization, i.e. using the intermediate categories of taxonomic classes. Moreover, at least orangoutangs proved able to learn intermediate-level categories as readily (or even more readily) as concrete categories, in line with the human developmental literature. On the contrary the gorillas seemed to find it easier to form categories at the concrete level. However, the small sample size prevents from making meaningful comparisons between the performance of the two species (orangoutangs and gorillas). Unfortunately the present study cannot provide any information on the role of experience with different living creatures in the relative facilitation for forming intermediate-level categories displayed by some of the subjects (for example see page 23, first paragraph). This most interesting issue can be assessed only by investigating the performance of subjects reared in strictly controlled environment and/or tested at a very early age after birth. Authors should be careful in their claims on this point (e.g., at page 24). The introduction contains some repeated information, which should be avoided (e.g. the work of Autier-D\u00e9rian et al., 2013 is described two times for no apparent reason). Moreover, the structure of the introduction itself is somehow unclear, some topics seem to assed repeatedly in different parts of the text. It would be helpful to make the introduction more dense and to provide a clearer structure for the reader. It could be also helpful to add, at the end of the introduction, a short paragraph highlighting the main conclusions and open issues originated from the existing literature. Similar problems are present also in the general discussion. The most problematic issue with the present paper is that it is not clear which are the conclusions that can be drawn from the results obtained. It is necessary to highlight the open questions that the study wanted to asses, and how (if) the results obtained provide novel information to answer these questions. Minor comments In the discussion session evidence is reported that categorization of own species images tends to occur at a more concrete level than categorization of distantly related species (page 24 and 25). This is likely to be due to the special status of own-species, which could be perceived as a category per se and elicit more detailed elaboration. It could be also appropriate to discuss this evidence in relation to the \"other species\" effect observed during face perception in humans. The abstract should mention which are the conclusions obtained from the present study. It would be very helpful to include images of the stimuli, since their perceptual appearance is crucial to the interpretation of the results.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #3 of \"Matching based on biological categories in Orangutans (Pongo abelii) and a Gorilla (Gorilla gorilla gorilla) (v0.1)\". PeerJ https://doi.org/10.7287/peerj.158v0.1/reviews/3\n",
+    "LLM_extracted_review": "['**Review 1:** Thank you for making the final corrections.\\n\\n**Review 2:**\\n1. The following sentence in the Introduction is not entirely clear to me. \u201cIn her view, basic level categories may be categorized perceptually before superordinate categories, but with regards to conceptual categories, which are based on kind as opposed to perceptual similarity, more global, abstract categories such as animals, foods etc. may emerge first.\u201d. Please can you improve it?\\n2. In the new sentence \u201cWe predicted that orangutans may readily category stimuli from both concrete and intermediate level categories, whereas the gorilla might categorize stimuli more readily at the concrete level.\u201d Correct category to categorize.\\n3. You use the word \u2018hoofstock\u2019 in the Introduction but it is not exactly clear what is meant by this.\\n4. The first sentence of the Results and Discussion section of Experiment 1 states \u201cAs shown in Figure 2 and confirmed by binomial tests, each subject performed significantly greater than chance..\u201d. Please insert \u2018at a level\u2019 after \u2018performed\u2019.\\n5. The second sentence under the heading Experiment 2 doubles up on commas and brackets - ,( and ,). Please choose which of these you want to use.\\n6. One of the reviewers said, \u201cI found the tables and figures rather confusing. In table two, the caption says \"percentages\", yet the numbers are all less than 1. And you replied, \u201cA: The numbers are less than one because percentage correct was calculated as a score between 0 and 1 (e.g. 55% = .55).\u201d. Tables 2 and 4 remain a problem because they are not percent scores. Please change to the 0 to 100% range.\\n7. In response to a reviewer you changed the following sentence: \u201cAlthough the current study was not designed to address developmental changes in concept acquisition, the fact that at least one gorilla showed greater facility with concrete level categories is interesting and should be explored further.\u201d Replace \u2018should be explored\u2019 by \u2018worth exploring\u2019.\\n\\n**Review 3:** Please consider all points raised by the reviewers and address them by making appropriate changes to your manuscript.\\n\\n**Review 4:**\\n1. At a general level, animals\\' discriminative skills should be discussed with some reference to perceptual differences.\\n2. There should be more discussion of generalisation from the relatively small number of animals used.\\n3. There should be more detail of how the apes were kept, particularly regarding proximity of other species and the subjects\\' experience of other primate species.\\n4. Page 8, paragraph 2 - this was not entirely clear regarding chimps. Please clarify what \"most difficulty with the most abstract distinctions\" means.\\n5. Page 9, lines 4-5 - needs clarification.\\n6. P.12. Please explain why different numbers of trials were used.\\n7. P.26. Discussion of how humans classify other species should acknowledge cross-cultural differences in classification.\\n8. I found the tables and figures rather confusing. In table two, the caption says \"percentages\", yet the numbers are all less than 1. The figures do not seem to show any significant improvement over time - please add more information.\\n\\n**Review 5:**\\n1. The discussion/conclusion is not as good and tight as the introduction and could be improved by referring back to the introduction.\\n2. There are a few minor grammar and other errors in the paper that the author should correct before final submission.\\n3. One glaring example was the formulation that traits are \u2018shared in common\u2019 \u2013 a bad logical error/doubling up.\\n4. Some of the sparkle is gone in the discussion and could be improved simply by using and expanding upon the same sharply defined theoretical concepts and issues as in the beginning of the paper.\\n\\n**Review 6:**\\n1. The introduction contains some repeated information, which should be avoided.\\n2. The structure of the introduction itself is somehow unclear; some topics seem to be addressed repeatedly in different parts of the text.\\n3. It would be helpful to make the introduction more dense and to provide a clearer structure for the reader.\\n4. The most problematic issue with the present paper is that it is not clear which conclusions can be drawn from the results obtained.\\n5. The abstract should mention which conclusions were obtained from the present study.\\n6. It would be very helpful to include images of the stimuli, since their perceptual appearance is crucial to the interpretation of the results.']"
+}

peerj_json_files/PeerJ_Json_115.json ADDED Viewed

The diff for this file is too large to render. See raw diff

peerj_json_files/PeerJ_Json_116.json ADDED Viewed

	@@ -0,0 +1,17 @@

+{
+    "v1_Abstract": "BD-Func (BiDirectional FUNCtional enrichment) is an algorithm that calculates functional enrichment by comparing lists of pre-defined genes that are known to be activated versus inhibited in a pathway or by a regulatory molecule. This paper shows that BD-Func can correctly predict cell line alternations and patient characteristics with accuracy comparable to popular algorithms, with a significantly faster run-time. BD-Func can compare scores for individual samples across multiple groups as well as provide predictive statistics and receiver operating characteristic (ROC) plots to quantify the accuracy of the signature associated with a binary phenotypic variable. BD-Func facilitates collaboration and reproducibility by encouraging users to share novel molecular signatures in the BD-Func discussion group, which is where the novel progesterone receptor and LBH589 signatures from this paper can be found. The novel LBH589 signature presented in this paper also serves as a case study showing how a custom signature using cell line data can accurately predict activity in vivo. This software is available to download at https://sourceforge.net/projects/bdfunc/. 2",
+    "v1_col_introduction": "introduction  : Systems-level analysis of the combined expression pattern of multiple genes can be more\ninformative than the expression pattern of an individual gene, and there are a number of tools to calculate functional enrichment of differentially expressed genes (Huang et al. 2009; Naeem et al. 2012; Nam & Kim 2008). However, many functional annotations merely list membership in a pathway or ontology without explicitly modelling genes that should show activation or inhibition. For example, consider the KEGG canonical Wnt signalling pathway (Figure 1) (Kanehisa & Goto 2000). This gene list includes molecules that both activate and inhibit the pathway, resulting in different phenotypes (Dellinger et al. 2012; Logan & Nusse 2004). However, many functional enrichment tools would expect all the members of the pathway to behave similarly (Figure 1C), such that up-regulation of a mix of activators and inhibitors can receive a higher score than selective up-regulation of only activators within the pathway. For example, the most standard method for functional enrichment is to calculate over-representation of one gene list in another gene list, possibly using a Fisher\u2019s exact test or hypergeometric test; in the example described above, this sort of statistical test would ask if a list of differentially expressed genes shows a higher number of Wnt signalling genes than expected by chance. This sort of test cannot differentiate the behavior of those genes unless more detailed gene lists are defined (such as Wnt-inhibitors versus Wnt-agonists). This is a basic problem that BD-Func (BiDirectional FUNCtional enrichment) is designed to overcome.\nMost functional enrichment tools either require upstream filtering of gene lists (FuncAssociate\n(Berriz et al. 2009), GATHER (Chang & Nevins 2006), DAVID (Huang et al. 2008), Connectivity Map (Lamb et al. 2006), WebGestalt (Zhang et al. 2005), GoMiner (Zeeberg et al. 2003), ErmineJ (Lee et al. 2005)) and/or a comparison of signal intensities between two groups (T-profiler (Boorsma et al. 2005), GSVA (Hanzelmann et al. 2013), PAGE (Kim & Volsky 2005), GSEA (Subramanian et al. 2005), ErmineJ (Lee et al. 2005)). However, BD-Func compares the relative expression levels between activated and inhibited genes, and we show that BD-Func can successfully analyze either fold-change values between populations or raw intensity / expression values (for both microarray and RNA-Seq data). Signalling Pathway Impact Analysis (Draghici et al. 2007; Tarca et al. 2008) can be used to model activation and inhibition within a graph, but that algorithm primarily focuses on network topology (which is not always known); in contrast, BD-Func uses a simpler assumption of binning genes into two categories (activation or inhibition). Additionally, the ability to analyze absolute expression values in a single sample is a unique\n21\n22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37\n38 39 40 41 42 43 44 45 46 47 48 49\nPeerJ reviewing PDF | (v2013:06:581:1:0:NEW 12 Aug 2013)\nR ev ie w in g M an\nus cr ip t\nfeature only present in a limited number of functional genomic tools. For example, ASSESS (Edelman et al. 2006) can theoretically predict functional enrichment in a single sample, but that score is not taken into context amongst other samples: in other words, BD-Func uniquely uses a single sample enrichment score as a classifier, which may be useful for personalized medicine research or for the development of novel molecular signatures where the user may need to quantify the utility of the signature as a classifier. Additionally, correlations between a single sample and various samples within a database (such as SPIED (Williams 2012)) can provide information about a single sample, but this requires having a database of samples for comparison (so, this strategy will not work without the presence of an external database). Finally, BD-Func users are encouraged to share their lists of activated and inhibited genes in a simple file format. This allows easy application of models that may not be in an existing database for molecular signatures.\nThis study tests the accuracy of BD-Func on datasets that were used to define gene sets in\nMSigDB (Molecular Signatures DataBase, (Subramanian et al. 2005)), in comparison to GSEA (Gene Set Enrichment Analysis, (Mootha et al. 2003; Subramanian et al. 2005)) and IPA (Ingenuity Pathway Analysis, Ingenuity\u00ae Systems, www.ingenuity.com). IPA was selected for comparison because the upstream regulator function utilizes a similar principle as BD-Func (activation and inhibition is predicted by comparing the proportion of activated or inhibited targets, based upon annotations in the proprietary IPA database). GSEA was selected for comparison for two reasons: 1) GSEA was specifically designed to analyze MSigDB signatures, thus serving as a good baseline for positive control datasets and 2) MSigDB contains some signatures for both up- and down-regulated genes, so it is useful to compare separate analysis of these signatures (using GSEA) versus a direct comparison of up-regulated to down-regulated gene expression (using BD-Func). Different models for TGF\u03b2 and progesterone receptor (PGR) activity are also tested for robustness by application to other datasets. Finally, the utility of BD-Func to study custom gene signatures is tested with a novel signature associated with progesterone receptor status in breast cancer patients as well as a novel LBH589 signature that was defined using previously published cell line data and is validated using novel in vivo data presented in this study.",
+    "v2_Abstract": "BD-Func (BiDirectional FUNCtional Enrichment) is an algorithm that calculates functional enrichment by comparing lists of genes that are known to be activated versus inhibited in a pathway or by a regulatory molecule. This paper shows that BD-Func can correctly predict cell line alternations and patient characteristics with accuracy comparable to popular algorithms, with a run-time that is up to 20 times faster. BD-Func is currently the only such algorithm which can make functional predictions on a single sample (and compare scores for individual samples across multiple groups), and BD-Func can provide predictive statistics and receiver operating characteristic (ROC) plots to quantify the accuracy of the signature associated with a binary phenotypic variable. BD-Func facilitates collaboration and reproducibility by encouraging users to share novel molecular signatures in the BD-Func discussion group, which is where the novel progesterone receptor and LBH589 signatures from this paper can be found. The novel LBH589 signature presented in this paper also serves as a case study showing how users can define a custom signature using cell line data that accurately predicts activity in vivo. This software is available to download at https://sourceforge.net/projects/bdfunc/. 1 PeerJ reviewing PDF | (v2013:06:581:0:0:NEW 13 Jun 2013) R ev ie w in g M an us cr ip t",
+    "v2_col_introduction": "introduction  : Systems-level analysis of the combined expression pattern of multiple genes can sometimes\nbe more informative than the expression pattern of an individual gene, and there are a number of tools to calculate functional enrichment of differentially expressed genes (Ashburner et al. 2000; Berriz et al. 2009; Boorsma et al. 2005; Chang & Nevins 2006; Draghici et al. 2007; Edelman et al. 2006; Hanzelmann et al. 2013; Huang et al. 2008; Huang et al. 2009; Kanehisa & Goto 2000; Kim et al. 2007; Kim & Volsky 2005; Lamb et al. 2006; Lee et al. 2005; Liberzon et al. 2011; McCormack et al. 2013; Naeem et al. 2012; Nam & Kim 2008; Sartor et al. 2009; Sartor et al. 2010; Subramanian et al. 2005; Tarca et al. 2008; Vazquez et al. 2010; Williams 2012; Xu et al. 2008; Zeeberg et al. 2003; Zhang et al. 2005). However, many functional annotations merely list membership in a pathway or ontology without explicitly modelling genes that should show activation or inhibition. For example, consider the KEGG canonical Wnt signalling pathway (Figure 1) (Kanehisa & Goto 2000). This pathway captures both exhibits molecules that either activate or inhibit the pathway, resulting in different phenotypes (Dellinger et al. 2012; Logan & Nusse 2004). However, many functional enrichment tools would expect all the members of the pathway to behave similarly (Figure 1C), such that up-regulation of a mix of activators and inhibitors can receive a higher score than selected up-regulation of only activators within the pathway. This is a basic problem that BD-Func (BiDirectional FUNCtional Enrichment) is designed to overcome.\nMost functional enrichment tools either require upstream filtering of gene lists (Berriz et al.\n2009; Chang & Nevins 2006; Huang et al. 2008; Lamb et al. 2006; Vazquez et al. 2010; Zhang et al. 2005) or a comparison of signal intensities between two groups (Boorsma et al. 2005; Kim & Volsky 2005; Subramanian et al. 2005). However, BD-Func compares the relative expression levels between activated and inhibited genes, and we show that BD-Func can successfully analyze either fold-change values between populations or raw intensity / expression values (for both microarray and RNA-Seq data). In fact, the ability to analyze absolute expression values in a single sample is a unique feature only present in a limited number of functional genomic tools (Edelman et al. 2006; Williams 2012). This may be useful for personalized medicine research or for the development of novel molecular signatures where the user may need to quantify the utility of the signature as a classifier. Additionally, BD-Func users are encouraged to share their lists of activated and inhibited genes in a simple file\n2PeerJ reviewing PDF | (v2013:06:581:0:0:NEW 13 Jun 2013)\nR ev ie w in g M an\nus cr ip t\nformat. This allows easy application of models that may not be in an existing database for molecular signatures.\nThis study tests the accuracy of BD-Func on datasets that were used to define gene sets in\nMSigDB (Molecular Signatures DataBase, (Subramanian et al. 2005)), in comparison to GSEA (Gene Set Enrichment Analysis, (Mootha et al. 2003; Subramanian et al. 2005)) and IPA (Ingenuity Pathway Analysis, Ingenuity\u00ae Systems, www.ingenuity.com). Different models for TGF\u03b2 and progesterone receptor (PGR) activity are also tested for robustness by application to other datasets. Finally, the utility of BD-Func to study custom gene signatures is tested with a novel signature associated with progesterone receptor status in breast cancer patients as well as a novel LBH589 signature that was defined using previously published cell line data and is validated using novel in vivo data presented in this study.",
+    "v1_text": "results : acknowledgements : We would like to thank Christine Brown, Mike Barish, and Thanh Dellinger for discussions that led to the creation of this algorithm. We would like to thank Xiwei Wu, Zheng Liu, and two anonymous reviewers for discussions regarding the BD-Func algorithm. We would like to thank the City of Hope Functional Genomics Core for processing the microarray data. 268 269 270 271 272 273 274 275 276 277 PeerJ reviewing PDF | (v2013:06:581:1:0:NEW 12 Aug 2013) R ev ie w in g M an us cr ip t discussion : Comparison of BD-Func to GSEA indicated that BD-Func can provide similar oncogenic signature predictions with much shorter run time (Table S3) and a more direct comparison of genes that are expected to be up- or down-regulated by the oncogenic regulators. One limitation to BD-Func is that it can only conduct functional enrichment for regulators with genes that are both up- and down-regulated, so there are many gene lists in MSigDB that cannot be analyzed using BD-Func (which instead should be 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 PeerJ reviewing PDF | (v2013:06:581:1:0:NEW 12 Aug 2013) R ev ie w in g M an us cr ip t analyzed using a tool such as GSEA). It is also worth noting that BD-Func can work with a wide range of sizes of gene lists for activated and inhibited genes (Table S4), but we would recommend using at least a few dozen genes when defining custom signatures. Comparison of the BD-Func oncogenic signatures to the IPA upstream regulators also showed that both programs could provide similar performance, which is not surprising giving the design of that module in IPA. One benefit to utilizing IPA is that IPA has a curated database which lists with a wider variety of regulators than the MSigDB oncogenic signatures that can be analyzed in BD-Func. In contrast, one major benefit to using BD-Func is the greater theoretical range of applications. For example, BD-Func provides an enrichment file for Gene Ontology (GO (Ashburner et al. 2000)) categories, but IPA does utilize this same strategy of analyzing functional ontologies by comparing the expression of positively and negatively regulated genes. The Connectivity Map is a commonly used tool to study gene expression profiles for drugs and other chemical perturbations (Lamb et al. 2006). There are no LBH589 / panobinostat treatments in the Connectivity Map database (although this database can certainly provide other useful information), so BDFunc provides a unique opportunity to test for gene signatures that show a strong positive or negative correlation with novel drug treatments (such as LBH589). Additionally, BD-Func is compatible with any gene mapping (in this case, gene symbol), whereas the Connectivity Map requires users to define their signatures in terms of HG-U133A probes. For example, affected gene symbols had to be converted to HG-U133A probes for this analysis. We believe that being able to define signatures based upon gene symbol is a substantial practical benefit. BD-Func also calculates a test statistic to represent functional activation or inhibition for each individual sample in a dataset, and this study shows how this statistic can be directly used as a classifier that can be used to quantify the predictive power of a given functional model. More specifically, this study shows the utility of using BD-Func for applying two novel predictive models (for progesterone receptor status in patients and for LBH589 drug treatment). The LBH589 signature provided biological confirmation that the results from an in vitro model can indeed apply to validation experiments in vivo. This is important because our hope is that the streamlined analysis, simple input file design, and BD-Func discussion board can be used to help scientists quickly and easily share novel predictive models. In short, BD-Func 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 PeerJ reviewing PDF | (v2013:06:581:1:0:NEW 12 Aug 2013) R ev ie w in g M an us cr ip t provides a novel framework for functional enrichment (by comparing the relative expression of activated versus inhibited genes) that is freely available with a user interface that is accessible to biologists without any coding experience. The results of this paper show that BD-Func provides accurate predictions matched by other popular tools, which make it a useful complement to standard analysis using tools like GSEA or IPA. in this table, bd-func is used to analyze fold-change values between pr+ and pr- patients. msigdb : = CLAUS_PGR_POSITIVE_MENINGIOMA signatures. COH = novel PGR signature developed in this study. PeerJ reviewing PDF | (v2013:06:581:1:0:NEW 12 Aug 2013) R ev ie w in g M an us cr ip t Table 3: Prediction of Progesterone Receptor Status in Patient Samples Cohort BD-Func (MSigDB) BD-Func (COH) GSEA (MSigDB) IPA GSE9438 (N=31) Yes Yes No (UP) No (DN) No Huang et al. 2003 (N=88) No Yes No (UP) No (DN) No Chin et al. 2006 (N=117) No Yes No (UP) No (DN) No Anders et al. 2008 (N=73) No Yes No (UP) No (DN) No Finak et al. 2008 (N=53) No Yes No (UP) No (DN) No expO (N=256) No Yes No (UP) No (DN) No TCGA (N=739) No Yes No (UP) No (DN) No 1 2 3 4 PeerJ reviewing PDF | (v2013:06:581:1:0:NEW 12 Aug 2013) R ev ie w in g M an us cr ip t material and methods : BD-Func Algorithm 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 PeerJ reviewing PDF | (v2013:06:581:1:0:NEW 12 Aug 2013) R ev ie w in g M an us cr ip t The basic principle behind BD-Func is to treat activated and inhibited genes as replicate observations in two populations. BD-Func is agnostic towards the type of signal used for analysis. For example, this paper uses BD-Func to study both fold-change values between two populations as well as raw signal intensities for a single column of signal values. In the paper, the t-test statistic is used to compare the expression patterns of activated and inhibited genes. However, BD-Func also allows users to compare the activated and inhibited distributions using a Mann-Whitney U test or Kolmogorov-Smirnov (K-S) test. BD-Func users have the option to calculate False Discovery Rates (FDR) using the method of Benjamini and Hochberg (Benjamini & Hochberg 1995) or the Storey q-value (Storey & Tibshirani 2003). BD-Func comes with four enrichment files: c2, c5, and c6 from MSigDB (Liberzon et al. 2011) and a list of functions defined directly from the human Gene Ontology (GO) annotations (Ashburner et al. 2000). All of these lists were created by searching for functions with both \u201cup\u201d and \u201cdown\u201d (or \u201cpositive regulation\u201d and \u201cnegative regulation\u201d) gene lists. For the human GO file, a functional annotation needed to contain at least 10 positively regulated genes and 10 negatively regulated genes. We also encourage users to share their own custom models on the BD-Func discussion group: http://sourceforge.net/p/bdfunc/discussion/ There are three different input files that can be analyzed using BD-Func (Figure S1): 1-D Input File: If the user supplies expression values for a single column of data, a density plot is created for the signal for the activated and inhibited genes. In this case, the BD-Func algorithm works exactly as described above. 2-D Input File: If the input file contains multiple columns of data, BD-Func is first run separately for each sample (represented by a column in the data matrix), as described above. Next, box-plots are created for test-statistic scores for each group (labelled in the header of the input file; Figure S2). Finally, an ANOVA p-value is provided to compare the test statistics between groups. Theoretically, test statistics could be used to make functional predictions for each sample in isolation. There are some examples in this paper where this strategy works OK. However, comparing test statistics across all samples within different groups is the only strategy that consistently works well for all the analysis presented in this paper. 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 PeerJ reviewing PDF | (v2013:06:581:1:0:NEW 12 Aug 2013) R ev ie w in g M an us cr ip t 2-D Input File for Classifier: If the input file contains multiple columns of data and two groups (called \u201cpositive\u201d and \u201cnegative\u201d), BD-Func will create a receiver operating characteristic (ROC) plot using the test-statistic as the classification score using the ROCR package (Sing et al. 2005). This is in addition to all the calculations and output files for a normal 2-D input file containing multiple columns for BD-Func to analyze (in this paper, this represents normalized signal intensity values). Sample Acquisition and Processing Microarray datasets were downloaded from GEO (Edgar et al. 2002) or Array-Express (Parkinson et al. 2009). When raw .CEL files were available, samples were RMA normalized (Irizarry et al. 2003). Otherwise, processed intensity values were used for microarray analysis. Fold-Change values for all of the cell line datasets (TGF\u03b2 (Padua et al. 2008; Qin et al. 2009; Renzoni et al. 2004; Sartor et al. 2010; Scandura et al. 2004), mTOR (Wei et al. 2006), p53 (Elkon et al. 2005), and BRCA1(Furuta et al. 2006)) and the MSigDB progesterone receptor dataset (Claus et al. 2008) were calculated using the method of the least-squares mean using Partek\u00ae Genomics SuiteTM (Partek Inc. 2012). All other clinical samples (Anders et al. 2008; Bild et al. 2006; Chin et al. 2006; Finak et al. 2008; Huang et al. 2003; Ivshina et al. 2006; Sotiriou et al. 2003; The Cancer Genome Atlas Network 2012) were downloaded and analyzed for differential expression using BRAVO (http://bravo.coh.org/) (Deng et al. 2013, unpublished data). The novel progesterone receptor gene signature presented in this paper was produced by identifying genes in the expO dataset (GEO Series GSE2109) with a |fold-change| > 2 and an False Discovery Rate (FDR) < 0.05 (where the FDR is calculated using the method of Benjamini and Hochberg (Benjamini & Hochberg 1995) to analyze the distribution of t-test p-values). This is how the samples in this particular paper were processed, but users are not required to use this particular set of tools for preparing BD-Func input files and/or creating gene lists for custom signatures. Reads Per Kilobase per Million mapped reads (RPKM (Mortazavi et al. 2008)) values for RNA- Seq data was downloaded from the TCGA web portal (The Cancer Genome Atlas Network 2012). RPKM values were transformed by addition of 0.1 (to avoid large fold-change values for low coverage reads) followed by a log2 transformation (to normalize the signal distribution). GSEA Comparison 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 PeerJ reviewing PDF | (v2013:06:581:1:0:NEW 12 Aug 2013) R ev ie w in g M an us cr ip t With the exception of the progesterone receptor signature (which utilized the CLAUS_PGR_POSITIVE_MENINGIOMA (Claus et al. 2008) signatures from MSigDB-c2, version 3.1), oncogenic signatures were defined using the following gene lists from MSigDB-c6 ((Liberzon et al. 2011), version 3.1) for GSEA analysis: TGFB_UP.V1 (Padua et al. 2008), MTOR_UP.N4.V1 (Wei et al. 2006), P53_DN.V2 (Elkon et al. 2005), and BRCA_DN.V1 (Furuta et al. 2006). GSEA ((Subramanian et al. 2005) ,version 2.0) calculated p-values by permutation over phenotypes whenever possible (Anders et al. 2008; Chin et al. 2006; Claus et al. 2008; Elkon et al. 2005; Finak et al. 2008; Huang et al. 2003; Padua et al. 2008; Sartor et al. 2010; Scandura et al. 2004; The Cancer Genome Atlas Network 2012; Wei et al. 2006), although there were a few datasets with less than 3 replicates for which gene sets had to be permuted instead of phenotypes (Furuta et al. 2006; Qin et al. 2009; Renzoni et al. 2004). For recovery of known perturbations of oncogenic regulators, GSEA results must either show a FWER p-value < 0.25 or a NOM p-value < 0.05, which are the default cut-offs IPA Comparison Ingenuity Pathway Analysis (IPA, Ingenuity\u00ae Systems, www.ingenuity.com) contains an \u201cUpstream Regulator\u201d module that compares the enrichment of activated and inhibited genes among up- and downregulated genes. So, the underlying principle is similar to BD-Func except it utilizes Ingenuity\u2019s propriety database of regulatory interactions and uses a z-score to calculate statistical significance between activated and inhibited genes. Gene lists in IPA were filtered for those genes showing |fold-change| > 1.5 while the entire gene list is used to define background enrichment. For recovery of known perturbations of oncogenic regulators, the upstream regulator must be identified as \u201cactivated\u201d or \u201cinhibited\u201d (|z-score| > 2), which are the default cut-offs LBH589 Signature Activated and inhibited genes were defined using overlapping gene lists defined from 3 cell line treatments that have been previously published (Kubo et al. 2013). That same study showed that LBH589 treatment significantly decreased tumor volume in exemestane (EXE) resistant MCF-7aro xenografts in mice. This study analyzes novel microarray data from EXE-resistant tumors treated with (EXE + LBH589) or without (EXE only) treatment of LBH589. All animal research procedures were approved by the City of 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 PeerJ reviewing PDF | (v2013:06:581:1:0:NEW 12 Aug 2013) R ev ie w in g M an us cr ip t Hope Institutional Animal Care and Use Committee. This novel microarray data is available in GEO series GSE47346. In order to be included in the novel BD-Func signature, genes must show differential expression in all 3 cell lines. Genes were defined as differentially expressed if they showed a |fold-change| > 1.5 and pvalue < 0.05 , and the BD-Func signature genes had to meet these conditions for each of the 3 LBH589 cell line treatments (with consistent direction of fold-change). P-values were calculated via 1-way ANOVA with appropriate linear contrast was used to compare data sets using Partek\u00ae Genomics SuiteTM (Partek, Inc., St. Louis, MO). Fold-change values were calculated based upon the least-squares mean, and data was normalized using robust multichip average (RMA) normalization (Irizarry et al. 2003). a bd-func density plot for fold-change values for activated and inhibited genes for exe + lbh589 : vs. Exe alone tumors. At a population level, the Exe + LBH589 tumors show higher expression of activated genes whereas the Exe tumors show increased expression of inhibited genes (p= 2.0 x 10-15). 1 qualitatively detected at 2 hours but not 4 hours, but activity is not significant with p-value < 0.05 for : either time-point PeerJ reviewing PDF | (v2013:06:581:1:0:NEW 12 Aug 2013) R ev ie w in g M an us cr ip t Table 2: Prediction of TGF\u03b2 Activity in Novel Datasets BD-Func (Fold-Change) BD-Func (Intensity) GSEA IPA GSE1724 Yes Yes Yes (UP) No (DN) Yes GSE1805 No No1 No (UP) Yes (DN) No GSE6653 Yes Yes Yes (UP) Yes (DN) Yes GSE17708 Yes Yes Yes (UP) No (DN) Yes PeerJ reviewing PDF | (v2013:06:581:1:0:NEW 12 Aug 2013) R ev ie w in g M an us cr ip t Table 3(on next page) b bd-func box-plot for single-sample signature scores. exe alone shows the greatest inhibition of : LBH-related gene expression whereas EXE + LBH shows less inhibition of LBH-related gene expression, as PeerJ reviewing PDF | (v2013:06:581:1:0:NEW 12 Aug 2013) R ev ie w in g M an us cr ip t Table 1(on next page) bd-func shows equal or greater performance to gsea and ipa for functional enrichment : Given the relative ease by which samples can be classified as having positive or negative activity for an individual biomarker, the accuracy of BD-Func was first tested by applying several MSigDB oncogenic signatures (Liberzon et al. 2011) to the datasets from which the signatures were defined (Claus et al. 2008; Elkon et al. 2005; Furuta et al. 2006; Padua et al. 2008; Wei et al. 2006). BD-Func was able to detect the activation or inhibition of all of these oncogenic signatures (Table 1, Figure 2). GSEA could detect all of the signatures except the Claus et al. (Claus et al. 2008) progesterone receptor signature. Among these 5 test datasets, IPA could only detect the activity of 2 of these genes; however, this is not a completely fair comparison because we would expect some over fitting of the MSigDB signatures for the GSEA and BD-Func analysis. Nevertheless, the significance of this analysis is that BD-Func can accurately detect perturbation of all of these biomarkers on datasets where we know that these specific genes will be altered. In order to test the performance of BD-Func, GSEA, and IPA on novel datasets (which were not used by MSigDB to define gene lists), we applied the 3 algorithms to four datasets with TGF\u03b2 treatments (Table 2). All 3 algorithms showed roughly equal performance for predicting TGF\u03b2 treatment in the appropriate samples. Overall, analyses of these nine datasets indicate that BD-Func can provide similar quality results as GSEA and IPA. 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 PeerJ reviewing PDF | (v2013:06:581:1:0:NEW 12 Aug 2013) R ev ie w in g M an us cr ip t Six breast cancer datasets were also used to test the robustness of the progesterone receptor (PGR) signature ((Anders et al. 2008; Chin et al. 2006; Finak et al. 2008; Huang et al. 2003; The Cancer Genome Atlas Network 2012)). Unfortunately, neither BD-Func, GSEA, nor IPA could predict progesterone status in all seven patient populations (Table 3). To be fair, the original Claus et al. (Claus et al. 2008) dataset was used to define progesterone receptor status in meningioma samples whereas the novel datasets tested were all breast cancer samples (where testing for over-expression of progesterone receptor is common (Bardou et al. 2003)). Nevertheless, BD-Func is designed to be able to utilize custom gene signatures with activated and inhibited, so we defined a novel progesterone receptor signature using the expO dataset (GEO Series GSE2109). This signature can identify progesterone receptor positive and negative patients for all 7 cohorts (1 meningioma and 6 breast cancer datasets), so it is robust enough for application to multiple cancer types. Another unique feature of BD-Func is the ability to use the activation versus inhibition test statistic as a classifier to define a predictive model. If a t-test statistic of 2 is used for the cut-off of distinguishing the positive and negative populations (roughly corresponding to a p-value < 0.05), it is clear that the MSigDB signature is extremely accurate at predicting PGR status in the original dataset but not in the breast cancer datasets (Table S1). Likewise, the TGF\u03b2 signature could differentiate between the treated and untreated groups if the test statistic of 2 was used as the threshold to distinguish the groups (Figure S2). However, this threshold does not work well in all circumstances: unlike the analysis of fold-change values, the p-value (for any statistical method) is not always the ideal statistic for assessment of functional enrichment on intensity values. For example, the mTOR and BRCA1 signatures (Figure 2B) show appropriate changes in test statistics that clearly distinguish treated and untreated groups, but activation and inhibition can\u2019t be defined based upon a pre-defined cut-off for the test statistic value (e.g. 2 or -2). For this reason, we provide an ANOVA p-value to quantify the difference in test-statistic between groups, where the test statistic serves as a score for a second calculation of statistical significance. Additionally, we believe that predictive statistics are a useful method for accessing BD-Func scores for individual samples within large patient populations. In order to quantify the accuracy of the model without depending on a pre-defined cut-off, BD-Func produces receiver operating characteristic 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 PeerJ reviewing PDF | (v2013:06:581:1:0:NEW 12 Aug 2013) R ev ie w in g M an us cr ip t (ROC) plots for each cohort and the area under the curve (AUC) is calculated for each ROC plot (where a 100% accurate model would have an AUC=1.00) (Figure S3, Table S1). The superior performance of the novel PGR signature on the breast cancer cohorts becomes even clearer when these predictive statistics are compared for the two models (Figure 3, Figure S4, and Table S2). Importantly, the novel signature showed the same level of accuracy on the TCGA breast cancer dataset as the original expO dataset. This is significant because the TCGA dataset is over twice as large as the expO dataset, and the TCGA dataset utilizes RNA-Seq while the expO dataset utilizes microarrays to quantify gene expression. In other words, this shows that BD-Func is capable of producing very robust predictions that translate across different genomic technologies. LBH589 (panobinostat) is a histone deacetylase inhibitor that has been previously shown to suppress the proliferation of aromatase inhibitor resistant breast cancer cells, which was a conclusion supported in part by functional enrichment analysis of commonly affected genes from 3 cell line experiments (Kubo et al. 2013). Gene lists derived from these cell line experiments can be easily used to define a BD-func signature, so we hypothesized that the results from this previous in vitro cell culture study could be used to predict LBH589 activity in vivo in an animal study. Specifically, we asked if the signature defined based upon LBH589 treatment in 3 cell lines (H295R, MCF-7her2, HeLa) could detect LBH589 activity in a mouse xenograft from a different cell line (MCF-7aro xenograft treated with EXE). Indeed, BD-Func correctly used the cell line LBH589 signature to identify common gene expression changes in the tumours treated with LBH589 and EXE compared to the mice that were only exposed to EXE treatment (Figure 4). This figure shows the output figures from BD-Func for the 5 MSigDB signatures tested on their original datasets. A. Density plots for fold-change values for activated genes (colored red) and inhibited genes (colored green). These plots are used to illustrate BD-Func analysis on a single column of data (in this case, fold-change values between the positive and negative populations). B. Box-plots of activation versus inhibition test statistics for all relevant samples in each of the 5 MSigDB datasets. Note that each box-plot describes the distribution of test statistics for each group \u2013 it does not represent the expression of an individual gene or a metagene. In each of these five examples, the test statistic shows very clear differences among the different groups. If the median t-test statistic is greater than 2, the box is colored red. If the median t-test statistic is less than -2, the box is colored green. PeerJ reviewing PDF | (v2013:06:581:1:0:NEW 12 Aug 2013) R ev ie w in g M an us cr ip t Figure 3 many pathways are characterized by a mix of activation and inhibition : This figure shows the initial signalling steps in the Wnt signalling pathway, as defined by KEGG ( Kanehisa & Goto 2000 ) . Up-regulated genes are shown in red and down-regulated genes are shown in green. A. Complete Activation All genes activating the Wnt signalling pathway are up-regulated and all inhibitors are down-regulated. B. Complete Inhibition All genes activating the Wnt signalling pathway are down-regulated and all inhibitors are up-regulated. C. Mixed Pattern All genes in the figure are up-regulated. It is unclear what the downstream expression levels should be, but one may hypothesize a mixed result from Figure 1A and Figure 1B. However, most functional enrichment tools would predict this as the pattern with the strongest up-regulation. Underneath each diagram is the expected signal distribution that would be produced by BD-Func. PeerJ reviewing PDF | (v2013:06:581:1:0:NEW 12 Aug 2013) R ev ie w in g M an us cr ip t Figure 2 custom progesterone signature outperforms msigdb signature on breast cancer patients : AUC (Area Under the Curve) values from the ROC plots for each patient cohort are displayed. Although the MSigBD gene set shows extremely high accuracy for the original meningioma dataset, it shows essentially random predictive power for the 6 larger breast cancer datasets. On the other hand, a custom progesterone receptor signature defined using the expO dataset shows high accuracy for all 7 cohorts, and the high accuracy is maintained even in the largest cohort (TCGA). PeerJ reviewing PDF | (v2013:06:581:1:0:NEW 12 Aug 2013) R ev ie w in g M an us cr ip t Figure 4 Novel Cell Line LBH589 Signature Can Accurately Detect Drug Activity In Vivo recoveryof known perturbations for selected msigdb oncogenic genes : BD-Func \u201cFold-Change\u201d corresponds to analysis of fold-change values calculated between the perturbed and unperturbed groups. BD-Func \u201cIntensity\u201d corresponds to calculation of activity vs. inhibition score for each sample in the dataset followed by a comparison in the distribution of test statistics for all of the samples. 1p53 signal changes with sign matching p53 expression (in this case P53_DN indicates genes down-regulated by knock-down of p53, not genes negatively related by p53) PeerJ reviewing PDF | (v2013:06:581:1:0:NEW 12 Aug 2013) R ev ie w in g M an us cr ip t Table 1: Recovery of Known Perturbations for Selected MSigDB Oncogenic Genes BD-Func (Fold-Change) BD-Func (Intensity) GSEA IPA TGF\u03b2 (E-TABM-420) Yes Yes Yes (UP) Yes (DN) Yes mTOR \u2013 N4 (GSE5824) Yes Yes Yes (UP) Yes (DN) No P53 \u2013 V21 (GSE1676 ) Yes Yes Yes (UP) No (DN) Yes BRCA1 (GSE4754) Yes Yes Yes (UP) Yes (DN) No PGR (GSE9438) Yes Yes No (UP) No (DN) No 1 2 PeerJ reviewing PDF | (v2013:06:581:1:0:NEW 12 Aug 2013) R ev ie w in g M an us cr ip t Table 2(on next page) prediction of tgf\u03b2 activity in novel datasets :",
+    "v2_text": "results : b bd-func box-plot for single-sample signature scores. exe alone shows the greatest inhibition of : LBH-related gene expression whereas EXE + LBH shows less inhibition of LBH-related gene expression, as PeerJ reviewing PDF | (v2013:06:581:0:0:NEW 13 Jun 2013) R ev ie w in g M an us cr ip t Table 1(on next page) acknowledgements : We would like to thank Christine Brown, Mike Barish, and Thanh Dellinger for discussions that led to the creation of this algorithm. We would like to thank Xiwei Wu and Zheng Liu for discussions regarding the BD-Func algorithm. We would like to thank the City of Hope Functional Genomics Core for processing the microarray data. 10PeerJ reviewing PDF | (v2013:06:581:0:0:NEW 13 Jun 2013) R ev ie w in g M an us cr ip t discussion : Comparison of BD-Func to GSEA indicated that BD-Func can provide similar oncogenic signature predictions with much shorter run time (Table S3) and a more direct comparison of genes that are expected to be up- or down-regulated by the oncogenic regulators. One limitation to BD-Func is that it can only conduct functional enrichment for regulators with genes that are both up- and down-regulated, so there are many gene lists in MSigDB that cannot be analyzed using BD-Func. It is also worth noting that BD-Func can work with a wide range of sizes of gene lists for activated and inhibited genes (Table S4), but we would recommend using at least a few dozen genes when defining custom signatures. Comparison of the BD-Func oncogenic signatures to the IPA upstream regulators also showed that both programs could provide similar performance, which is not surprising giving the design of that module in IPA. One benefit to utilizing IPA is that IPA has a curated database which lists with a wider variety of regulators than the MSigDB oncogenic signatures that can be analyzed in 9PeerJ reviewing PDF | (v2013:06:581:0:0:NEW 13 Jun 2013) R ev ie w in g M an us cr ip t BD-Func. In contrast, one major benefit to using BD-Func is the greater theoretical range of applications. For example, BD-Func provides an enrichment file for Gene Ontology (GO (Ashburner et al. 2000)) categories, but IPA does utilize this same strategy of analyzing functional ontologies by comparing the expression of positively and negatively regulated genes. BD-Func also calculates a test statistic to represent functional activation or inhibition for each individual sample in a dataset, and this study shows how this statistic can be directly used as a classifier that can be used to quantify the predictive power of a given functional model. More specifically, this study shows the utility of using BD-Func for applying two novel predictive models (for progesterone receptor status in patients and for LBH589 drug treatment). The LBH589 signature provided biological confirmation that the results from an in vitro model can indeed apply to validation experiments in vivo. This is important because our hope is that the streamlined analysis, simple input file design, and BD-Func discussion board can be used to help scientists quickly and easily share novel predictive models. in this table, bd-func is used to analyze fold-change values between pr+ and pr- patients. msigdb : = CLAUS_PGR_POSITIVE_MENINGIOMA signatures. COH = novel PGR signature developed in this study. PeerJ reviewing PDF | (v2013:06:581:0:0:NEW 13 Jun 2013) R ev ie w in g M an us cr ip t Table 3: Prediction of Progesterone Receptor Status in Patient Samples Cohort BD-Func (MSigDB ) BD-Func (COH) GSEA (MSigDB) IPA GSE9438 (N=31) Yes Yes No (UP) No (DN) No Huang et al. 2003 (N=88) No Yes No (UP) No (DN) No Chin et al. 2006 (N=117) No Yes No (UP) No (DN) No Anders et al. 2008 (N=73) No Yes No (UP) No (DN) No Finak et al. 2008 (N=53) No Yes No (UP) No (DN) No expO (N=256) No Yes No (UP) No (DN) No TCGA (N=739) No Yes No (UP) No (DN) No PeerJ reviewing PDF | (v2013:06:581:0:0:NEW 13 Jun 2013) R ev ie w in g M an us cr ip t a bd-func density plot for fold-change values for activated and inhibited genes for exe + lbh589 : vs. Exe alone tumors. At a population level, the Exe + LBH589 tumors show higher expression of activated genes whereas the Exe tumors show increased expression of inhibited genes (p= 2.0 x 10-15). material and methods : BD-Func Algorithm The basic principle behind BD-Func is to treat activated and inhibited genes as replicate observations in two populations. BD-Func is agnostic towards the type of signal used for analysis. For example, this paper uses BD-Func to study both fold-change values between two populations as well as raw signal intensities for an individual sample. In the paper, the t-test statistic is used to compare the expression patterns of activated and inhibited genes. However, BD-Func also allows users to compare the activated and inhibited distributions using a Mann-Whitney U test or Kolmogorov-Smirnov (K-S) test. BD-Func users have the option to calculate False Discovery Rates (FDR) using the method of Benjamini and Hochberg (Benjamini & Hochberg 1995) or the Storey q-value (Storey & Tibshirani 2003). BD-Func comes with four enrichment files: c2, c5, and c6 from MSigDB (Liberzon et al. 2011) and a list of functions defined directly from the human Gene Ontology (GO) annotations (Ashburner et al. 2000). All of these lists were created by searching for functions with both \u201cup\u201d and \u201cdown\u201d (or \u201cpositive regulation\u201d and \u201cnegative regulation\u201d) gene lists. For the human GO file, a functional annotation needed to contain at least 10 positively regulated genes and 10 negatively regulated 3PeerJ reviewing PDF | (v2013:06:581:0:0:NEW 13 Jun 2013) R ev ie w in g M an us cr ip t genes. We also encourage users to share their own custom models on the BD-Func discussion group: http://sourceforge.net/p/bdfunc/discussion/ There are three different output figures that can be created by BD-Func. If the user supplies expression values for a single column of data, a density plot is created for the signal for the activated and inhibited genes. If the input file contains multiple columns of data, box-plots are created for test-statistic scores for each group (labelled in the header of the input file; Figure S1) and an ANOVA p-value is provided to compare the test statistics between groups. If the input file contains multiple columns of data and two groups (called \u201cpositive\u201d and \u201cnegative\u201d), BD-Func will create a receiver operating characteristic (ROC) plot using the test-statistic as the classification score using the ROCR package (Sing et al. 2005). Sample Acquisition and Processing Microarray datasets were downloaded from GEO (Edgar et al. 2002) or Array-Express (Parkinson et al. 2009). When raw .CEL files were available, samples were RMA normalized (Irizarry et al. 2003). Otherwise, processed intensity values were used for microarray analysis. Fold-Change values for all of the cell line datasets (Elkon et al. 2005; Furuta et al. 2006; Padua et al. 2008; Qin et al. 2009; Renzoni et al. 2004; Sartor et al. 2010; Scandura et al. 2004; Wei et al. 2006) and the MSigDB progesterone receptor dataset (Claus et al. 2008) were calculated using the method of the least-squares mean using Partek\u00ae Genomics SuiteTM (Partek Inc. 2012). All other clinical samples (Anders et al. 2008; Bild et al. 2006; Chin et al. 2006; Finak et al. 2008; Huang et al. 2003; Ivshina et al. 2006; Sotiriou et al. 2003; The Cancer Genome Atlas Network 2012) were downloaded and analyzed for differential expression using BRAVO (http://bravo.coh.org/) (Deng et al.). The novel progesterone receptor gene signature presented in this paper was produced by identifying genes in the expO dataset (GEO Series GSE2109) with a |fold-change| > 2 and an False Discovery Rate (FDR) < 0.05 (where the FDR is calculated using the method of Benjamini and Hochberg (Benjamini & Hochberg 1995) to analyze the distribution of t-test p-values), again calculated within BRAVO (Deng et al.). Reads Per Kilobase per Million mapped reads (RPKM (Mortazavi et al. 2008)) values for RNA-Seq data was downloaded from the TCGA web portal (The Cancer Genome Atlas Network 4PeerJ reviewing PDF | (v2013:06:581:0:0:NEW 13 Jun 2013) R ev ie w in g M an us cr ip t 2012). RPKM values were transformed by addition of 0.1 (to avoid large fold-change values for low coverage reads) followed by a log2 transformation (to normalize the signal distribution). GSEA Parameters With the exception of the progesterone receptor signature (which utilized the CLAUS_PGR_POSITIVE_MENINGIOMA (Claus et al. 2008) signatures from MSigDB-c2, version 3.1), oncogenic signatures were defined using the following gene lists from MSigDB-c6 ((Liberzon et al. 2011), version 3.1) for GSEA analysis: TGFB_UP.V1 (Padua et al. 2008), MTOR_UP.N4.V1 (Wei et al. 2006), P53_DN.V2 (Elkon et al. 2005), and BRCA_DN.V1 (Furuta et al. 2006). GSEA ((Subramanian et al. 2005) ,version 2.0) calculated p-values by permutation over phenotypes whenever possible (Anders et al. 2008; Chin et al. 2006; Claus et al. 2008; Elkon et al. 2005; Finak et al. 2008; Huang et al. 2003; Padua et al. 2008; Sartor et al. 2010; Scandura et al. 2004; The Cancer Genome Atlas Network 2012; Wei et al. 2006), although there were a few datasets with less than 3 replicates for which gene sets had to be permuted instead of phenotypes (Furuta et al. 2006; Qin et al. 2009; Renzoni et al. 2004). For recovery of known perturbations of oncogenic regulators, GSEA results must either show a FWER p-value < 0.25 or a NOM p-value < 0.05. IPA Parameters Ingenuity Pathway Analysis (IPA, Ingenuity\u00ae Systems, www.ingenuity.com) contains an \u201cUpstream Regulator\u201d module that compares the enrichment of activated and inhibited genes among up- and down-regulated genes. So, the underlying principal is similar to BD-Func except it utilizes Ingenuity\u2019s propriety database of regulatory interactions and uses a z-score to calculate statistical significance between activated and inhibited genes. Gene lists in IPA were filtered for those genes showing |fold-change| > 1.5 while the entire gene list is used to define background enrichment. For recovery of known perturbations of oncogenic regulators, the upstream regulator must be identified as \u201cactivated\u201d or \u201cinhibited\u201d (|z-score| > 2). LBH589 Signature Activated and inhibited genes were defined using overlapping gene lists defined from 3 cell line treatments that have been previously published (Kubo et al. 2013). That same study showed that LBH589 treatment significantly decreased tumor volume in exemestane (EXE) resistant MCF-7aro 5PeerJ reviewing PDF | (v2013:06:581:0:0:NEW 13 Jun 2013) R ev ie w in g M an us cr ip t xenografts in mice. This study analyzes novel microarray data from EXE-resistant tumors treated with (EXE + LBH589) or without (EXE only) treatment of LBH589. All animal research procedures were approved by the City of Hope Institutional Animal Care and Use Committee. This novel microarray data is available in GEO series GSE47346. In order to be included in the novel BD-Func signature, genes must show differential expression in all 3 cell lines. Genes were defined as differentially expressed if they showed a |fold-change| > 1.5 and p-value < 0.05 , and the BD-Func signature genes had to meet these conditions for each of the 3 LBH589 cell line treatments (with consistent direction of fold-change). P-values were calculated via 1-way ANOVA with appropriate linear contrast was used to compare data sets using Partek\u00ae Genomics SuiteTM (Partek, Inc., St. Louis, MO). Fold-change values were calculated based upon the least-squares mean, and data was normalized using robust multichip average (RMA) normalization (Irizarry et al. 2003). For Connectivity Map analysis, genes were defined as differentially expressed if they showed |fold-change| > 1.5 and p-value < 0.05 (the same as for an individual cell line treatment). NetAffx was used to map gene symbols to HG-U133A probes (Liu et al. 2003). 1 qualitatively detected at 2 hours but not 4 hours, but activity is not significant with p-value < 0.05 for : either time-point PeerJ reviewing PDF | (v2013:06:581:0:0:NEW 13 Jun 2013) R ev ie w in g M an us cr ip t Table 2: Prediction of TGF\u03b2 Activity in Novel Datasets BD-Func (Fold-Change) BD-Func (Intensity) GSEA IPA GSE1724 Yes Yes Yes (UP) No (DN) Yes GSE1805 No No1 No (UP) Yes (DN) No GSE6653 Yes Yes Yes (UP) Yes (DN) Yes GSE17708 Yes Yes Yes (UP) No (DN) Yes PeerJ reviewing PDF | (v2013:06:581:0:0:NEW 13 Jun 2013) R ev ie w in g M an us cr ip t Table 3(on next page) recovery of known perturbations for selected msigdb oncogenic genes : BD-Func \u201cFold-Change\u201d corresponds to analysis of fold-change values calculated between the perturbed and unperturbed groups. BD-Func \u201cIntensity\u201d corresponds to calculation of activity vs. inhibition score for each sample in the dataset followed by a comparison in the distribution of test statistics for all of the samples. 1p53 signal changes with sign matching p53 expression (in this case P53_DN indicates genes down-regulated by knock-down of p53, not genes negatively related by p53) PeerJ reviewing PDF | (v2013:06:581:0:0:NEW 13 Jun 2013) R ev ie w in g M an us cr ip t Table 1: Recovery of Known Perturbations for Selected MSigDB Oncogenic Genes BD-Func (Fold-Chang e) BD-Func (Intensity) GSEA IPA TGF\u03b2 (E-TABM-420) Yes Yes Yes (UP) Yes (DN) Yes mTOR \u2013 N4 (GSE5824) Yes Yes Yes (UP) Yes (DN) No P53 \u2013 V21 (GSE1676 ) Yes Yes Yes (UP) No (DN) Yes BRCA1 (GSE4754) Yes Yes Yes (UP) Yes (DN) No PGR (GSE9438) Yes Yes No (UP) No (DN) No PeerJ reviewing PDF | (v2013:06:581:0:0:NEW 13 Jun 2013) R ev ie w in g M an us cr ip t Table 2(on next page) bd-func shows equal or greater performance to gsea and ipa for functional enrichment : Given the relative ease by which samples can be classified as having positive or negative activity for an individual biomarker, the accuracy of BD-Func was first tested by applying several MSigDB oncogenic signatures (Liberzon et al. 2011) to the datasets from which the signatures were defined (Elkon et al. 2005; Furuta et al. 2006; Padua et al. 2008; Renzoni et al. 2004; Wei et al. 2006). BD-Func was able to detect the activation or inhibition of all of these oncogenic signatures (Table 1, Figure 2). GSEA could detect all of the signatures except the Claus et al. (Elkon et al. 2005; Furuta et al. 2006; Renzoni et al. 2004; Wei et al. 2006) progesterone receptor signature. Among these 5 test datasets, IPA could only detect the activity of 2 of these genes; however, this is not a completely fair comparison because we would expect some over fitting of the MSigDB signatures for the GSEA and BD-Func analysis. Nevertheless, the significance of this analysis is that BD-Func can detect 6PeerJ reviewing PDF | (v2013:06:581:0:0:NEW 13 Jun 2013) R ev ie w in g M an us cr ip t accurately detect perturbation of all of these biomarkers on datasets where we know that these specific genes will be altered. In order to test the performance of BD-Func, GSEA, and IPA on novel datasets (which were not used by MSigDB to define gene lists), we applied the 3 algorithms to four datasets with TGF\u03b2 treatments (Table 2). All 3 algorithms showed roughly equal performance for predicting TGF\u03b2 treatment in the appropriate samples. Overall, analyses of these nine datasets indicate that BD-Func can provide similar quality results as GSEA and IPA. Six breast cancer datasets were also used to test the robustness of the progesterone receptor (PGR) signature ((Anders et al. 2008; Chin et al. 2006; Finak et al. 2008; Huang et al. 2003; The Cancer Genome Atlas Network 2012)). Unfortunately, neither BD-Func, GSEA, nor IPA could predict progesterone status in these two patient populations (Table 3). To be fair, the original Claus et al. (Claus et al. 2008) dataset was used to define progesterone receptor status in meningioma samples whereas the novel datasets tested were all breast cancer samples (where testing for over-expression of progesterone receptor is common (Bardou et al. 2003)). Nevertheless, BD-Func is designed to be able to utilize custom gene signatures with activated and inhibited, so we defined a novel progesterone receptor signature using the expO dataset (GEO Series GSE2109). This signature can identify progesterone receptor positive and negative patients for all 7 cohorts (1 meningioma and 6 breast cancer datasets), so it is robust enough for application to multiple cancer types. Another unique feature of BD-Func is the ability to use the activation versus inhibition test statistic as a classifier to define a predictive model. If a t-test statistic of 2 is used for the cut-off of distinguishing the positive and negative populations (roughly corresponding to a p-value < 0.05), it is clear that the MSigDB signature is extremely accurate at predicting PGR status in the original dataset but not in the breast cancer datasets (Table S1). Likewise, the TGF\u03b2 signature could differentiate between the treated and untreated groups if the test statistic of 2 was used as the threshold to distinguish the groups (Figure S1). However, this threshold does not work well in all circumstances: unlike the analysis of fold-change values, the p-value (for any statistical method) is not always the ideal statistic for assessment of functional enrichment on intensity values. For example, the mTOR and BRCA1 signatures (Figure 2B) show appropriate changes in test statistics that clearly distinguish 7PeerJ reviewing PDF | (v2013:06:581:0:0:NEW 13 Jun 2013) R ev ie w in g M an us cr ip t treated and untreated groups, but activation and inhibition can\u2019t be defined based upon a pre-defined cut-off for the test statistic value (e.g. 2 or -2). For this reason, we provide an ANOVA p-value to quantify the difference in test-statistic between groups, where the test statistic serves as a score for a second calculation of statistical significance. Additionally, we believe that predictive statistics are a useful method for accessing BD-Func scores for individual samples within large patient populations. In order to quantify the accuracy of the model without depending on a pre-defined cut-off, BD-Func produces receiver operating characteristic (ROC) plots for each cohort and the area under the curve (AUC) is calculated for each ROC plot (where a 100% accurate model would have an AUC=1.00) (Figure S2). The superior performance of the novel PGR signature on the breast cancer cohorts becomes even clearer when these predictive statistics are compared for the two models (Figure 3, Figure S3). Importantly, the novel signature showed the same level of accuracy on the TCGA breast cancer dataset as the original expO dataset. This is significant because the TCGA dataset is over twice as large as the expO dataset, and the TCGA dataset utilizes RNA-Seq while the expO dataset utilizes microarrays to quantify gene expression. In other words, this shows that BD-Func is capable of producing very robust predictions that translate across different genomic technologies. LBH589 (panobinostat) is a histone deacetylase inhibitor that has been previously shown to suppress the proliferation of aromatase inhibitor resistant breast cancer cells, which was a conclusion supported in part by functional enrichment analysis of commonly affected genes from 3 cell line experiments (Kubo et al. 2013). Gene lists derived from these cell line experiments can be easily used to define a BD-func signature, so we hypothesized that the results from this previous in vitro cell culture study could be used to predict LBH589 activity in vivo in an animal study. Specifically, we asked if the signature defined based upon LBH589 treatment in 3 cell lines (H295R, MCF-7her2, HeLa) could detect LBH589 activity in a mouse xenograft from a different cell line (MCF-7aro xenograft treated with EXE). Indeed, BD-Func correctly used the cell line LBH589 signature to identify common gene expression changes in the tumours treated with LBH589 and EXE compared to the mice that were only exposed to EXE treatment (Figure 4). 8PeerJ reviewing PDF | (v2013:06:581:0:0:NEW 13 Jun 2013) R ev ie w in g M an us cr ip t Next, we compared the BD-Func results to those profiles defined using the Connectivity Map, which is a commonly used tool to study gene expression profiles for drugs and other chemical perturbations (Lamb et al. 2006). Unfortunately, there are no LBH589 / panobinostat treatments in the Connectivity Map database. To be fair, this database can provide useful information: for example, the Connectivity Map indicates that LBH589 shows a gene expression signature that is correlated with fulvestrant treatment (an estrogen receptor antagonist used for metastatic breast cancer in postmenopausal women) and anti-correlated with estradiol and alpha-estradiol treatment. Likewise, the upstream regulator function in IPA predicts inhibition of beta-estradiol due to LBH589 treatment. All of these results match the expectation that LBH589 is an inhibitory modulator of aromatase (estrogen synthetase), resulting in the reduction of estrogen levels (57). Nevertheless, it should be emphasized that BD-Func can uniquely allow users to compare their own experiments to the LBH589 signature presented in this paper, which is not possible to achieve using these other very popular tools. Additionally, BD-Func is compatible with any gene mapping (in this case, gene symbol), whereas the Connectivity Map requires users to define their signatures in terms of HG-U133A probes. For example, affected gene symbols had to be converted to HG-U133A probes for this analysis. We believe that being able to define signatures based upon gene symbol is a substantial practical benefit. This figure shows the output figures from BD-Func for the 5 MSigDB signatures tested on their original dataset. A. Density plots for fold-change values for activated genes (colored red) and inhibited genes (colored green). These plots are used to illustrate BD-Func analysis on a single sample (in this case, fold-change values between the positive and negative populations). B. Box-plots of activation versus inhibition test statistics for all relevant samples in each of the 5 MSigDB datasets. Note that each box-plot describes the distribution of test statistics for each group \u2013 it does not represent the expression of an individual gene or a metagene. In each of these five examples, the test statistic shows very clear differences among the different groups. If the median t-test statistic is greater than 2, the box is colored red. If the median t-test statistic is less than -2, the box is colored green. PeerJ reviewing PDF | (v2013:06:581:0:0:NEW 13 Jun 2013) R ev ie w in g M an us cr ip t Figure 3 many pathways are characterized by a mix of activation and inhibition : This figure shows the initial signalling steps in the Wnt signalling pathway, as defined by KEGG ( Kanehisa & Goto 2000 ) . Up-regulated genes are shown in red and down-regulated genes are shown in green. A. Complete Activation All genes activating the Wnt signalling pathway are up-regulated and all inhibitors are down-regulated. B. Complete Inhibition All genes activating the Wnt signalling pathway are down-regulated and all inhibitors are up-regulated. C. Mixed Pattern All genes in the figure are up-regulated. It is unclear what the downstream expression levels should be, but one may hypothesize a mixed result from Figure 1A and Figure 1B. However, most functional enrichment tools would predict this as the pattern with the strongest up-regulation. Underneath each diagram is the expected signal distribution that would be produced by BD-Func. PeerJ reviewing PDF | (v2013:06:581:0:0:NEW 13 Jun 2013) R ev ie w in g M an us cr ip t Figure 2 custom progesterone signature outperforms msigdb signature on breast cancer patients : AUC (Area Under the Curve) values from the ROC plots for each patient cohort are displayed. Although the MSigBD gene set shows extremely high accuracy for the original meningioma dataset, it shows essentially random predictive power for the 6 larger breast cancer datasets. On the other hand, a custom progesterone receptor signature defined using the expO dataset shows high accuracy for all 7 cohorts, and the high accuracy is maintained even in the largest cohort (TCGA). PeerJ reviewing PDF | (v2013:06:581:0:0:NEW 13 Jun 2013) R ev ie w in g M an us cr ip t Figure 4 Novel Cell Line LBH589 Signature Can Accurately Detect Drug Activity In Vivo prediction of tgf\u03b2 activity in novel datasets :",
+    "url": "https://peerj.com/articles/160/reviews/",
+    "review_1": "William Jungers \u00b7 Aug 23, 2013 \u00b7 Academic Editor\nACCEPT\nThe authors have responded quickly and thoroughly to reviewer suggestions, and I believe the ms. has been improved as a result of these additions and emendations. This study provides a fascinating window into the function of an evolutionary novelty, and I believe it merits publication in its revised form.",
+    "review_2": "William Jungers \u00b7 Aug 21, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nBoth reviewers believe this is an interesting and well-written article that combines 3DGM with FEM. Both reviewers request that the authors provide an illustration with attachment sites for the muscle(s) included in the models, and I encourage the authors to do so.\nOne reviewer has no other substantive comments, but the second challenges some of the specific conclusions -- and these should be addressed in the revision. I tend to agree that the unloaded model unduly impacts the ordination space, and could be excluded without loss of information. I also agree with the reviewer's comment about Figure 4. Please detail pecisely how you have responded to the various suggestions and queries in the revision response files (including rebuttals).",
+    "review_3": "Reviewer 1 \u00b7 Aug 13, 2013\nBasic reporting\nNo comments\nExperimental design\nNo comments\nValidity of the findings\nNo comments\nAdditional comments\nThis is an excellent, well written article on a fascinating newly discovered species of rodent. I think that it would be very useful to include an additional figure that shows attachment sites and lines of action (or vectors) of the jaw adductor muscles used in the analyses.\n\nOne minor editorial note - line 137: Following Cox et al.........\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Masticatory biomechanics of the Laotian rock rat, Laonastes aenigmamus, and the function of the zygomaticomandibularis muscle (v0.1)\". PeerJ https://doi.org/10.7287/peerj.160v0.1/reviews/1",
+    "review_4": "Reviewer 2 \u00b7 Aug 8, 2013\nBasic reporting\nNo comments\nExperimental design\nNo comments\nValidity of the findings\nNo comments\nAdditional comments\nThis is a well written and interesting paper that uses a 3D analytical approach, combining both geometric morphometrics and finite element analysis, to understand the function of the Zygomaticomandibularis muscle.\n\nI have some suggestions below that the authors should consider (particularly interpretation of the results in the discussion section), and I especially think the ms would benefit from an additional figure that shows precisely the origins and insertions of the muscles used to make the FEM \u2013 this would make the results entirely more easy to interpret for the reader.\n\nLn17: \u201cspecific diversity\u201d should be \u201ctaxonomic diversity\u201d\nLn78: insert, \u201ce.g.\u201d before \u201cDumont\u201d ?\nLn104: \u201cCox et al.\u201d typo, and same for the next ref.\nLn153: A ratio is expressed as two numbers separated by a colon e.g. \u201c2:3\u201d. It\u2019s not the same as a fraction, percentage or proportion. The values expressed in Figure 3 relating to mechanical efficiency are incorrectly described as a ratio.\nLn185: Table 4 shows not demonstrates\nLn204: it is not clear why you included the unloaded model here \u2013 this clearly causes all of the variance in all other models to be clumped together on the right hand side of the plot, as a consequence of those all being very different from the unloaded model. Would it not be more informative to produce a plot excluding the unloaded model?\nLn209: maybe you should include a reference to Fig. 2 here (e.g. also shown in Fig. 2) after \u201cregion\u201d\nLn209: insert \u201cby PC2\u201d at the end of the sentence\nLn214: looking at Fig. 2 what appears to actually be happening is dorsal movement (causing stress) of the area of the skull above the molars, i.e. between the orbits at the beginning of the calvaria surface. With procrustes superimposition, you spread the deformation across all your landmarks.\nLn221: \u201c\u201dindicating that even more deformation is occurring in these models\u201d \u2013 so we conclude that the position of the IOZM, or at least the reason it\u2019s there at all, is to limit the amount of stress in the skull (by limiting the amount of strain/displacement etc)\nLn222: this is not inferred\nLn226: insert \u201crestrained\u201d after loaded\nLn236: or the geometry of the zygomatic is adapted to restrain this?\nLn239: I wonder if this is likely? A system is adapted whereby expenditure from the temporalis would be required every time the masseter is used, otherwise the zygomatic gets snapped.\nLn244: should read \u201cbut is\u201d\nLn248: this conclusion is difficult to evaluate without a figure showing us the FEM \u2013 i.e. the muscle insertions and origins\nLn254: I think this is an incorrect conclusion: look at Fig. 4, the position of attachment (again we need a figure with this) has little effect. BUT the presence or absence of IOZM is having an effect. Look at the orange symbols (no IOZM) in Fig 4, these are separate from the others because the absence of the IOZM is causing more PC shape change (which we know is a resultant of more strain = more stress). Mean VM stress might not pick this up because it's so vague and the pattern maybe similar (this depends again on where the other muscle attachments are, which we don't know) but this is likely controlled by skull morphology (mostly).\nLn259: But the reason it\u2019s there, as shown in Fig. 4 is that it reduces landmark displacement/strain (and hence stress magnitude)\nLn263: but again, I think the data show that despite the IOZM being relatively small and having long muscle fibres, it is very effective at reducing the strain in the skull.\nLn264: this is because stress distribution is largely controlled by skull morphology in this case\nLn267: \u201creduces the bit force generated at all teeth\u201d - but increases strain (Landmark displacement in the PCA, orange dots) i.e. removing the IOZM leads to a weaker bite AND a more strained/stressed skull. The mean stress and stress distribution may not appear different in Fig. 2, but the magnitudes of the stresses within this similar distribution must be higher with IOZM removed. This appears correct in Fig. 2 along the zygomatic arch, although images are inconclusive...\nLn268: again need a figure with muscle insertions\u2026\nLn277: not sure that is correct - your data show that the presence of the IOZM evolved to increase bite force and decrease strain. Moving the IOZM has no result in Fig 4 on strain.\n\nFigure 4: the small images of the skulls showing somehow deformation on the PC axes, are difficult to read \u2013 it\u2019s hard to quickly and clearly see which regions of the skull are deforming \u2013 could the authors not add vectors to these graphics to show the direction of change, or a wireframe?\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"Masticatory biomechanics of the Laotian rock rat, Laonastes aenigmamus, and the function of the zygomaticomandibularis muscle (v0.1)\". PeerJ https://doi.org/10.7287/peerj.160v0.1/reviews/2",
+    "pdf_1": "https://peerj.com/articles/160v0.2/submission",
+    "pdf_2": "https://peerj.com/articles/160v0.1/submission",
+    "all_reviews": "Review 1: William Jungers \u00b7 Aug 23, 2013 \u00b7 Academic Editor\nACCEPT\nThe authors have responded quickly and thoroughly to reviewer suggestions, and I believe the ms. has been improved as a result of these additions and emendations. This study provides a fascinating window into the function of an evolutionary novelty, and I believe it merits publication in its revised form.\nReview 2: William Jungers \u00b7 Aug 21, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nBoth reviewers believe this is an interesting and well-written article that combines 3DGM with FEM. Both reviewers request that the authors provide an illustration with attachment sites for the muscle(s) included in the models, and I encourage the authors to do so.\nOne reviewer has no other substantive comments, but the second challenges some of the specific conclusions -- and these should be addressed in the revision. I tend to agree that the unloaded model unduly impacts the ordination space, and could be excluded without loss of information. I also agree with the reviewer's comment about Figure 4. Please detail pecisely how you have responded to the various suggestions and queries in the revision response files (including rebuttals).\nReview 3: Reviewer 1 \u00b7 Aug 13, 2013\nBasic reporting\nNo comments\nExperimental design\nNo comments\nValidity of the findings\nNo comments\nAdditional comments\nThis is an excellent, well written article on a fascinating newly discovered species of rodent. I think that it would be very useful to include an additional figure that shows attachment sites and lines of action (or vectors) of the jaw adductor muscles used in the analyses.\n\nOne minor editorial note - line 137: Following Cox et al.........\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Masticatory biomechanics of the Laotian rock rat, Laonastes aenigmamus, and the function of the zygomaticomandibularis muscle (v0.1)\". PeerJ https://doi.org/10.7287/peerj.160v0.1/reviews/1\nReview 4: Reviewer 2 \u00b7 Aug 8, 2013\nBasic reporting\nNo comments\nExperimental design\nNo comments\nValidity of the findings\nNo comments\nAdditional comments\nThis is a well written and interesting paper that uses a 3D analytical approach, combining both geometric morphometrics and finite element analysis, to understand the function of the Zygomaticomandibularis muscle.\n\nI have some suggestions below that the authors should consider (particularly interpretation of the results in the discussion section), and I especially think the ms would benefit from an additional figure that shows precisely the origins and insertions of the muscles used to make the FEM \u2013 this would make the results entirely more easy to interpret for the reader.\n\nLn17: \u201cspecific diversity\u201d should be \u201ctaxonomic diversity\u201d\nLn78: insert, \u201ce.g.\u201d before \u201cDumont\u201d ?\nLn104: \u201cCox et al.\u201d typo, and same for the next ref.\nLn153: A ratio is expressed as two numbers separated by a colon e.g. \u201c2:3\u201d. It\u2019s not the same as a fraction, percentage or proportion. The values expressed in Figure 3 relating to mechanical efficiency are incorrectly described as a ratio.\nLn185: Table 4 shows not demonstrates\nLn204: it is not clear why you included the unloaded model here \u2013 this clearly causes all of the variance in all other models to be clumped together on the right hand side of the plot, as a consequence of those all being very different from the unloaded model. Would it not be more informative to produce a plot excluding the unloaded model?\nLn209: maybe you should include a reference to Fig. 2 here (e.g. also shown in Fig. 2) after \u201cregion\u201d\nLn209: insert \u201cby PC2\u201d at the end of the sentence\nLn214: looking at Fig. 2 what appears to actually be happening is dorsal movement (causing stress) of the area of the skull above the molars, i.e. between the orbits at the beginning of the calvaria surface. With procrustes superimposition, you spread the deformation across all your landmarks.\nLn221: \u201c\u201dindicating that even more deformation is occurring in these models\u201d \u2013 so we conclude that the position of the IOZM, or at least the reason it\u2019s there at all, is to limit the amount of stress in the skull (by limiting the amount of strain/displacement etc)\nLn222: this is not inferred\nLn226: insert \u201crestrained\u201d after loaded\nLn236: or the geometry of the zygomatic is adapted to restrain this?\nLn239: I wonder if this is likely? A system is adapted whereby expenditure from the temporalis would be required every time the masseter is used, otherwise the zygomatic gets snapped.\nLn244: should read \u201cbut is\u201d\nLn248: this conclusion is difficult to evaluate without a figure showing us the FEM \u2013 i.e. the muscle insertions and origins\nLn254: I think this is an incorrect conclusion: look at Fig. 4, the position of attachment (again we need a figure with this) has little effect. BUT the presence or absence of IOZM is having an effect. Look at the orange symbols (no IOZM) in Fig 4, these are separate from the others because the absence of the IOZM is causing more PC shape change (which we know is a resultant of more strain = more stress). Mean VM stress might not pick this up because it's so vague and the pattern maybe similar (this depends again on where the other muscle attachments are, which we don't know) but this is likely controlled by skull morphology (mostly).\nLn259: But the reason it\u2019s there, as shown in Fig. 4 is that it reduces landmark displacement/strain (and hence stress magnitude)\nLn263: but again, I think the data show that despite the IOZM being relatively small and having long muscle fibres, it is very effective at reducing the strain in the skull.\nLn264: this is because stress distribution is largely controlled by skull morphology in this case\nLn267: \u201creduces the bit force generated at all teeth\u201d - but increases strain (Landmark displacement in the PCA, orange dots) i.e. removing the IOZM leads to a weaker bite AND a more strained/stressed skull. The mean stress and stress distribution may not appear different in Fig. 2, but the magnitudes of the stresses within this similar distribution must be higher with IOZM removed. This appears correct in Fig. 2 along the zygomatic arch, although images are inconclusive...\nLn268: again need a figure with muscle insertions\u2026\nLn277: not sure that is correct - your data show that the presence of the IOZM evolved to increase bite force and decrease strain. Moving the IOZM has no result in Fig 4 on strain.\n\nFigure 4: the small images of the skulls showing somehow deformation on the PC axes, are difficult to read \u2013 it\u2019s hard to quickly and clearly see which regions of the skull are deforming \u2013 could the authors not add vectors to these graphics to show the direction of change, or a wireframe?\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"Masticatory biomechanics of the Laotian rock rat, Laonastes aenigmamus, and the function of the zygomaticomandibularis muscle (v0.1)\". PeerJ https://doi.org/10.7287/peerj.160v0.1/reviews/2\nReview 5: nan\nReview 6: \n",
+    "LLM_extracted_review": "[\"Review 1: The authors have responded quickly and thoroughly to reviewer suggestions, and I believe the manuscript has been improved as a result of these additions and emendations. This study provides a fascinating window into the function of an evolutionary novelty, and I believe it merits publication in its revised form.\\n\\nReview 2: Both reviewers believe this is an interesting and well-written article that combines 3DGM with FEM. Both reviewers request that the authors provide an illustration with attachment sites for the muscle(s) included in the models. One reviewer has no other substantive comments, but the second challenges some of the specific conclusions -- and these should be addressed in the revision. I tend to agree that the unloaded model unduly impacts the ordination space, and could be excluded without loss of information. I also agree with the reviewer's comment about Figure 4. \\n\\nReview 3: This is an excellent, well-written article on a fascinating newly discovered species of rodent. I think that it would be very useful to include an additional figure that shows attachment sites and lines of action (or vectors) of the jaw adductor muscles used in the analyses.\\n\\nReview 4: This is a well-written and interesting paper that uses a 3D analytical approach, combining both geometric morphometrics and finite element analysis, to understand the function of the Zygomaticomandibularis muscle. I have some suggestions below that the authors should consider, particularly interpretation of the results in the discussion section. I especially think the manuscript would benefit from an additional figure that shows precisely the origins and insertions of the muscles used to make the FEM. \\n\\nReview 5: nan\\n\\nReview 6: \"]"
+}

peerj_json_files/PeerJ_Json_117.json ADDED Viewed

	@@ -0,0 +1,18 @@

+{
+    "v1_Abstract": "The Laotian rock rat, Laonastes aenigmamus, is one of the most recently discovered species of rodent, and displays a cranial morphology that is highly specialised. The rostrum of L. aenigmamus is exceptionally elongate and bears a large attachment site for the infraorbital portion of the zygomaticomandibularis muscle (IOZM), which is particularly well-developed in this species. In this study, we used finite element analysis to investigate the biomechanical performance of the Laotian rock rat cranium and to elucidate the function of the IOZM. A finite element model of the skull of L. aenigmamus was constructed and solved for biting on each of the teeth (incisors, premolar and molars). Further load cases were created and solved in which the origin of the IOZM had been moved anteriorly and posteriorly along the rostrum. Finally, a set of load cases were produced in which the IOZM was removed entirely, and its force was redistributed between the remaining masticatory muscles. The analysis showed that, during biting, the most stressed areas of the skull were the zygomatic and orbital regions. Compared to other rodents, L. aenigmamus is highly efficient at incisor gnawing, but less efficient at molar chewing. However, a relatively constant bite force across the molar tooth row may be an adaptation to folivory. Movement of the origin of the IOZM had little on the patterns of von Mises stresses, or the overall stress experienced by the cranium. However, removal of the IOZM had a substantial effect on the total deformation experienced by the skull. In addition, the positioning and presence of the IOZM had large impact on bite force. Moving the IOZM origin to the anterior tip of the rostrum led to a substantially reduced bite force at all teeth. This was hypothesised to be a result of the increasing horizontal component to the pull of this muscle as it is moved anteriorly along the rostrum. Removal of the IOZM also resulted in a reduced bite force, even when the total input muscle force was maintained at the same level. It was thus concluded that the function of the IOZM in L. aenigmamus is to increase bite force whilst reducing cranial deformation. If the IOZM can be",
+    "v1_col_introduction": "introduction  : The Rodentia is the most speciose of all mammalian orders, with over 2,200 extant species (Wilson & Reeder, 2005) in addition to a large number of fossil forms. Despite such taxonomic diversity, morphological variation within the order is relatively limited, particularly with regard to the skull and mandible (Wood, 1965; Hautier et al., 2011). Hence, determining relationships between rodent species based on morphology is difficult, and rodent taxonomy has been historically controversial, with two competing classifications arising in the second half of the 19 th century. First, Brandt (1855) split the rodents into three suborders (Sciuromorpha, Hystricomorpha and Myomorpha) based on the morphology of the masseter muscle and its attachment to the rostrum (see Cox & Jeffery, 2011 for details). Later, Tullberg (1899) divided the rodents into two groups (Sciurognathi and Hystricognathi) based on the position of the angular process of the mandible relative to the incisor. Recent molecular phylogenies (Blanga-Kanfi et al., 2009; Churakov et al., 2010; Fabre et al., 2012) have shown that neither of these classifications accurately resolves the evolutionary relationships between the rodents. The three suborders of Brandt (1855), also used by Simpson (1945), have now been discarded, as they are thought to represent polyphyletic groupings of rodents, although the names have been retained in their adjectival form (sciuromorphous, etc) to describe the three morphotypes of the skull and\n1\n2\n3\n4\n5\n6\n7\n8\n9\n10\n11\n12\n13\n14\n15\n16\n17\n18\nPeerJ reviewing PDF | (v2013:07:700:1:0:CHECK 23 Aug 2013)\nR ev ie w in g M an\nus cr ip t\nmasseter (Wood, 1965; Cox & Jeffery, 2011). The classification of Tullberg (1899) has fared slightly better, with the Hystricognathi still recognised as a monophyletic clade (the Sciurognathi is paraphyletic with respect to the Hystricognathi). However, recent work (Hautier et al., 2011) has questioned the usefulness of sciurognathy and hystricognathy as morphological terms, noting that there exists a continuum of mandibular forms rather than two discrete morphologies, and that some members of the Hystricognathi actually have mandibles that appear to be almost sciurognathous in form.\nThe Laotian rock rat, Laonastes aenigmamus (Jenkins et al., 2005), is a recently discovered species of rodent from south-east Asia. It shows an unusual mixture of cranial, mandibular and muscular morphologies, combining a large part of the zygomaticomandibularis muscle that extends through the enlarged infraorbital foramen to attach to the rostrum (Hautier & Saksiri, 2009) with a weak lateral displacement of the angular process of the mandible (Hautier et al., 2011). Thus, the Laotian rock rat brings together a hystricomorphous skull and masseter with a lower jaw that is intermediate between sciurognathous and hystricognathous. This combination of characters has made its phylogenetic relationships difficult to ascertain. When first described, a new family, the Laonastidae, was created to house L. aenigmamus (Jenkins et al., 2005). This family was placed within the Hystricognathi as the sister-group to African mole-rats (Bathyergidae) or the dassie rat (Petromuridae). A subsequent analysis (Dawson et al., 2006) showed that L. aenigmamus was in fact a member of the Diatomyidae, a family of rodents previously thought to have gone extinct in the Miocene. Further work then showed the Laotian rock rat to be the sister-taxon to the Ctenodactylidae (Huchon et al., 2007), a family of rodents that also display the combination of a hystricomorphous skull with a sciurognathous mandible (Hautier, 2010). The Ctenodactylidae and Diatomyidae together form the sister-group the Hystricognathi, within the more inclusive clade Ctenohystrica (Fabre et al., 2012).\nOne particularly notable characteristic of the Laotian rock rat is the morphology of the zygomaticomandibularis muscle. This muscle is the innermost layer of the masseter, and in the hystricomorph condition extends anteriorly through the orbit and the grossly enlarged infraorbital foramen to take an attachment on the rostrum (Wood, 1965; Cox & Jeffery, 2011). In L. aenigmamus, the infraorbital portion of the zygomaticomandibularis (IOZM) is especially welldeveloped, extending over halfway along the rostrum (Hautier & Saksiri, 2009), which is itself exceptionally elongated compared to other rodents (Herrel et al., 2012). Various functions have\n19\n20\n21\n22\n23\n24\n25\n26\n27\n28\n29\n30\n31\n32\n33\n34\n35\n36\n37\n38\n39\n40\n41\n42\n43\n44\n45\n46\n47\n48\n49\nPeerJ reviewing PDF | (v2013:07:700:1:0:CHECK 23 Aug 2013)\nR ev ie w in g M an\nus cr ip t\nbeen proposed for the IOZM. Becht (1953) noted that its origin on the rostrum between the incisors and cheek teeth would enable it to function as a second class lever during chewing at the molars, but as a third class lever during gnawing at the incisors. Thus, it can produce both fine control at the incisors and strong pressure at the molars. Using electromyography, Weijs & Dantuma (1975) found that the IOZM was firing at low intensities during jaw opening in rats, and suggested that it may have a role in the fine control of the opening phase of mastication. With regard to L. aenigmamus specifically, Herrel et al. (2012) proposed that the strong development of the IOZM and its anterior origin on the rostrum would produce a strong horizontal force component to the bite. This would result in optimal functioning at low gape angles and the generation of uniform force along the tooth row, both of which would be beneficial for a folivorous diet, which has been suggested for the Laotian rock rat (Scopin et al., 2011).\nThe aim of this study is to investigate the biomechanics of feeding in Laonastes aenigmamus. Finite element analysis (FEA) will be used to examine the response of the cranium during gnawing at the incisors and chewing at the molars. FEA is a computer-based engineering technique that enables the prediction of stress, strain and deformation in a complex geometric object subjected to a load (Rayfield, 2007). It has been successfully used to study the biomechanics of feeding in a number of mammalian groups (e.g. Dumont, Piccirillo & Grosse, 2005; Kupczik et al., 2007; Wroe, 2010; Bright & Rayfield, 2011) including rodents (Cox et al., 2011, 2012). The major advantage of FEA in the study of biological systems is that elements such as muscle or bone can be modified or removed at will, without the practical and ethical concerns that would arise with in vivo work. In this study, the contribution of the IOZM to feeding will be investigated by changing its attachment site and removing it altogether. In this way, we aim to understand the function of the IOZM, which is so highly developed in the Laotian rock rat, and to elucidate why such an unusual morphology has evolved in this species.",
+    "v2_Abstract": "The Laotian rock rat, Laonastes aenigmamus, is one of the most recently discovered species of rodent, and displays a cranial morphology that is highly specialised. The rostrum of L. aenigmamus is exceptionally elongate and bears a large attachment site for the infraorbital portion of the zygomaticomandibularis (IOZM), which is particularly well-developed in this species. In this study, we used finite element analysis to investigate the biomechanical performance of the Laotian rock rat cranium and to elucidate the function of the IOZM muscle. A finite element model of the skull of L. aenigmamus was constructed and solved for biting on each of the teeth (incisors, premolar and molars). Further load cases were created and solved in which the origin of the IOZM had been moved anteriorly and posteriorly along the rostrum. Finally, a set of load cases were produced in which the IOZM was removed entirely, and its force was redistributed between the remaining masticatory muscles. The analysis showed that, during biting, the most stressed areas of the skull were the zygomatic and orbital regions. Compared to other rodents, L. aenigmamus is highly efficient at incisor gnawing, but less efficient at molar chewing. However, a relatively constant bite force across the molar tooth row may be an adaptation to folivory. Neither movement nor removal of the IOZM had much effect on the patterns of von Mises stresses, or the overall stress experienced by the cranium. However, the positioning and presence of the IOZM did have a large effect on bite force. Moving the IOZM origin to the anterior tip of the rostrum led to a substantially reduced bite force at all teeth. This was hypothesised to be a result of the increasing horizontal component to the pull of this muscle as it moves anteriorly along the rostrum. Removal of the IOZM also resulted in a reduced bite force, even when the total input muscle force is maintained at the same level, and in addition increased the total deformation across the skull. It was thus concluded that the function of the IOZM in L. aenigmamus is to increase bite force whilst reducing cranial deformation. If the IOZM can be shown to have this function in other rodent groups, this may help explain the evolution of this muscle, and may also provide an understanding of why it has evolved independently several times within rodents.",
+    "v2_col_introduction": "introduction  : The Rodentia is the most speciose of all mammalian orders, with over 2,200 extant species (Wilson & Reeder, 2005) in addition to a large number of fossil forms. Despite such specific diversity, morphological variation within the order is relatively limited, particularly with regard to the skull and mandible (Wood, 1965; Hautier et al., 2011). Hence, determining relationships between rodent species based on morphology is difficult, and rodent taxonomy has been historically controversial, with two competing classifications arising in the second half of the 19 th century. First, Brandt (1855) split the rodents into three suborders (Sciuromorpha, Hystricomorpha and Myomorpha) based on the morphology of the masseter muscle and its attachment to the rostrum (see Cox & Jeffery, 2011 for details). Later, Tullberg (1899) divided the rodents into two groups (Sciurognathi and Hystricognathi) based on the position of the angular process of the mandible relative to the incisor. Recent molecular phylogenies (Blanga-Kanfi et al., 2009; Churakov et al., 2010; Fabre et al., 2012) have shown that neither of these classifications accurately resolves the evolutionary relationships between the rodents. The three suborders of Brandt (1855), also used by Simpson (1945), have now been discarded, as they are thought to represent polyphyletic groupings of rodents, although the names have been retained in their adjectival form (sciuromorphous, etc) to describe the three morphotypes of the skull and\n1\n2 3 4 5 6 7 8\n9 10 11 12 13\n14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30\nPeerJ reviewing PDF | (v2013:07:700:0:1:NEW 1 Aug 2013)\nR ev ie w in g M an\nus cr ip t\nmasseter (Wood, 1965; Cox & Jeffery, 2011). The classification of Tullberg (1899) has fared slightly better, with the Hystricognathi still recognised as a monophyletic clade (the Sciurognathi is paraphyletic with respect to the Hystricognathi). However, recent work (Hautier et al., 2011) has questioned the usefulness of sciurognathy and hystricognathy as morphological terms, noting that there exists a continuum of mandibular forms rather than two discrete morphologies, and that some members of the Hystricognathi actually have mandibles that appear to be almost sciurognathous in form.\nThe Laotian rock rat, Laonastes aenigmamus (Jenkins et al., 2005), is a recently discovered species of rodent from south-east Asia. It shows an unusual mixture of cranial, mandibular and muscular morphologies, combining a large part of the zygomaticomandibularis muscle that extends through the enlarged infraorbital foramen to attach to the rostrum (Hautier & Saksiri, 2009) with a weak lateral displacement of the angular process of the mandible (Hautier et al., 2011). Thus, the Laotian rock rat brings together a hystricomorphous skull and masseter with a lower jaw that is intermediate between sciurognathous and hystricognathous. This combination of characters has made its phylogenetic relationships difficult to ascertain. When first described, a new family, the Laonastidae, was created to house L. aenigmamus (Jenkins et al., 2005). This family was placed within the Hystricognathi as the sister-group to African mole-rats (Bathyergidae) or the dassie rat (Petromuridae). A subsequent analysis (Dawson et al., 2006) showed that L. aenigmamus was in fact a member of the Diatomyidae, a family of rodents previously thought to have gone extinct in the Miocene. Further work then showed the Laotian rock rat to be the sister-taxon to the Ctenodactylidae (Huchon et al., 2007), a family of rodents that also display the combination of a hystricomorphous skull with a sciurognathous mandible (Hautier, 2010). The Ctenodactylidae and Diatomyidae together form the sister-group the Hystricognathi, within the more inclusive clade Ctenohystrica (Fabre et al., 2012).\nOne particularly notable characteristic of the Laotian rock rat is the morphology of the zygomaticomandibularis muscle. This muscle is the innermost layer of the masseter, and in the hystricomorph condition extends anteriorly through the orbit and the grossly enlarged infraorbital foramen to take an attachment on the rostrum (Wood, 1965; Cox & Jeffery, 2011). In L. aenigmamus, the infraorbital portion of the zygomaticomandibularis (IOZM) is especially well-developed, extending over halfway along the rostrum (Hautier & Saksiri, 2009), which is itself exceptionally elongated compared to other rodents (Herrel et al., 2012). Various functions\n31 32 33 34 35 36 37\n38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54\n55 56 57 58 59 60 61\nPeerJ reviewing PDF | (v2013:07:700:0:1:NEW 1 Aug 2013)\nR ev ie w in g M an\nus cr ip t\nhave been proposed for the IOZM. Becht (1953) noted that its origin on the rostrum between the incisors and cheek teeth would enable it to function as a second class lever during chewing at the molars, but as a third class lever during gnawing at the incisors. Thus, it can produce both fine control at the incisors and strong pressure at the molars. Using electromyography, Weijs & Dantuma (1975) found that the IOZM was firing at low intensities during jaw opening in rats, and suggested that it may have a role in the fine control of the opening phase of mastication. With regard to L. aenigmamus specifically, Herrel et al. (2012) proposed that the strong development of the IOZM and its anterior origin on the rostrum would produce a strong horizontal force component to the bite. This would result in optimal functioning at low gape angles and the generation of uniform force along the tooth row, both of which would be beneficial for a folivorous diet, which has been suggested for the Laotian rock rat (Scopin et al., 2011).\nThe aim of this study is to investigate the biomechanics of feeding in Laonastes aenigmamus. Finite element analysis (FEA) will be used to examine the response of the cranium during gnawing at the incisors and chewing at the molars. FEA is a computer-based engineering technique that enables the prediction of stress, strain and deformation in a complex geometric object subjected to a load (Rayfield, 2007). It has been successfully used to study the biomechanics of feeding in a number of mammalian groups (Dumont, Piccirillo & Grosse, 2005; Kupczik et al., 2007; Wroe, 2010; Bright & Rayfield, 2011) including rodents (Cox et al., 2011, 2012). The major advantage of FEA in the study of biological systems is that elements such as muscle or bone can be modified or removed at will, without the practical and ethical concerns that would arise with in vivo work. In this study, the contribution of the IOZM to feeding will be investigated by changing its attachment site and removing it altogether. In this way, we aim to understand the function of the IOZM, which is so highly developed in the Laotian rock rat, and to elucidate why such an unusual morphology has evolved in this species.",
+    "v1_text": "materials & methods : results : Figure 3 shows the von Mises stress patterns generated across the skull of L. aenigmamus during biting at the incisor, first molar and third molar. It can be seen that, aside from the biting tooth, the zygomatic arch is the most stressed region of the skull, followed by the orbital region. The rostrum experiences a moderate degree of stress during incisor gnawing, but is unstressed during molar chewing, and the occipital region is unstressed during all bites. From visual inspection of the stress distribution figures it is difficult to determine a great deal of variation between bites at different points along the tooth row, even between incisors and molars. However, by studying the mean von Mises stresses across the skull (Table 3) it can be seen that there are indeed subtle differences between bites on different teeth. In general, overall stress increases as the bite point moves closer to the jaw articulation from the premolar to the third molar. However, the incisor bite does not fit this trend, and shows the greatest mean stress across the skull of all bites, presumably because the rostrum is stressed in incisor bites but not in molar bites. The maximum bite force predicted at each tooth is given in Table 4. As would be expected from simple lever mechanics, bite force increases the closer the bite point (the biting tooth) is to the fulcrum (the TMJ). By dividing the bite forces by total input muscle force, the mechanical efficiency of biting has been calculated and can be compared with previous analyses on other rodents (Cox et al., 2012). It can be seen (Figure 4) that the Laotian rock rat performs well at the incisors and premolars, outcompeting the squirrel and guinea pig, and even matching the rat for efficiency at the incisors. However, compared to the other rodents, the mechanical efficiency tails off towards the distal molars and at the third molar, L. aenigmamus has the least efficient bite. The impact of the IOZM muscle was investigated by altering the position of its origin on the rostrum in VOX-FE. Figure 3 shows the von Mises stress patterns generated across the skull in models with the origin of the IOZM moved anteriorly and posteriorly, as well as in its original position. Despite quite large changes in muscle origins, very few differences can be seen between the models. This situation is confirmed by examining the mean von Mises stresses across the skull (Table 3). It can be seen that moving the IOZM origin anteriorly or posteriorly results in very little change (less than 6%) in mean von Mises stress during both incisor gnawing and molar 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 PeerJ reviewing PDF | (v2013:07:700:1:0:CHECK 23 Aug 2013) R ev ie w in g M an us cr ip t biting. Although not presented here, the same lack of difference was found on examination of the principal strains across the skull. Predicted bite force was also affected very little by moving the IOZM origin caudally to the most posterior part of the rostrum. However, Table 4 shows that an anterior shift of the IOZM origin resulted in quite a substantial reduction in bite force (between 10 and 13%). The effect of removing the IOZM muscle (and redistributing its force between the remaining masticatory muscles) can be seen in Figure 3. As when the IOZM origin was moved, few differences could be detected between the von Mises stress patterns by visual inspection alone, and again, this assertion is supported by examination of the mean von Mises stresses across the skull (Table 3). However, it can be seen from Table 4 that the action of the IOZM has a noticeable effect on bite force. It is clear that the presence of the IOZM enables L. aenigmamus to generate a greater bite force than if it were absent, even when the total input muscle force is the same. The reduction in bite force resulting from removal of the IOZM is approximately 10% for bites on both incisors and molars, which is similar to the effect of shifting the origin of this muscle to the anterior part of the rostrum. In order to analyse the subtle differences between the deformation patterns generated by the various load cases described above, a geometric morphometric analysis was performed. Landmark data from the original solved models, the models with the IOZM moved anteriorly and posteriorly, and models with the IOZM removed and its force redistributed, as well as landmarks from the unsolved model, were all subjected to Procrustes superimposition and PCA. The plot of the first two principal components (together comprising over 98% of the variation; PC1, 63.2%; PC2, 35.1%) is shown in Figure 5. The first principal component largely separates the unloaded model from the solved load cases. The point representing the unloaded model (star symbol) is on the far left of the plot whilst the points representing the loaded models are spread out down the right hand side of the diagram. The warped reconstructions indicate that the difference in deformation between the loaded and unloaded models is largely concentrated in the zygomatic region (also shown in Fig. 3). The incisor bites (squares) are clearly separated from bites on the other teeth on PC2. The cheek teeth bites are more closely grouped together, but separable into bites on each of the different teeth and positioned in order along the second principal component from the premolar (diamonds) to the third molar (lines). It can be seen from the reconstructions at the extremes of PC2 that incisor bites tend to deflect the rostrum dorsally (relative to the orbito- 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 PeerJ reviewing PDF | (v2013:07:700:1:0:CHECK 23 Aug 2013) R ev ie w in g M an us cr ip t temporal region), but that molar bites lead to a dorsal movement of the orbital region (relative to the rostrum). Within bites on each tooth, the models with the IOZM origin in its original position (blue symbols), moved anteriorly (green symbols) and moved posteriorly (red symbols) all group together closely at a similar distance from the unloaded model. This demonstrates that a similar amount of deformation is occurring in each of these load cases. However, the symbols representing the models with the IOZM muscle force redistributed between the other masticatory muscles (orange symbols), are positioned further from the unloaded model than the other load cases, indicating that even more deformation is occurring in these models. As the displacement of these symbols is along PC1, it can be seen that the redistribution of the IOZM is leading to greater deformation mainly in the zygomatic region. acknowledgements : We thank Jean-Pierre Hugot for providing access to the specimens and Dominique Adriaens, Loes Brabant and Luc Van Hoorebeke for scanning the specimens at the University of Ghent CT 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 PeerJ reviewing PDF | (v2013:07:700:1:0:CHECK 23 Aug 2013) R ev ie w in g M an us cr ip t facility (UGCT). We are grateful to Paul O\u2019Higgins and Michael Fagan for access to VOX-FE finite element software. Thanks are due to Laura Fitton for help with model construction, Peter Bazira for technical support with high-performance computing, and Andrew McIntosh and Thomas P\u00fcschel for assistance with GMM software. discussion : A finite element model of the skull of the Laotian rock rat, Laonastes aenigmamus, was created, loaded, constrained and solved. It was shown that the area of the skull experiencing the highest levels of stress was the zygomatic arch, including the zygomatic processes of the maxillary and frontal bones, which is likely to be a result of the large amount of masticatory musculature that attaches directly to this area. Similarly high zygomatic stresses have been noted in other rodents (Cox et al., 2012) as well as in other mammalian groups (Dumont, Piccirillo & Grosse, 2005; Bright & Rayfield, 2011; Dumont et al., 2011). It has been suggested that, in primates, the downward pull of the masseter muscle on the zygomatic arch may be counterbalanced to some degree by the upward pull of a soft tissue structure, namely the temporal fascia (Curtis et al., 2011). Despite many careful dissections (e.g. Baverstock, Jeffery & Cobb, 2013), no temporal fascia has been found in rodents, and this is also true of L. aenigmamus (Hautier & Saksiri, 2009). Thus, for the time being, it must be assumed that, although the zygomatic stresses are high in L. aegnimamus, they are not so high as to pose a danger of bone fracture. The bite forces predicted in this study demonstrate that the skull of L. aenigmamus can generate bites of 29 N during gnawing and between 53 and 74 N during chewing. By dividing these values by the total input muscle force, the mechanical efficiency of biting was calculated. It was shown that the Laotian rock rat is particularly efficient at incisor biting, having a mechanical efficiency greater than squirrels or guinea pigs, and similar to that of rats (Cox et al., 2012), but is less efficient at molar bites compared to these three rodent species. This would seem to indicate that L. aenigmamus is well-adapted for gnawing and less so for chewing, a conclusion that at first 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 PeerJ reviewing PDF | (v2013:07:700:1:0:CHECK 23 Aug 2013) R ev ie w in g M an us cr ip t glance appears somewhat at odds with the suggested diet of this species, which is thought to be largely folivorous (Scopin et al., 2011). However, Onoda et al. (2011) have proposed that the generation of a uniform bite force across the tooth row may be beneficial in the processing of fibrous plant material such as leaves. Viewed in this light, L. aenigmamus is well-adapted to folivory with a difference of just 20 N between its premolar and M3 bites, compared to 30-35 N in squirrels and guinea pigs. One of the more intriguing results of this study is that the position of the IOZM muscle has little effect on the overall stress experienced by the skull of L. aenigmamus. It can be seen from both Figure 3 and Table 3 that neither the pattern of von Mises stresses nor the mean stress across the skull are greatly affected by moving the origin of the IOZM forwards or backwards. This is may be because the IOZM is contributing a relatively small proportion of the total muscle force (around 8%). Although the IOZM appears quite a large muscle in lateral view (see Hautier & Saksiri, 2009), it is relatively small compared to the superficial and deep masseters. Moreover, it also possesses the longest muscle fibres of all masticatory muscles, and so its PCSA (and hence muscle force) is comparatively reduced (Table 1). Thus, its ability to impact the overall stress patterns across the skull may be fairly limited and, furthermore, these patterns may be more strongly influenced by skull morphology than by muscle geometry. In contrast to the position of its attachment, the presence or absence of the IOZM has a much stronger effect on the cranial biomechanics of L. aenigmamus. Figure 5 shows that the symbols representing the models without the IOZM are positioned further from the unloaded model that the other load cases, indicating that the removal of the IOZM leads to greater deformation (and therefore strain) across the skull. Thus, one of the major functions of the IOZM, at least in L. aenigmamus, appears to be to minimise strain during feeding. This conclusion holds true at all bites, both gnawing and chewing. In addition to its effect on cranial deformation, it was also found that the IOZM has a strong impact on the bite force produced by L. aenigmamus (Table 4). Specifically, removing the IOZM altogether (and redistributing its force between the remaining masticatory muscles) reduces the bite force generated at all teeth. Similarly, moving the origin of the IOZM to the anterior tip of the rostrum reduces bite force to a similar degree. It is likely that this effect is a result of the wrapping of the IOZM around the zygomatic process of the maxilla, which means that as the 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 PeerJ reviewing PDF | (v2013:07:700:1:0:CHECK 23 Aug 2013) R ev ie w in g M an us cr ip t IOZM is moved forward, its vector of pull becomes more horizontal (Fig.1), and so less able to generate bite force. On the other hand, moving the origin of the IOZM posteriorly appears to have little effect on bite force. Interestingly, the reduction in bite force resulting from the anterior movement or removal of the IOZM is largely constant along the tooth row \u2013 around 10%. Therefore, it would seem that the extension of the zygomaticomandibularis on to the rostrum evolved in order to increase bite force (as proposed by Becht, 1953). Given that there is minimal difference in bite force between the models with the IOZM in its original position and those with the origin moved posteriorly, the position of the IOZM origin halfway along the rostrum may simply have coevolved with the lengthening of the rostrum in this species. Alternatively, Herrel et al. (2012) have suggested that the anterior insertion of the IOZM in L. aenigmamus increases the horizontal component of biting, which leads to anterior displacement of the mandible during jaw closing, which may be advantageous for the processing of leaves. Whatever the driving force behind the anterior extension of the IOZM, the results here indicate that the reason that the IOZM does not extend any further along the rostrum, as is seen in some other hystricomorph rodents such as the springhare (Offermans & De Vree, 1989) and capybara (M\u00fcller, 1933), is because this would lead to a reduction in bite force at both the incisors and cheek teeth (Table 4). The results of this study have provided some important insights into the role of the IOZM muscle in the feeding behaviour of L. aenigmamus. Further investigations into other rodents, particularly other hystricomorphs, will enable us to understand whether the ability of the IOZM to increase bite force is unique to L. aenigmamus or common to all rodents that possess it. If the latter scenario is true, this could provide a selective advantage that may have driven the evolution of the IOZM, and could explain why it has evolved independently in several rodent groups (Ctenohystrica, Dipodidae, Anomaluroidea and Gliridae). However, finite element models can only shed light on static loading, and therefore cannot inform about the dynamic processes to which the IOZM may contribute, such as propalineal (antero-posterior) movements of the lower jaw, jaw opening or the fine control of gnawing. These activities can be addressed with dynamic modelling techniques, such as multibody dynamics analysis (e.g. Jones et al., 2012), and may well provide a fruitful avenue of research in the future. corresponding author: philip g. cox : Address: Centre for Anatomical and Human Sciences, Hull York Medical School, University of Hull, Cottingham Road, Hull, HU6 7RX, UK Telephone: 01904 321744 Email: philip.cox@hyms.ac.uk model solution and analysis : The finite element model of Laonastes aenigmamus was solved for biting at each tooth along the dental arcade, using VOX-FE. Gnawing was assumed always to be bilateral, as a result of the close apposition of the incisors, whereas chewing at the premolars and molars was modelled unilaterally on the left side. Von Mises stress patterns across the skull and bite forces at the teeth were recorded from the solved models. Following Cox et al. (2011) and O\u2019Higgins et al. (2011), geometric morphometrics was used to study the deformation patterns across the cranium. Thus, a set of landmarks (3D co-ordinate data) was recorded from each loaded skull, as well as from the original unsolved model. The landmark set was partially based on that in Cox et al. (2011) and is shown in Figure 2. Landmarks were concentrated in the rostral, orbital and zygomatic regions, as these were the areas experiencing the highest strains. The landmarks were then subjected to a geometric morphometric analysis using MorphoJ software (Klingenberg, 2011). This consisted of co-registering the landmarks via Procrustes superimposition followed by a principal components analysis (PCA). Cranial deformations along PC axes were visualised using the EVAN toolbox (http://www.evan-society.org). The results generated by the Laotian rock rat model were compared to those found in other rodents, such as squirrels, guinea pigs and rats (Cox et al., 2012). As it difficult to compare absolute bite force between models of different sizes, the mechanical efficiency of biting was calculated. Mechanical efficiency of biting is the ratio of output bite force to input muscle force, and represents the proportion of muscle force that is translated into bite force i.e. is not lost to deformation of the mandible or generation of joint reaction force at the condyles (Dumont et al., 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 PeerJ reviewing PDF | (v2013:07:700:1:0:CHECK 23 Aug 2013) R ev ie w in g M an us cr ip t 2011). As it is a proportion, mechanical efficiency is size-independent and facilitates clearer comparisons between skulls of varying sizes. authors: philip g cox1, joanna kirkham2,3 and anthony herrel4 : 1Centre for Anatomical and Human Sciences, Hull York Medical School, University of Hull, Hull, UK 2Centre for Anatomical and Human Sciences, Hull York Medical School, University of York, York, UK 3College of Medical and Dental Sciences, University of Birmingham, Birmingham, UK 4UMR CNRS/MNHN 7179, M\u00e9canismes adaptatifs: des organismes aux communaut\u00e9s, Paris, France sample and model creation : MicroCT scans of an adult female Laonastes aenigmamus (specimen number KY213), previously obtained for an earlier research project (Herrel et al., 2012), were provided by AH. Voxels were isometric and the voxel resolution was 0.137 mm. Further details of the scanning protocol can be found in Herrel et al. (2012). 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 PeerJ reviewing PDF | (v2013:07:700:1:0:CHECK 23 Aug 2013) R ev ie w in g M an us cr ip t Avizo 7.0 (Visualization Sciences Group, Burlington, MA, USA) was used to create a 3D volume reconstruction of the cranium of the Laotian rock rat. The bone and teeth were segmented separately so that different elastic properties could be assigned to them. Within the incisors, the enamel, dentine and pulp were not differentiated, as the scan resolution was not sufficient to distinguish these materials from one another. In addition, varying the material properties of these components has been shown to make little difference to the overall deformation of the skull (Cox et al., 2011). For similar reasons, the periodontal ligament was not included (Wood et al., 2011). The model was converted to an eight-noded FE mesh by direct voxel conversion, using VOX-FE, custom-built FE software (Liu et al., 2012), resulting in a model of 1734787 elements. Material properties, constraints and loads Bone was assigned a Young\u2019s modulus of 17 GPa, and the teeth were given a Young\u2019s modulus of 30 GPa. Both materials were modelled to be linearly elastic and isotropic with a Poisson\u2019s ratio of 0.3. These values were based on previous nano-indentation work on rodents (Cox et al., 2012) and FE studies on other mammals (Kupczik et al., 2012). The model was constrained at three locations: the left and right temporo-mandibular joints (TMJ) and the biting tooth. The TMJs were constrained on the ventral surface of the zygomatic process of the squamosal in all three dimensions. However, the biting tooth was only constrained in the dorsoventral axis (i.e. perpendicular to the occlusal plane). The number of nodes constrained at each location varied between 191 and 221. Loads were added to the model to represent the forces generated by the following muscles: superficial masseter; deep masseter; zygomaticomandibularis (anterior, posterior and infraorbital parts); temporalis (main, orbital and posterior parts); internal pterygoid; and external pterygoid. Muscle attachment sites and directions of pull were assigned based on the detailed dissections presented in Hautier & Saksiri (2009) and are shown in Figure 1. Table 1 gives the muscle mass and mean fibre length of each masticatory muscle, measured from the dissection of specimen KY213. Muscle masses were converted to volumes, assuming a muscle density of 1.0564 gcm-3 (Murphy & Beardsley, 1974), and the physiological cross-sectional area (PCSA) of each muscle (given in Table 1) was calculated by dividing the muscle volume by mean fibre length. Muscle forces (Table 2) were calculated by multiplying PCSAs by an intrinsic muscle stress value of 0.3 Nmm-2 (van Spronsen et al., 1989). In order to investigate the function of the IOZM, alternate 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 PeerJ reviewing PDF | (v2013:07:700:1:0:CHECK 23 Aug 2013) R ev ie w in g M an us cr ip t versions of the models were created with the origin of this muscle moved anteriorly and posteriorly, and also with the IOZM omitted completely on each side. In the models with a posteriorly shifted IOZM, the insertion point of the IOZM was kept the same, resulting in a more vertically directed muscle vector. However, in the models with an anteriorly shifted IOZM, the insertion point was moved to reflect the wrapping of the muscle around the zygomatic process of the maxilla, resulting in a highly horizontally directed muscle vector (see Fig. 1). To facilitate comparisons between models with and without the IOZM whilst retaining the same overall input muscle load, where it was omitted, the force of the IOZM was redistributed across the remaining masticatory muscles. This was done in such a way as to preserve their relative proportions. The muscle forces applied in the absence of the IOZM are given in Table 2. masticatory biomechanics of the laotian rock rat, : Laonastes aenigmamus, and the Function of the Zygomaticomandibularis Muscle The Laotian rock rat, Laonastes aenigmamus, is one of the most recently discovered species of rodent, and displays a cranial morphology that is highly specialised. The rostrum of L. aenigmamus is exceptionally elongate and bears a large attachment site for the infraorbital portion of the zygomaticomandibularis muscle (IOZM), which is particularly well-developed in this species. In this study, we used finite element analysis to investigate the biomechanical performance of the Laotian rock rat cranium and to elucidate the function of the IOZM. A finite element model of the skull of L. aenigmamus was constructed and solved for biting on each of the teeth (incisors, premolar and molars). Further load cases were created and solved in which the origin of the IOZM had been moved anteriorly and posteriorly along the rostrum. Finally, a set of load cases were produced in which the IOZM was removed entirely, and its force was redistributed between the remaining masticatory muscles. The analysis showed that, during biting, the most stressed areas of the skull were the zygomatic and orbital regions. Compared to other rodents, L. aenigmamus is highly efficient at incisor gnawing, but less efficient at molar chewing. However, a relatively constant bite force across the molar tooth row may be an adaptation to folivory. Movement of the origin of the IOZM had little on the patterns of von Mises stresses, or the overall stress experienced by the cranium. However, removal of the IOZM had a substantial effect on the total deformation experienced by the skull. In addition, the positioning and presence of the IOZM had large impact on bite force. Moving the IOZM origin to the anterior tip of the rostrum led to a substantially reduced bite force at all teeth. This was hypothesised to be a result of the increasing horizontal component to the pull of this muscle as it is moved anteriorly along the rostrum. Removal of the IOZM also resulted in a reduced bite force, even when the total input muscle force was maintained at the same level. It was thus concluded that the function of the IOZM in L. aenigmamus is to increase bite force whilst reducing cranial deformation. If the IOZM can be PeerJ reviewing PDF | (v2013:07:700:1:0:CHECK 23 Aug 2013) R ev ie w in g M an us cr ip t shown to have this function in other rodent groups, this may help explain the evolution of this muscle, and may also provide an understanding of why it has evolved independently several times within rodents. PeerJ reviewing PDF | (v2013:07:700:1:0:CHECK 23 Aug 2013) R ev ie w in g M an us cr ip t 16, midpoint between 15 and 17; 17, anteriormost attachment of zygomatic arch to rostrum; : 18; posteriormost point of infraorbital margin; 19, point on ventrolateral margin of zygomatic : arch in same coronal plane as midpoint of M1; 20, apex of tubercle on anterior orbital margin; 21, dorsalmost point on orbital margin; 22, ventralmost point on orbital margin; 23, midpoint : between 21 and 24; 24, posteriormost point on orbital margin. Landmarks 11-24recorded on both sides. PeerJ reviewing PDF | (v2013:07:700:1:0:CHECK 23 Aug 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:07:700:1:0:CHECK 23 Aug 2013) R ev ie w in g M an us cr ip t Figure 3 Predicted distribution of von Mises stresses across the skull of Laonastes aenigmamus. Arrows indicate the biting tooth. First column, incisor bites; second column, M1 bites; third column, M3 bites. First line, original models; second line, origin of IOZM moved anteriorly to front of rostrum; third line, origin of IOZM moved posteriorly to back of rostrum; fourth line, IOZM muscle force removed and redistributed proportionally between the remaining masticatory muscles. PeerJ reviewing PDF | (v2013:07:700:1:0:CHECK 23 Aug 2013) R ev ie w in g M an us cr ip t Figure 4 Mechanical efficiency of biting at each tooth Predicted from FE models of squirrel, guinea pig, rat and Laotian rock rat skulls. Data for squirrel, guinea pig and rat from Cox et al. (2012). Abbreviations: I, incisor; PM, premolar (absent in rats); M, molar. PeerJ reviewing PDF | (v2013:07:700:1:0:CHECK 23 Aug 2013) R ev ie w in g M an us cr ip t Figure 5 GMM analysis of cranial deformations in Laonastes aengimamus. Plot of the first two principal components from a GMM analysis of 24 cranial landmarks. Cranial reconstructions and thin-plate splines indicate shape changes (x 200) along the first and second principal components. PC1 and PC2 not to same scale. Key: star, unloaded model; squares, incisor bites; diamonds, premolar bites; triangles, M1 bites; circles, M2 bites; lines, M3 bites; blue, IOZM in original position; green, IOZM moved anteriorly; red, IOZM moved posteriorly; orange, IOZM force redistributed between other masticatory muscles. PeerJ reviewing PDF | (v2013:07:700:1:0:CHECK 23 Aug 2013) R ev ie w in g M an us cr ip t Table 1(on next page) Muscle mass (g), mean fibre length (mm) and PCSA (cm2) of masticatory muscles of Laonastes aenigmamus, specimen KY213. abbreviations: zm, zygomaticomandibularis : PeerJ reviewing PDF | (v2013:07:700:1:0:CHECK 23 Aug 2013) R ev ie w in g M an us cr ip t Mass (g) Fibre length (mm) PCSA (cm2) Superficial masseter 0.348 8.73 0.376 Deep masseter 0.235 7.71 0.287 Anterior ZM 0.285 9.16 0.293 Posterior ZM 0.033 3.70 0.084 Infraorbital ZM 0.152 9.71 0.148 Main temporalis 0.047 8.58 0.052 Orbital temporalis 0.019 4.67 0.037 Posterior temporalis 0.017 3.70 0.042 Internal pterygoid 0.188 6.34 0.280 External pterygoid 0.118 4.67 0.239 PeerJ reviewing PDF | (v2013:07:700:1:0:CHECK 23 Aug 2013) R ev ie w in g M an us cr ip t Table 2(on next page) Muscle loads (N) applied to each side of each model in the presence and absence of the IOZM. PeerJ reviewing PDF | (v2013:07:700:1:0:CHECK 23 Aug 2013) R ev ie w in g M an us cr ip t With IOZM Without IOZM Superficial masseter 11.31 12.30 Deep masseter 8.62 9.37 Anterior ZM 8.80 9.57 Posterior ZM 2.50 2.72 Infraorbital ZM 4.44 - Main temporalis 1.56 1.70 Orbital temporalis 1.12 1.21 Posterior temporalis 1.27 1.38 Internal pterygoid 8.40 9.13 External pterygoid 7.15 7.78 Total 55.17 55.17 PeerJ reviewing PDF | (v2013:07:700:1:0:CHECK 23 Aug 2013) R ev ie w in g M an us cr ip t Table 3(on next page) Mean von Mises stress (MPa) across the skull. Abbreviations: I, incisor; PM, premolar; M, molar. PeerJ reviewing PDF | (v2013:07:700:1:0:CHECK 23 Aug 2013) R ev ie w in g M an us cr ip t Original model IOZM origin moved anteriorly IOZM origin moved posteriorly IOZM force redistributed I 2.31 2.27 2.32 2.34 PM 2.14 2.15 2.13 2.15 M1 2.29 2.28 2.23 2.28 M2 2.66 2.59 2.60 2.57 M3 3.04 2.86 2.99 2.95 PeerJ reviewing PDF | (v2013:07:700:1:0:CHECK 23 Aug 2013) R ev ie w in g M an us cr ip t Table 4(on next page) Predicted bite forces (N) across the skull. Abbreviations: I, incisor; PM, premolar; M, molar. PeerJ reviewing PDF | (v2013:07:700:1:0:CHECK 23 Aug 2013) R ev ie w in g M an us cr ip t Original model IOZM origin moved anteriorly IOZM origin moved posteriorly IOZM force redistributed I 29.27 25.74 29.05 26.61 PM 53.74 46.93 53.72 48.80 M1 61.00 54.93 61.70 55.78 M2 68.97 61.92 70.32 62.05 M3 73.93 65.35 74.38 66.61 PeerJ reviewing PDF | (v2013:07:700:1:0:CHECK 23 Aug 2013) R ev ie w in g M an us cr ip t",
+    "v2_text": "materials & methods : results : Figure 2 shows the von Mises stress patterns generated across the skull of L. aenigmamus during biting at the incisor, first molar and third molar. It can be seen that, aside from the biting tooth, the zygomatic arch is the most stressed region of the skull, followed by the orbital region. The rostrum experiences a moderate degree of stress during incisor gnawing, but is unstressed during molar chewing, and the occipital region is unstressed during all bites. From visual inspection of the stress distribution figures it is difficult to determine a great deal of variation between bites at different points along the tooth row, even between incisors and molars. However, by studying the mean von Mises stresses across the skull (Table 3) it can be seen that there are indeed subtle differences between bites on different teeth. In general, overall stress increases as the bite point moves closer to the jaw articulation from the premolar to the third molar. However, the incisor bite does not fit this trend, and shows the greatest mean stress across the skull of all bites, presumably because the rostrum is stressed in incisor bites but not in molar bites. The maximum bite force predicted at each tooth is given in Table 4. As would be expected from simple lever mechanics, bite force increases the closer the bite point (the biting tooth) is to the fulcrum (the TMJ). By dividing the bite forces by total input muscle force, the mechanical efficiency of biting has been calculated and can be compared with previous analyses on other rodents (Cox et al., 2012). It can be seen (Figure 3) that the Laotian rock rat performs well at the incisors and premolars, outcompeting the squirrel and guinea pig, and even matching the rat for efficiency at the incisors. However, compared to the other rodents, the mechanical efficiency tails off towards the distal molars and at the third molar, L. aenigmamus has the least efficient bite. The impact of the IOZM muscle was investigated by altering the position of its origin on the rostrum in VOX-FE. Figure 2 shows the von Mises stress patterns generated across the skull in models with the origin of the IOZM moved anteriorly and posteriorly, as well as in its original position. Despite quite large changes in muscle origins, very few differences can be seen between the models. This situation is confirmed by examining the mean von Mises stresses across the skull (Table 3). It can be seen that moving the IOZM origin anteriorly or posteriorly results in very little change (less than 6%) in mean von Mises stress during both incisor gnawing and molar 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 PeerJ reviewing PDF | (v2013:07:700:0:1:NEW 1 Aug 2013) R ev ie w in g M an us cr ip t biting. Although not presented here, the same lack of difference was found on examination of the principal strains across the skull. Predicted bite force was also affected very little by moving the IOZM origin caudally to the most posterior part of the rostrum. However, Table 4 demonstrates that an anterior shift of the IOZM origin resulted in quite a substantial reduction in bite force (between 10 and 13%). The effect of removing the IOZM muscle (and redistributing its force between the remaining masticatory muscles) can be seen in Figure 2. As when the IOZM origin was moved, few differences could be detected between the von Mises stress patterns by visual inspection alone, and again, this assertion is supported by examination of the mean von Mises stresses across the skull (Table 3). However, it can be seen from Table 4 that the action of the IOZM has a noticeable effect on bite force. It is clear that the presence of the IOZM enables L. aenigmamus to generate a greater bite force than if it were absent, even when the total input muscle force is the same. The reduction in bite force resulting from removal of the IOZM is approximately 10% for bites on both incisors and molars, which is similar to the effect of shifting the origin of this muscle to the anterior part of the rostrum. In order to analyse the subtle differences between the deformation patterns generated by the various load cases described above, a geometric morphometric analysis was performed. Landmark data from the original solved models, the models with the IOZM moved anteriorly and posteriorly, and models with the IOZM removed and its force redistributed, as well as landmarks from the unsolved model, were all subjected to Procrustes superimposition and PCA. The plot of the first two principal components (together comprising over 98% of the variation; PC1, 63.2%; PC2, 35.1%) is shown in Figure 4. The first principal component largely separates the unloaded model from the solved load cases. The point representing the unloaded model (star symbol) is on the far left of the plot whilst the points representing the loaded models are spread out down the right hand side of the diagram. The warped reconstructions indicate that the difference in deformation between the loaded and unloaded models is largely concentrated in the zygomatic region. The incisor bites (squares) are clearly separated from bites on the other teeth. The cheek teeth bites are more closely grouped together, but separable into bites on each of the different teeth and positioned in order along the second principal component from the premolar (diamonds) to the third molar (lines). It can be seen from the reconstructions at the extremes of PC2 that incisor bites tend to deflect the rostrum dorsally (relative to the orbito-temporal region), 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 PeerJ reviewing PDF | (v2013:07:700:0:1:NEW 1 Aug 2013) R ev ie w in g M an us cr ip t but that molar bites lead to a ventral deflection of the rostrum. Within bites on each tooth, the models with the IOZM origin in its original position (blue symbols), moved anteriorly (green symbols) and moved posteriorly (red symbols) all group together closely at a similar distance from the unloaded model. This demonstrates that a similar amount of deformation is occurring in each of these load cases. However, the symbols representing the models with the IOZM muscle force redistributed between the other masticatory muscles (orange symbols), are positioned further from the unloaded model than the other load cases, indicating that even more deformation is occurring in these models. As the displacement of these symbols is along PC1, it can be inferred that the redistribution of the IOZM is leading to greater deformation in the zygomatic region. acknowledgements : We thank Jean-Pierre Hugot for providing access to the specimens and Dominique Adriaens, Loes Brabant and Luc Van Hoorebeke for scanning the specimens at the University of Ghent CT facility (UGCT). We are grateful to Paul O\u2019Higgins and Michael Fagan for access to VOX-FE finite element software. Thanks are due to Laura Fitton for help with model construction, Peter 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 PeerJ reviewing PDF | (v2013:07:700:0:1:NEW 1 Aug 2013) R ev ie w in g M an us cr ip t Bazira for technical support with high-performance computing, and Andrew McIntosh and Thomas P\u00fcschel for assistance with GMM software. discussion : A finite element model of the skull of the Laotian rock rat, Laonastes aenigmamus, was created, loaded and solved. It was shown that the area of the skull experiencing the highest levels of stress was the zygomatic arch, including the zygomatic processes of the maxillary and frontal bones, which is likely to be a result of the large amount of masticatory musculature that attaches directly to this area. Similarly high zygomatic stresses have been noted in other rodents (Cox et al., 2012) as well as in other mammalian groups (Dumont, Piccirillo & Grosse, 2005; Bright & Rayfield, 2011; Dumont et al., 2011). It has been suggested that, in primates, the downward pull of the masseter muscle on the zygomatic arch may be counterbalanced to some degree by the upward pull of a soft tissue structure, namely the temporal fascia (Curtis et al., 2011). Despite many careful dissections (e.g. Baverstock, Jeffery & Cobb, 2013), no temporal fascia has been found in rodents, and this is also true of L. aenigmamus (Hautier & Saksiri, 2009). However, there is a possibility that one, or part of one, of the masticatory muscles may be acting as a counterbalance. Specifically, if part of the temporalis were to run from the temporal fossa to the zygomatic arch it could have this role. This function of the temporalis is at the moment hypothetical and will require more detailed dissections of L. aenigmamus for confirmation. The bite forces predicted in this study demonstrate that the skull of L. aenigmamus can generate bites of 29 N during gnawing and between 53 and 74 N during chewing. By dividing these values by the total input muscle force, the mechanical efficiency of biting was calculated. It was shown that the Laotian rock rat is particularly efficient at incisor biting, having a mechanical efficiency greater than squirrels or guinea pigs, and similar to that of rats (Cox et al., 2012), but are less 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 PeerJ reviewing PDF | (v2013:07:700:0:1:NEW 1 Aug 2013) R ev ie w in g M an us cr ip t efficient at molar bites compared to these three rodent species. This would seem to indicate that L. aenigmamus is well-adapted for gnawing and less so for chewing, a conclusion that at first glance appears somewhat at odds with the suggested diet of this species, which is thought to be largely folivorous (Scopin et al., 2011). However, Onoda et al. (2011) have proposed that the generation of a uniform bite force across the tooth row may be beneficial in the processing of fibrous plant material such as leaves. Viewed in this light, L. aenigmamus is well-adapted to folivory with a difference of just 20 N between its premolar and M3 bites, compared to 30-35 N in squirrels and guinea pigs. One of the more intriguing results of this study is that the presence and position of the IOZM muscle has little effect on the overall stress experienced by the skull of L. aenigmamus. It can be seen from both Figure 2 and Table 3 that neither the pattern of von Mises stresses nor the mean stress across the skull are affected by moving the origin of the IOZM forwards or backwards, or by removing the IOZM altogether (and redistributing its force between the remaining muscles). This is probably because the IOZM is contributing a relatively small proportion of the total muscle force (around 8%). Although the IOZM appears quite a large muscle in lateral view (see Hautier & Saksiri, 2009), it is relatively small compared to the superficial and deep masseters. Moreover, it also possesses the longest muscle fibres of all masticatory muscles, and so its PCSA (and hence muscle force) is comparatively reduced (Table 1). Thus, its ability to impact the overall stress patterns across the skull may be fairly limited. In contrast to the lack of influence on stress distributions, it was found that the IOZM has a strong impact on the bite force produced by L. aenigmamus (Table 4). Specifically, removing the IOZM altogether (and redistributing its force between the remaining masticatory muscles) reduces the bite force generated at all teeth. Similarly, moving the origin of the IOZM to the anterior tip of the rostrum reduces bite force to a similar degree. It is likely that this effect is a result of the wrapping of the IOZM around the zygomatic process of the maxilla, which means that as the IOZM is moved forward, its vector of pull becomes more horizontal, and so less able to generate bite force. On the other hand, moving the origin of the IOZM posteriorly appears to have little effect on bite force. Interestingly, the reduction in bite force resulting from the anterior movement or removal of the IOZM is largely constant along the tooth row \u2013 around 10%. Furthermore, the GMM analysis of deformations (Figure 4) has shown that the absence of the IOZM leads to increased deformations across the skull, again at bites on all teeth. Therefore, it would seem that 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 PeerJ reviewing PDF | (v2013:07:700:0:1:NEW 1 Aug 2013) R ev ie w in g M an us cr ip t the extension of the zygomaticomandibularis on to the rostrum evolved in order to increase bite force (as proposed by Becht, 1953) and to decrease cranial deformation. Given that there is minimal difference in bite force between the models with the IOZM in its original position and those with the origin moved posteriorly, the position of the IOZM origin halfway along the rostrum may simply have coevolved with the lengthening of the rostrum in this species. Alternatively, Herrel et al. (2012) have suggested that the anterior insertion of the IOZM in L. aenigmamus increases the horizontal component of biting, which leads to anterior displacement of the mandible during jaw closing, which may be advantageous for the processing of leaves. Whatever the driving force behind the anterior extension of the IOZM, the results here indicate that the reason that the IOZM does not extend any further along the rostrum, as is seen in some other hystricomorph rodents such as the springhare (Offermans & De Vree, 1989) and capybara (M\u00fcller, 1933), is because this would lead to a reduction in bite force at both the incisors and cheek teeth. The results of this study have provided some important insights into the role of the IOZM muscle in the feeding behaviour of L. aenigmamus. Further investigations into other rodents, particularly other hystricomorphs, will enable us to understand whether the ability of the IOZM to increase bite force is unique to L. aenigmamus or common to all rodents that possess it. If the latter scenario is true, this could provide a selective advantage that may have driven the evolution of the IOZM, and could explain why it has evolved independently in several rodent groups (Ctenohystrica, Dipodidae, Anomaluroidea and Gliridae). However, finite element models can only shed light on static loading, and therefore cannot inform about the dynamic processes to which the IOZM may contribute, such as propalineal (antero-posterior) movements of the lower jaw, jaw opening or the fine control of gnawing. These activities can be addressed with dynamic modelling techniques, such as multibody dynamics analysis (e.g. Jones et al., 2012), and may well provide a fruitful avenue of research in the future. corresponding author: philip g. cox : Address: Centre for Anatomical and Human Sciences, Hull York Medical School, University of Hull, Cottingham Road, Hull, HU6 7RX, UK Telephone: 01904 321744 Email: philip.cox@hyms.ac.uk tables : Table 1. Muscle mass (g), mean fibre length (mm) and PCSA (cm2) of masticatory muscles in Laonastes aenigmamus, specimen KY213. Mass (g) Fibre length (mm) PCSA (cm2) Superficial masseter 0.348 8.73 0.376 Deep masseter 0.235 7.71 0.287 Anterior ZM 0.285 9.16 0.293 Posterior ZM 0.033 3.70 0.084 Infraorbital ZM 0.152 9.71 0.148 Main temporalis 0.047 8.58 0.052 Orbital temporalis 0.019 4.67 0.037 Posterior temporalis 0.017 3.70 0.042 Internal pterygoid 0.188 6.34 0.280 External pterygoid 0.118 4.67 0.239 Abbreviations: ZM, zygomaticomandibularis. Table 2. Muscle loads (N) applied to each side of each model in the presence and absence of the IOZM. 407 408 409 410 411 412 414 415 416 PeerJ reviewing PDF | (v2013:07:700:0:1:NEW 1 Aug 2013) R ev ie w in g M an us cr ip t Original model IOZM origin moved anteriorly IOZM origin moved posteriorly IOZM force redistribut ed I 29.27 25.74 29.05 26.61 PM 53.74 46.93 53.72 48.80 M1 61.00 54.93 61.70 55.78 M2 68.97 61.92 70.32 62.05 M3 73.93 65.35 74.38 66.61 418 419 420 PeerJ reviewing PDF | (v2013:07:700:0:1:NEW 1 Aug 2013) R ev ie w in g M an us cr ip t Figure 1 Landmarks used in GMM analysis of skull deformations. Reconstruction of skull of Laonastes aenigmamus in (A) dorsal, (B) ventral and (C) left lateral view. 1, anteriormost point on internasal suture; 2, midpoint on cranium between anterior roots of zygomatic arch; 3, midpoint between medialmost points on orbital margins; 4, midpoint on skull roof between zygomatic processes of the squamosal; 5, posteriormost point in dorsal midline; 6, midpoint between ventral margins of incisal alveoli; 7, midpoint between anteriormost points of premolars; 8, posteriormost midline point on palate; 9, midpoint between posterior margins of pterygoid flanges; 10, ventralmost point on margin of foramen magnum; 11, anteriormost point on naso-frontal suture; 12, dorsalmost point on incisal alveolar margin; 13, rostralmost point of infraorbital fossa; 14, midpoint between incisor and premolar on ventrolateral rostral margin; 15, midpoint of dorsal margin of infraorbital fossa; 16, midpoint between 15 and 17; 17, anteriormost attachment of zygomatic arch to rostrum; 18; posteriormost point of infraorbital margin; 19, point on ventrolateral margin of zygomatic arch in same coronal plane as midpoint of M1; 20, apex of tubercle on anterior orbital margin; 21, dorsalmost point on orbital margin; 22, ventralmost point on orbital margin; 23, midpoint between 21 and 24; 24, posteriormost point on orbital margin. Landmarks 11-24 recorded on both sides. PeerJ reviewing PDF | (v2013:07:700:0:1:NEW 1 Aug 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:07:700:0:1:NEW 1 Aug 2013) R ev ie w in g M an us cr ip t Figure 2 Predicted distribution of von Mises stresses across the skull of Laonastes aenigmamus. Arrows indicate the biting tooth. Top row, incisor bites; middle row, M1 bites; bottom row, M3 bites. First column, original models; second column, origin of IOZM moved anteriorly to front of rostrum; third column, origin of IOZM moved posteriorly to back of rostrum; fourth column, IOZM muscle force removed and redistributed proportionally between the remaining masticatory muscles. PeerJ reviewing PDF | (v2013:07:700:0:1:NEW 1 Aug 2013) R ev ie w in g M an us cr ip t Figure 3 Mechanical efficiency of biting at each tooth. Predicted from FE models of squirrel, guinea pig, rat and Laotian rock rat skulls. Data for squirrel, guinea pig and rat from Cox et al. (2012). Abbreviations: I, incisor; PM, premolar (absent in rats); M, molar. PeerJ reviewing PDF | (v2013:07:700:0:1:NEW 1 Aug 2013) R ev ie w in g M an us cr ip t Figure 4 GMM analysis of cranial deformations in Laonastes aenigmamus. Plot of the first two principal components from a GMM analysis of 24 cranial landmarks. Cranial reconstructions indicate shape changes (x 100) along the first and second principal components. Key: star, unloaded model; squares, incisor bites; diamonds, premolar bites; triangles, M1 bites; circles, M2 bites; lines, M3 bites; blue, IOZM in original position; green, IOZM moved anteriorly; red, IOZM moved posteriorly; orange, IOZM force redistributed between other masticatory muscles. PeerJ reviewing PDF | (v2013:07:700:0:1:NEW 1 Aug 2013) R ev ie w in g M an us cr ip t model solution and analysis : The finite element model of Laonastes aenigmamus was solved for biting at each tooth along the dental arcade, using VOX-FE. Gnawing was assumed always to be bilateral, as a result of the close apposition of the incisors, whereas chewing at the premolars and molars was modelled unilaterally on the left side. Von Mises stress patterns across the skull and bite forces at the teeth were recorded from the solved models. Following, Cox et al. (2011) and O\u2019Higgins et al. (2011), geometric morphometrics was used to study the deformation patterns across the cranium. Thus, a set of landmarks (3D co-ordinate data) was recorded from each loaded skull, as well as from the original unsolved model. The landmark set was partially based on that in Cox et al. (2011) and is shown in Figure 1. Landmarks were concentrated in the rostral, orbital and zygomatic regions, as these were the areas experiencing the highest strains. The landmarks were then subjected to a geometric morphometric analysis using MorphoJ software (Klingenberg, 2011). This consisted of co-registering the landmarks via Procrustes superimposition followed by a principal components analysis (PCA). Cranial deformations along PC axes were visualised using the EVAN toolbox (http://www.evan-society.org). The results generated by the Laotian rock rat model were compared to those found in other rodents, such as squirrels, guinea pigs and rats (Cox et al., 2012). As it difficult to compare absolute bite force between models of different sizes, the mechanical efficiency of biting was calculated. Mechanical efficiency of biting is the ratio of output bite force to input muscle force, and represents the proportion of muscle force that is translated into bite force i.e. is not lost to deformation of the mandible or generation of joint reaction force at the condyles (Dumont et al., 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 PeerJ reviewing PDF | (v2013:07:700:0:1:NEW 1 Aug 2013) R ev ie w in g M an us cr ip t 2011). As it is a ratio, mechanical efficiency is size-independent and facilitates clearer comparisons between skulls of varying sizes. authors: philip g cox1, joanna kirkham2,3 and anthony herrel4 : 1Centre for Anatomical and Human Sciences, Hull York Medical School, University of Hull, Hull, UK 2Centre for Anatomical and Human Sciences, Hull York Medical School, University of York, York, UK 3College of Medical and Dental Sciences, University of Birmingham, Birmingham, UK 4UMR CNRS/MNHN 7179, M\u00e9canismes adaptatifs: des organismes aux communaut\u00e9s, Paris, France sample and model creation : MicroCT scans of an adult female Laonastes aenigmamus (specimen number KY213), previously obtained for an earlier research project (Herrel et al., 2012), were provided by AH. Voxels were isometric and the voxel resolution was 0.137 mm. Further details of the scanning protocol can be found in Herrel et al. (2012). 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 PeerJ reviewing PDF | (v2013:07:700:0:1:NEW 1 Aug 2013) R ev ie w in g M an us cr ip t Avizo 7.0 (Visualization Sciences Group, Burlington, MA, USA) was used to create a 3D volume reconstruction of the cranium of the Laotian rock rat. The bone and teeth were segmented separately so that different elastic properties could be assigned to them. Within the incisors, the enamel, dentine and pulp were not differentiated, as the scan resolution was not sufficient to distinguish these materials from one another. In addition, varying the material properties of these components has been shown to make little difference to the overall deformation of the skull (Cox et al., 2011). For similar reasons, the periodontal ligament was not included (Wood et al., 2011). The model was converted to an eight-noded FE mesh by direct voxel conversion, using VOX-FE, custom-built FE software (Liu et al., 2012), resulting in a model of 1734787 elements. Material properties, constraints and loads Bone was assigned a Young\u2019s modulus of 17 GPa, and the teeth were given a Young\u2019s modulus of 30 GPa. Both materials were modelled to be linearly elastic and isotropic with a Poisson\u2019s ratio of 0.3. These values were based on previous nano-indentation work on rodents (Cox et al.., 2012) and FE studies on other mammals (Kupczik et al.., 2012). The model was constrained at three locations: the left and right temporo-mandibular joints (TMJ) and the biting tooth. The TMJs were constrained on the ventral surface of the zygomatic process of the squamosal in all three dimensions. However, the biting tooth was only constrained in the dorsoventral axis (i.e. perpendicular to the occlusal plane). The number of nodes constrained at each location varied between 191 and 221. Loads were added to the model to represent the forces generated by the following muscles: superficial masseter; deep masseter; zygomaticomandibularis (anterior, posterior and infraorbital parts); temporalis (main, orbital and posterior parts); internal pterygoid; and external pterygoid. Muscle attachment sites and directions of pull were assigned based on the detailed dissections presented in Hautier & Saksiri (2009). Table 1 gives the muscle mass and mean fibre length of each masticatory muscle, measured from the dissection of specimen KY213. Muscle masses were converted to volumes, assuming a muscle density of 1.0564 gcm-3 (Murphy & Beardsley, 1974), and the physiological cross-sectional area (PCSA) of each muscle (given in Table 1) was calculated by dividing the muscle volume by mean fibre length. Muscle forces (Table 2) were calculated by multiplying PCSAs by an intrinsic muscle stress value of 0.3 Nmm-2 (van Spronsen et al., 1989). In order to investigate the function of the IOZM, alternate versions of the models 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 PeerJ reviewing PDF | (v2013:07:700:0:1:NEW 1 Aug 2013) R ev ie w in g M an us cr ip t were created with the origin of this muscle moved anteriorly and posteriorly, and also with the IOZM omitted completely on each side. In the models with a posteriorly shifted IOZM, the insertion point of the IOZM was kept the same, resulting in a more vertically directed muscle vector. However, in the models with an anteriorly shifted IOZM, the insertion point was moved to reflect the wrapping of the muscle around the zygomatic process of the maxilla, resulting in a highly horizontally directed muscle vector. To facilitate comparisons between models with and without the IOZM whilst retaining the same overall input muscle load, where it was omitted, the force of the IOZM was redistributed across the remaining masticatory muscles. This was done in such a way as to preserve their relative proportions. The muscle forces applied in the absence of the IOZM are given in Table 2.",
+    "url": "https://peerj.com/articles/161/reviews/",
+    "review_1": "Marco Iacoboni \u00b7 Aug 27, 2013 \u00b7 Academic Editor\nACCEPT\nThe addition of Experiment 2 not only clarifies the nature of the results in Experiment 1 but also provides strong evidence of how temporal dynamics are powerful modulators of these effects. An excellent study that will undoubtedly inspire more research on this important topic.",
+    "review_2": "Marco Iacoboni \u00b7 Aug 22, 2013 \u00b7 Academic Editor\nMAJOR REVISIONS\nThe reviewers made a number of important comments that you may want to address in your revision. I also read the paper and have some comments, which partially overlap the comments of the reviewers.\n\nYour paper reports a clever new experimental variation of the automatic imitation task (AIT). You propose that this new task, the controlled imitation task (CIT), can be used in studies \u201cof self-other control processes and more generally, internally generated versus externally triggered actions.\u201d\n\nI think that with regard to the self-other control processes the task lacks an important control condition, that is, non biological cues requiring subjects to override their initially planned action. In this control condition, rather than seeing an action 100 ms after the cue had given subjects instructions on which action to perform, subjects would see either another symbolic or a spatial cue. The RT costs in the incongruent trials of this control task, subtracted to the RT costs in the CIT, would give us a true estimate of self-other control process. Without this control task, the only thing we can conclude is that the whopping RT costs (120 ms!) shown in this study are costs in reprogramming an action. Indeed, I am not even sure one can conclude that the task in its current form can give us an index of \u201cinternally generated versus externally triggered actions\u201d, since the first planned action is also externally cued.\n\nTwo other minor comments:\n\nWhy 100 ms? What happens if you manipulate that time interval? Does the RT cost get bigger, smaller, it disappear completely? These are definitely interesting follow up experimental manipulations.\n\nThere are documented orthogonal spatial s-r effects. I think you should look into that literature and make sure that those effects do no play a role here.",
+    "review_3": "Jonas Kaplan \u00b7 Aug 15, 2013\nBasic reporting\nI find the manuscript to be well-written, clear, and concise. I enjoyed reading it. The reporting is simple and straight-forward.\nExperimental design\nThe experiment as it stands appears to be well-executed, but I do feel that the experiment is lacking an important control which would allow the authors to draw conclusions that they draw about the usefulness of this task for studying self-other issues. In order to say this result has anything to do with imitation, it would be important to use a cue that was not another person's movement. The current results, which show that people were slower to respond when their instruction was changed mid-trial by a hand moving, may be explained rather simply without invoking any explanation relating to imitation: people make slower responses when their instruction is changed mid-movement compared to when they are not. In fact, it\u2019s hard to see how any other result would be obtained regardless of what the interrupting cue was. It is clearly going to take more time to change a response than it is to not change a response.\n\nTherefore, to make the case that this task has any relevance to imitation, the authors would need to show that the imitative cue itself has some special power to interfere with self-inhibition, over and above the simple cost of changing action mid-stream. This concept is briefly discussed in the discussion, but it seems to me that it is crucial to the interpretation that the task involves a conflict between self and other representations, which is the way the manuscript frames the usefulness of this task. It would be a fairly simple control condition to add to the experiment.\n\n(\n-The N is fairly small for a simple behavioral study, although the results are significant nevertheless. I wonder about the impact of the handedness of the participants, given that 3 of the 12 were left handed. It appears that everyone used the right hand to respond and the cues from the figure appear to be a left hand such that participants were doing \"mirror imitation\".\n\n-How is reaction time measured? Is it the time when the key on the keyboard is released?\n)\nValidity of the findings\nAs discussed above, it seems to me that the results don't directly speak to the issue of imitation or self/other representation, and instead may be explained in simple motor terms: changing a target response midstream leads to slower responses compared with not changing it. I don't think the manuscript succeeds in making the case that this novel task is measuring anything about representations of self-other control as promised in the title.\n\nThat said, I like the idea of this task, and if it turns out that the interference effect is related to representations of observed movement it has the potential to be very useful as a complement to the AIT.\nAdditional comments\nNo comments.\nCite this review as\nKaplan J (2013) Peer Review #1 of \"The controlled imitation task: a new paradigm for studying self-other control (v0.1)\". PeerJ https://doi.org/10.7287/peerj.161v0.1/reviews/1",
+    "review_4": "Reviewer 2 \u00b7 Aug 6, 2013\nBasic reporting\nThe authors introduce a novel task, based on the automatic imitation task, which aims at testing for prioritization of imitative responses. The basic rationale of the task is a useful addition, and if constructed well is likely to spawn a large number of studies using this addition, or extensions thereof.\n- It would be helpful to elaborate in more detail why the effects in the AIT are not explained by other processes, such as preparation and consequent competition resolution. In other words, what is so special about this specific process, or is there a unifying account \u2013 intuitively, having dedicated but separated systems for imitation and selection seems odd. For example, to which degree can prior expectation explain the reviewed results \u2013 participants have to respond against an expectation? It is hardly surprising that action representations with higher prior probability will be faster to express than unlikely ones, irrespective of the type of information on which such an expectation is build? See also my point below regarding the CIT. I also note that I am not an expert on action observation or imitation, so the difference may be obvious and thus easy to point out.\n- As far as I understand, the CIT is essentially a countermanding task, with matched probabilities. Other cousins include reprogramming tasks (cf work by Rushworth) and stopping tasks. I can see how the AIT is incomplete, and the CIT is therefore a useful addition, but embedding the task into the vast literature on related inhibition tasks would seem useful. This point is essential as it concerns the construct validity of the task, and this reviewer is not entirely sure the task divorces itself from the various paradigms in this literature.\n- It seems to me that the authors basically aim to extend the AIT to become a fully factorial task. Instead of being complementary, this author wonders if they are indeed both necessary, i.e. accurate inference requires both to be tested. One may be tempted to attribute any outcome to \u2018other\u2019 vs \u2018self\u2019 generated movements, but this essentially requires comparison of both, and neither task achieves that alone. Viewed in this way, the authors basically add another factor to the AIT, which distinguishes reference frame. That may or may not be seen as a novel task, but in any case advocating the use of the CIT as standalone seems to suffer from the same limitations as using the AIT alone, i.e. poor control for \u201creference frame\u201d.\n- Then again, having said the above, comparability between the AIT and CIT is somewhat limited because the temporal structure is different. One may deem this unlikely to explain any observed differences but in terms of experimental design and control this seems to be a problem. It would be useful to device a novel task in which the temporal structure in both tasks is matched\n- Please report measures of effect size\n- tDCS is a terrible method for anatomical inference. Such suggestions should be avoided as they permeate the field in unhealthy and incorrect ways. More generally, anatomical predictions would seem only the second step. First, what about the novel hypotheses concerning processing of observed actions that can now be addressed? As per my comment above, it would seem also that any anatomical prediction would have to include the extensive work on conflict, inhibition etc, and this reviewer wonders whether the current task would reveal anything surprising here.\n- Related, if the authors want to advocate the use of this task, it may be useful to anticipate modifications. One could vary the likelihood of incongruent/congruent trials, for example, which may help to address the previous issue about preparation/prior expectation. The authors use a fixed interval of 100ms between cue and target stimulus, which may not be optimal (nb. How was this interval chosen?). In any case, it may be useful to explore the temporal nature of the process, by varying this interval similar to SSRT tasks. Note also that there is debate about the nature of these processes, e.g. whether the selection and stop process occur in parallel or not, and what the dynamics of this process are.\nExperimental design\nsee above\nValidity of the findings\nsee above\nAdditional comments\nsee above\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"The controlled imitation task: a new paradigm for studying self-other control (v0.1)\". PeerJ https://doi.org/10.7287/peerj.161v0.1/reviews/2",
+    "pdf_1": "https://peerj.com/articles/161v0.2/submission",
+    "pdf_2": "https://peerj.com/articles/161v0.1/submission",
+    "review_5": "Hugo Th\u00e9oret \u00b7 Aug 1, 2013\nBasic reporting\nNo comments\nExperimental design\nNo comments\nValidity of the findings\nNo comments\nAdditional comments\nIn this paper, the authors present a novel task that requires participants to suppress self-activated motor reponses following an initial cue. The authors argue that this task, which complement the well known \u00ab automatic imitation task \u00bb, may be used to study self-other representations.\n\nThe paper presents a good argument for the development of such a task, and the protocol and task appear well thought. I do not have concerns regarding the manuscript as it stands. I wonder, however, how performance on the CIT correlates with performance on the classical AIT\u2026 It would have been quite interesting to have participants perform both tasks\u2026\nCite this review as\nTh\u00e9oret H (2013) Peer Review #3 of \"The controlled imitation task: a new paradigm for studying self-other control (v0.1)\". PeerJ https://doi.org/10.7287/peerj.161v0.1/reviews/3",
+    "all_reviews": "Review 1: Marco Iacoboni \u00b7 Aug 27, 2013 \u00b7 Academic Editor\nACCEPT\nThe addition of Experiment 2 not only clarifies the nature of the results in Experiment 1 but also provides strong evidence of how temporal dynamics are powerful modulators of these effects. An excellent study that will undoubtedly inspire more research on this important topic.\nReview 2: Marco Iacoboni \u00b7 Aug 22, 2013 \u00b7 Academic Editor\nMAJOR REVISIONS\nThe reviewers made a number of important comments that you may want to address in your revision. I also read the paper and have some comments, which partially overlap the comments of the reviewers.\n\nYour paper reports a clever new experimental variation of the automatic imitation task (AIT). You propose that this new task, the controlled imitation task (CIT), can be used in studies \u201cof self-other control processes and more generally, internally generated versus externally triggered actions.\u201d\n\nI think that with regard to the self-other control processes the task lacks an important control condition, that is, non biological cues requiring subjects to override their initially planned action. In this control condition, rather than seeing an action 100 ms after the cue had given subjects instructions on which action to perform, subjects would see either another symbolic or a spatial cue. The RT costs in the incongruent trials of this control task, subtracted to the RT costs in the CIT, would give us a true estimate of self-other control process. Without this control task, the only thing we can conclude is that the whopping RT costs (120 ms!) shown in this study are costs in reprogramming an action. Indeed, I am not even sure one can conclude that the task in its current form can give us an index of \u201cinternally generated versus externally triggered actions\u201d, since the first planned action is also externally cued.\n\nTwo other minor comments:\n\nWhy 100 ms? What happens if you manipulate that time interval? Does the RT cost get bigger, smaller, it disappear completely? These are definitely interesting follow up experimental manipulations.\n\nThere are documented orthogonal spatial s-r effects. I think you should look into that literature and make sure that those effects do no play a role here.\nReview 3: Jonas Kaplan \u00b7 Aug 15, 2013\nBasic reporting\nI find the manuscript to be well-written, clear, and concise. I enjoyed reading it. The reporting is simple and straight-forward.\nExperimental design\nThe experiment as it stands appears to be well-executed, but I do feel that the experiment is lacking an important control which would allow the authors to draw conclusions that they draw about the usefulness of this task for studying self-other issues. In order to say this result has anything to do with imitation, it would be important to use a cue that was not another person's movement. The current results, which show that people were slower to respond when their instruction was changed mid-trial by a hand moving, may be explained rather simply without invoking any explanation relating to imitation: people make slower responses when their instruction is changed mid-movement compared to when they are not. In fact, it\u2019s hard to see how any other result would be obtained regardless of what the interrupting cue was. It is clearly going to take more time to change a response than it is to not change a response.\n\nTherefore, to make the case that this task has any relevance to imitation, the authors would need to show that the imitative cue itself has some special power to interfere with self-inhibition, over and above the simple cost of changing action mid-stream. This concept is briefly discussed in the discussion, but it seems to me that it is crucial to the interpretation that the task involves a conflict between self and other representations, which is the way the manuscript frames the usefulness of this task. It would be a fairly simple control condition to add to the experiment.\n\n(\n-The N is fairly small for a simple behavioral study, although the results are significant nevertheless. I wonder about the impact of the handedness of the participants, given that 3 of the 12 were left handed. It appears that everyone used the right hand to respond and the cues from the figure appear to be a left hand such that participants were doing \"mirror imitation\".\n\n-How is reaction time measured? Is it the time when the key on the keyboard is released?\n)\nValidity of the findings\nAs discussed above, it seems to me that the results don't directly speak to the issue of imitation or self/other representation, and instead may be explained in simple motor terms: changing a target response midstream leads to slower responses compared with not changing it. I don't think the manuscript succeeds in making the case that this novel task is measuring anything about representations of self-other control as promised in the title.\n\nThat said, I like the idea of this task, and if it turns out that the interference effect is related to representations of observed movement it has the potential to be very useful as a complement to the AIT.\nAdditional comments\nNo comments.\nCite this review as\nKaplan J (2013) Peer Review #1 of \"The controlled imitation task: a new paradigm for studying self-other control (v0.1)\". PeerJ https://doi.org/10.7287/peerj.161v0.1/reviews/1\nReview 4: Reviewer 2 \u00b7 Aug 6, 2013\nBasic reporting\nThe authors introduce a novel task, based on the automatic imitation task, which aims at testing for prioritization of imitative responses. The basic rationale of the task is a useful addition, and if constructed well is likely to spawn a large number of studies using this addition, or extensions thereof.\n- It would be helpful to elaborate in more detail why the effects in the AIT are not explained by other processes, such as preparation and consequent competition resolution. In other words, what is so special about this specific process, or is there a unifying account \u2013 intuitively, having dedicated but separated systems for imitation and selection seems odd. For example, to which degree can prior expectation explain the reviewed results \u2013 participants have to respond against an expectation? It is hardly surprising that action representations with higher prior probability will be faster to express than unlikely ones, irrespective of the type of information on which such an expectation is build? See also my point below regarding the CIT. I also note that I am not an expert on action observation or imitation, so the difference may be obvious and thus easy to point out.\n- As far as I understand, the CIT is essentially a countermanding task, with matched probabilities. Other cousins include reprogramming tasks (cf work by Rushworth) and stopping tasks. I can see how the AIT is incomplete, and the CIT is therefore a useful addition, but embedding the task into the vast literature on related inhibition tasks would seem useful. This point is essential as it concerns the construct validity of the task, and this reviewer is not entirely sure the task divorces itself from the various paradigms in this literature.\n- It seems to me that the authors basically aim to extend the AIT to become a fully factorial task. Instead of being complementary, this author wonders if they are indeed both necessary, i.e. accurate inference requires both to be tested. One may be tempted to attribute any outcome to \u2018other\u2019 vs \u2018self\u2019 generated movements, but this essentially requires comparison of both, and neither task achieves that alone. Viewed in this way, the authors basically add another factor to the AIT, which distinguishes reference frame. That may or may not be seen as a novel task, but in any case advocating the use of the CIT as standalone seems to suffer from the same limitations as using the AIT alone, i.e. poor control for \u201creference frame\u201d.\n- Then again, having said the above, comparability between the AIT and CIT is somewhat limited because the temporal structure is different. One may deem this unlikely to explain any observed differences but in terms of experimental design and control this seems to be a problem. It would be useful to device a novel task in which the temporal structure in both tasks is matched\n- Please report measures of effect size\n- tDCS is a terrible method for anatomical inference. Such suggestions should be avoided as they permeate the field in unhealthy and incorrect ways. More generally, anatomical predictions would seem only the second step. First, what about the novel hypotheses concerning processing of observed actions that can now be addressed? As per my comment above, it would seem also that any anatomical prediction would have to include the extensive work on conflict, inhibition etc, and this reviewer wonders whether the current task would reveal anything surprising here.\n- Related, if the authors want to advocate the use of this task, it may be useful to anticipate modifications. One could vary the likelihood of incongruent/congruent trials, for example, which may help to address the previous issue about preparation/prior expectation. The authors use a fixed interval of 100ms between cue and target stimulus, which may not be optimal (nb. How was this interval chosen?). In any case, it may be useful to explore the temporal nature of the process, by varying this interval similar to SSRT tasks. Note also that there is debate about the nature of these processes, e.g. whether the selection and stop process occur in parallel or not, and what the dynamics of this process are.\nExperimental design\nsee above\nValidity of the findings\nsee above\nAdditional comments\nsee above\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"The controlled imitation task: a new paradigm for studying self-other control (v0.1)\". PeerJ https://doi.org/10.7287/peerj.161v0.1/reviews/2\nReview 5: Hugo Th\u00e9oret \u00b7 Aug 1, 2013\nBasic reporting\nNo comments\nExperimental design\nNo comments\nValidity of the findings\nNo comments\nAdditional comments\nIn this paper, the authors present a novel task that requires participants to suppress self-activated motor reponses following an initial cue. The authors argue that this task, which complement the well known \u00ab automatic imitation task \u00bb, may be used to study self-other representations.\n\nThe paper presents a good argument for the development of such a task, and the protocol and task appear well thought. I do not have concerns regarding the manuscript as it stands. I wonder, however, how performance on the CIT correlates with performance on the classical AIT\u2026 It would have been quite interesting to have participants perform both tasks\u2026\nCite this review as\nTh\u00e9oret H (2013) Peer Review #3 of \"The controlled imitation task: a new paradigm for studying self-other control (v0.1)\". PeerJ https://doi.org/10.7287/peerj.161v0.1/reviews/3\nReview 6: \n",
+    "LLM_extracted_review": "['Review 1: The addition of Experiment 2 clarifies the results in Experiment 1 and provides strong evidence of how temporal dynamics modulate these effects. An excellent study that will inspire more research on this important topic.\\n\\nReview 2: The task lacks an important control condition for self-other control processes. Without this control task, the RT costs observed may only reflect reprogramming an action. The manipulation of the 100 ms interval and the consideration of orthogonal spatial s-r effects are suggested for further exploration.\\n\\nReview 3: The experiment is well-executed but lacks a control that would allow conclusions about imitation. The results may simply reflect slower responses when instructions change mid-trial, rather than any specific imitation effect. The small sample size and handedness of participants could impact results. The measurement of reaction time needs clarification.\\n\\nReview 4: The rationale of the task is useful, but the authors should elaborate on why AIT effects are not explained by other processes. The CIT may be similar to other inhibition tasks, and its construct validity is questioned. The temporal structure of both tasks should be matched for better comparison. Effect size measures should be reported, and anatomical inferences should be approached cautiously.\\n\\nReview 5: The authors present a well-thought-out task for studying self-other representations. However, it would be interesting to see how performance on the CIT correlates with the classical AIT.']"
+}

peerj_json_files/PeerJ_Json_118.json ADDED Viewed

	@@ -0,0 +1,16 @@

+{
+    "v1_Abstract": "2 3",
+    "v2_Abstract": "2 3",
+    "v1_text": "stimuli & apparatus : Experimental stimuli & apparatus were similar to experiment 1, with two exceptions: i) black dots were superimposed over the finger-nails of each finger, and ii) on trials where the finger or dot moved, the movement occurred at once, rather than containing intermediate movement stages (cf. Catmur & Heyes, 2011; Press, Bird, Flach, & Heyes, 2005)(Figure 3). On trials where the fingers moved, the dots remained in position; whereas on trials where the dots moved the fingers remained in position. The final locations of the dots in the spatial control conditions were positioned at the final locations of the finger movements in the imitative conditions. procedure : The experiment took place in a dimly lit cubicle. Participants were seated at the computer, told they would be holding down the v and b keys of the keyboard with their right index and middle fingers throughout the experiment, and would need to execute an index or middle finger lift as fast and accurately as possible when the number 1 or 2 was presented onscreen, respectively. Responses were made by releasing the v or b keys on a standard Lenovo keyboard. Further, participants were told: \u201cif the hand onscreen moves, cancel your response and imitate what you see regardless of the movement you were planning to make.\u201d The study began with a practice session containing 24 trials, which observed by the experimenter to ensure that participants understood the task. If the participant made more than 5 mistakes, the practice section was re-run, which only occurred for one participant in the sample. The experiment proper was divided into six blocks of 32 trials, each containing four congruent index finger trials, four congruent middle finger trials, four incongruent index finger trials, four incongruent middle finger trials, eight baseline index finger trials, and eight baseline middle finger trials. Each trial began with a still image of the hand on screen (300, 400, 500, 600, 700, or 800ms), then the response cue was presented, and after a delay period (50, 70, or 90ms) the hand either moved or stayed still (see Stimuli & Apparatus for a description of the hand movement). Results Data Exclusion As described in the Methods section, the baseline condition was in place purely to ensure that participants actually prepared the cued response on each trial, and was excluded from further analysis. Reaction time data for index and middle finger actions were collapsed and responses that were 3 SD\u2019s above or below the mean within each experimental condition (i.e. congruent, or 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 PeerJ reviewing PDF | (v2013:07:696:1:0:NEW 23 Aug 2013) R ev ie w in g M an us cr ip t incongruent) were excluded from statistical analysis, resulting in the removal of 0.39% of the data. The remaining congruent and incongruent trial data was then subjected to inferential statistical analysis. The CIT and AIT each contained five experimental conditions randomized within blocks: congruent dot, incongruent dot, congruent finger, incongruent finger, and baseline (Figure 3). The trial events in the CIT were nearly identical to experiment 1, except experiment 2 contained a longer neutral stimulus and a shorter blue screen. Participants completed a practice CIT block (32 trials: 4 congruent dot, 4 incongruent dot, 4 congruent finger, 4 incongruent finger, 16 baseline) 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 PeerJ reviewing PDF | (v2013:07:696:1:0:NEW 23 Aug 2013) R ev ie w in g M an us cr ip t followed by experimental CIT blocks (4 blocks x 48 trials) that contained a total of 24 trials per operant (i.e. congruent, incongruent) condition. Akin to experiment 1, on the CIT blocks participants were instructed: \u201cif the hand or the dot onscreen moves, cancel your response and match what you see regardless of the movement you were planning to make.\u201d Participants who made > 5 errors during the practice block in the CIT (n=4) performed the practice twice. In experiment 2, participants completed a practice AIT block and four experimental AIT blocks containing the same randomized trial type distributions as the CIT. The order of presentation for the two tasks was counterbalanced across the sample. Prior to the AIT, participants were instructed to perform index and middle finger lifts in response to the numbers 1 or 2, respectively, and to execute that plan \u201cno matter what the hand or dot onscreen does.\u201d Thus, overall the stimuli and task demands required for the AIT and CIT were identical, except in the former enforcing self was required to inhibit other, whereas in the latter enforcing other was required to imitate and cancel a planned response. Results Data Exclusion As in experiment 1, the baseline condition was in place to ensure that participants expected to perform the cued response on the majority of trials, and was excluded from further analysis. Within-subjects data exclusion was identical to the criterion set in experiment 1, resulting in the removal of 0.33% of all data. Further, a between-subjects criterion of 3 SD away from the mean was violated by one participant\u2019s mean RT in the congruent finger condition, and this participant was removed from the dataset (leaving n=15). It is worthwhile to note that the subsequently reported inferential results were identical with or without the outlying participant. CIT and AIT RT Analyses 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 PeerJ reviewing PDF | (v2013:07:696:1:0:NEW 23 Aug 2013) R ev ie w in g M an us cr ip t Mean correct response times on the CIT blocks were entered into a 2 (stimulus type: dot, finger) x 2 (congruence: dot, finger) repeated measures ANOVA (rmANOVA). There was no effect of stimulus type (p=.14), but the effect of congruence was highly significant (F1,14=80.95, p<.001, \u03b72=.85). Most importantly, the CIT RT analysis yielded a significant interaction between stimulus type and congruence (F1,14=6.00, p<.05, \u03b72=.30). To elucidate the interaction, interference effects were calculated for the dot and finger conditions, which were then entered into a rmANOVA with one factor (stimulus type: dot, finger). The mean interference effect for the spatially-matched dot control (M=74, SD=34) was significantly smaller than the interference effect for the imitation condition (M=98, SD=48; F1,14=5.93, p<.05, \u03b72=.30). Next, mean correct RT\u2019s on the AIT blocks were entered into a 2 (stimulus type: dot, finger) x 2 (congruence: dot, finger) rmANOVA. Neither of the main effects were significant (ps>.1), but importantly the interaction between stimulus type and congruence was significant on the AIT (F1,14=8.21, p<.05, \u03b72=.37). Akin to the CIT analysis, the interference effects on the AIT were computed and entered into a rmANOVA with one factor (stimulus type: dot, finger). Again, akin to the CIT, the dot control (M=7, SD=38) had a much smaller effect than the imitative condition (M=30, SD=49; F1,14=8.23, p<.05, \u03b72=.37) in the AIT context. The lack of a main effect of congruence in the AIT was surprising, given the fact that previous AIT studies have found a main effect of congruence even when a similar spatiallymatched control condition was included (Cook & Bird, 2011, 2012). One possible reason for this discrepancy could be that the order of the two tasks (i.e. CIT then AIT vs. AIT then CIT) confounded congruence in the AIT. To isolate any order effects, the 2x2 rmANOVA\u2019s described for the CIT and AIT were re-run as analyses of covariance (rmANCOVA), regressing out the task 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 PeerJ reviewing PDF | (v2013:07:696:1:0:NEW 23 Aug 2013) R ev ie w in g M an us cr ip t order term. Interestingly, the rmANCOVA\u2019s for the CIT (p=.19), and the AIT (p=.56), no longer contained significant interaction effects. Participants who performed the AIT before the CIT showed a significant interaction on the CIT, with dots (M=78, SD=28) having a smaller interference effect than fingers (M=110, SD=54; F1,6=6.06, p<.05, \u03b72=.50), whereas participants who performed the CIT immediately did not show a stimulus type preference (p=.29). Conversely, participants who performed the CIT first showed a significant stimulus type preference on the AIT, with dots (M=17, SD=46) having a smaller effect than fingers (M=42, SD=53; F1,7=6.39, p<.05, \u03b72=.48), and participants who performed the AIT immediately showing no difference as a function of stimulus type (p=.18). Discussion After documenting an large CIT interference effect in experiment 1, we conducted experiment 2 in order to replicate the CIT, and to differentiate controlled imitation from response modification by including a non-social cue condition. The results from experiment 2 demonstrate again that the size of the RT cost associated with cancelling a self-related movement in favor of imitation is quite large (98ms), and that it is greater than the cost in a well-matched spatial-cue condition (dots: 74ms). The finding of a greater interference effect to the finger stimulus in both the CIT and the AIT is likely due to the greater degree of overlap between the conflicting representations relative to the moving dot conditions. Whereas the spatial location and movement affordances are matched between the finger and dot stimuli, the former contains a topographically congruent or incongruent movement by a model that is not present in the latter. Without saying anything about their ontogeny (Cooper et al., 2012), the present finding is consistent with suggestions that, once developed, imitative compatibility reflects a form of stimulus-response compatibility that is unique, or at least more pronounced, than non-imitative 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 PeerJ reviewing PDF | (v2013:07:696:1:0:NEW 23 Aug 2013) R ev ie w in g M an us cr ip t spatial-cue-driven effects (Catmur & Heyes, 2011). Further, we suggest that experiment 2 utilizes a useful control condition for the CIT, which future studies attempting to manipulate imitative self-other control should include to differentiate controlled imitation from general response modification processes. General Discussion In two experiments, we report a novel task requiring the online control of competing self- and other-generated motor representations. We have called this task the controlled imitation task (CIT), and suggest that it is a useful complementary task to the well-studied AIT. On each experimental trial, participants had to prepare a cued response, but on 50% of trials they were required to suppress this response and instead prioritize the production of an imitative response that was either congruent or incongruent with the cue. The robust result is that participants were significantly better at imitating an observed movement when it matched the response cue, relative to when it did not match the cue (experiment 1), and that this effect was reliable and of a much larger magnitude than the traditional AIT (experiment 2). We suggest the CIT provides a valuable method for examining the influence of the self with respect to self-other control in the imitative context. To the extent that the production of a symbolically-cued response reflects a task specific self-activated process, the incongruent condition in the CIT can be thought of as requiring suppression of this process in favor of producing an imitative response. Thus, the present nicely complement typical AIT data: in the AIT interference effects reflect the cost of inhibiting an other-activated motor representation in 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 PeerJ reviewing PDF | (v2013:07:696:1:0:NEW 23 Aug 2013) R ev ie w in g M an us cr ip t favor of a cued response, whereas in the CIT interference effects reflect the cost of inhibiting a self-activated motor representation in favor of imitation. Experiment 2 enables us to suggest not only that controlled and automatic imitation are more pronounced than general response modification and inhibition, respectively, but also that this biological preference is driven by recent experience. Typical demonstrations of biologicalspecificity assume they are present a priori (Kanwisher, McDermott, & Chun, 1997; Kilner, Paulignan, & Blakemore, 2003), but biological-specificity in the present experiment varied as a function of task order. Participants who engaged in the CIT initially, showed a finger-preference on the subsequent AIT, and participants who did the AIT first showed this preference on the subsequent CIT, but neither group showed a stimulus preference during the initial task. This suggests that inhibiting imitation at time 1 (AIT) makes it more difficult to perform imitation at time 2 (CIT), which was not true for the dot control condition. Conversely, performing imitation at time 1 (CIT) makes it harder to inhibit imitation at time 2 (AIT), which again was not the case for the dot control condition. Thus, it appears the shared self-other representational system that is invoked to explain the traditional AIT is also engaged by the CIT. Furthermore, this system seems to be strongly affected by recent experience (Catmur, Walsh, & Heyes, 2007; Heyes, Bird, Johnson, & Haggard, 2005; Hogeveen & Obhi, 2012). Beyond the relationship between the CIT and AIT, there are several interesting issues and possibilities for future work that could be studied with the CIT. Specifically, in light of a recent training study designed to improve self-other control (Santiesteban, White, et al., 2012), the CIT provides an excellent and quite well matched experimental paradigm with which to contrast AIT, and investigate task-appropriate shifts in processing towards self or other in the motor domain. Their data demonstrate that imitation-inhibition training on day 1 suppresses the automatic 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 PeerJ reviewing PDF | (v2013:07:696:1:0:NEW 23 Aug 2013) R ev ie w in g M an us cr ip t tendency to imitate on day 2, in order to facilitate task performance which necessitates production of the cued response. Whereas the size of the AIT interference effect was reduced after training, accuracy on a visual perspective-taking task was improved, suggesting that processing of the other person was enhanced or inhibited in a task-appropriate manner (Santiesteban, White, et al., 2012). Despite the impressive nature of this finding, the CIT provides a closer other-centric homologue for AIT than visual perspective taking. Thus, if counter-imitation training has a functional impact on task dependent control of self- and other-generated motor representations, we would expect to observe an increased ability to suppress the self-activated motor representation in the CIT. Thus, the control process in the CIT, at least in the incongruent trials, requires the suppression of a self-related representation whereas the control process in the standard AIT requires the suppression of an other-related representation. Seeing a pattern of results that is consistent with suppression of self-related activity in the CIT combined with suppression of other-related activity in the AIT would bolster the notion that the counter-imitation training improves task specific control of self-other processing. In addition to comparing the AIT and CIT directly after a self-other control manipulation, the extant difference between the two tasks could be used to index individual differences in selfrelated bias. On their own, the size of the CIT or AIT interference effects may reflect self-other distinction, control, or both. However, by subtracting the AIT from the CIT, experimenters could quantify the extent to which self-other control is biased in favor of the self, which could then be related to personality variables hypothesized to relate to self-other control. It remains for future work to investigate this suggestion further. In the non-social context, another very interesting aspect of the CIT is that it can be clearly contrasted with the truncation paradigm, in which internal generation of action is pitted 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 PeerJ reviewing PDF | (v2013:07:696:1:0:NEW 23 Aug 2013) R ev ie w in g M an us cr ip t against external cueing of action in non-imitative contexts (Obhi & Haggard, 2004; Obhi, Matkovich, & Chen, 2009; Obhi et al., 2008). In particular, in the truncation paradigm participants begin the trial preparing to make an index finger movement at a time of their own choosing, but are interrupted on a portion of trials by a tone prompting production of the same movement. That, is, in the truncation paradigm, participants prepare to make a self-paced action and are interrupted with a non-social external cue to produce the same action or a different action, whereas in the current study the external cue to action is a movement by another person. Given the dissociation between the ability to inhibit imitation in the AIT and the inhibition of other overlearned responses (Brass et al., 2005; Spengler, von Cramon, et al., 2010), it would be interesting to determine whether the CIT and action modification in the truncation paradigm depend on similar or specialized neural substrates. Implementation of this line of enquiry would necessitate modification to the CIT to include a self-paced action, as opposed to a cued action. Such experiments could be useful for understanding the domain-generality or domain-specificity of the functional mechanisms underlying self-other control. In summary, in two experiments we introduce a new paradigm that we have termed the \u201ccontrolled imitation task\u201d. Using an identical stimulus set to the established AIT, the CIT yields robust interference and facilitation effects that appear to shed light on self-other control processes, and that specifically tap into the capacity to suppress, or harness self-activated motor representations to facilitate fast production of an other-activated response (cf. Brass et al., 2009). We have outlined just two possible areas of inquiry where the CIT could a useful paradigm; selfother control, and the more general exploration of internally versus externally triggered action generation. It remains for future work to consider exactly how this paradigm can be employed to further understanding in these areas of study. 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 PeerJ reviewing PDF | (v2013:07:696:1:0:NEW 23 Aug 2013) R ev ie w in g M an us cr ip t References Bertenthal, B. I., & Scheutz, M. (2013). In praise of a model but not its conclusions: commentary on Cooper, Catmur, and Heyes (2012). Cognitive Science, 37(4), 631\u201341; discussion 642\u20135. doi:10.1111/cogs.12039 Brass, M., Bekkering, H., Wohlschl\u00e4ger, A., & Prinz, W. (2000). Compatibility between observed and executed finger movements: Comparing symbolic, spatial, and imitative cues. Brain and Cognition, 44(2), 124\u2013143. doi:10.1006/brcg.2000.1225 Brass, M., Derrfuss, J., & von Cramon, D. Y. (2005). The inhibition of imitative and overlearned responses: A functional double dissociation. Neuropsychologia, 43(1), 89\u201398. doi:10.1016/j.neuropsychologia.2004.06.018 Brass, M., Ruby, P., & Spengler, S. (2009). Inhibition of imitative behaviour and social cognition. Philosophical Transactions of the Royal Society of London. Series B, Biological sciences, 364(1528), 2359\u20132367. doi:10.1098/rstb.2009.0066 Brass, M., Zysset, S., & von Cramon, Y. (2001). The inhibition of imitative response tendencies. NeuroImage, 14, 1416\u20131423. Catmur, C., & Heyes, C. (2011). Time course analyses confirm independence of imitative and spatial compatibility. Journal of Experimental Psychology: Human Perception and Performance, 37(2), 409\u2013421. doi:10.1037/a0019325 Catmur, Caroline, Walsh, V., & Heyes, C. (2007). Sensorimotor learning configures the human mirror system. Current Biology, 17(17), 1527\u201331. doi:10.1016/j.cub.2007.08.006 Cavallo, A., Heyes, C., Becchio, C., Bird, G., & Catmur, C. (2013). Timecourse of mirror and counter-mirror effects measured with transcranial magnetic stimulation. Social Cognitive and Affective Neuroscience, in press. doi:10.1093/scan/nst085 Cook, J. L., & Bird, G. (2011). Social attitudes differentially modulate imitation in adolescents and adults. Experimental brain research. Experimentelle Hirnforschung. Exp\u00e9rimentation c\u00e9r\u00e9brale, 211(3-4), 601\u201312. doi:10.1007/s00221-011-2584-4 Cook, J. L., & Bird, G. (2012). Atypical social modulation of imitation in autism spectrum conditions. Journal of Autism and Developmental Disorders, 42(6), 1045\u20131051. doi:10.1007/s10803-011-1341-7 Cooper, R. P., Catmur, C., & Heyes, C. (2012). Are automatic imitation and spatial compatibility mediated by different processes? Cognitive Science, 37(4), 605\u2013630. doi:10.1111/j.15516709.2012.01252.x Cross, K. A., Torrisi, S., Losin, E. A. R., & Iacoboni, M. (2013). Controlling automatic imitative tendencies: Interactions between mirror neuron and cognitive control systems. NeuroImage, 83, 493\u2013504. doi:10.1016/j.neuroimage.2013.06.060 Fadiga, L., Fogassi, L., Pavesi, G., & Rizzolatti, R. (1995). Motor facilitation during action observation: A magnetic stimulation study. Journal of Neurophysiology, 73(6), 2608\u20132611. Heyes, C. (2011). Automatic imitation. Psychological Bulletin, 137(3), 463\u201383. doi:10.1037/a0022288 Heyes, C., Bird, G., Johnson, H., & Haggard, P. (2005). Experience modulates automatic imitation. Brain Research, 22(2), 233\u2013240. doi:10.1016/j.cogbrainres.2004.09.009 Hogeveen, J., & Obhi, S. S. (2011). Altogether now: Activating interdependent self-construal induces hypermotor resonance. Cognitive Neuroscience, 2(2), 74\u201382. doi:10.1080/17588928.2010.533164 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 PeerJ reviewing PDF | (v2013:07:696:1:0:NEW 23 Aug 2013) R ev ie w in g M an us cr ip t Hogeveen, J., & Obhi, S. S. (2012). Social interaction enhances motor resonance for observed human actions. The Journal of Neuroscience, 32(17), 5984\u20135989. doi:10.1523/JNEUROSCI.5938-11.2012 Hogeveen, J., & Obhi, S. S. (2013). Automatic imitation is automatic, but less so for narcissists. Experimental brain research. Experimentelle Hirnforschung. Exp\u00e9rimentation c\u00e9r\u00e9brale, 224(4), 613\u201321. doi:10.1007/s00221-012-3339-6 Kanwisher, N., McDermott, J., & Chun, M. M. (1997). The Fusiform Face Area : A Module in Human Extrastriate Cortex Specialized for Face Perception. The Journal of Neuroscience, 17(11), 4302\u20134311. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/9151747 Kilner, J., Paulignan, Y., & Blakemore, S. (2003). An interference effect of observed biological movement on action. Current Biology, 13, 1274\u20131277. doi:10.1016/S Obhi, S. S., & Haggard, P. (2004). Internally generated and externally triggered actions are physically distinct and independently controlled. Experimental Brain Research., 156(4), 518\u2013523. doi:10.1007/s00221-004-1911-4 Obhi, S. S., & Hogeveen, J. (2010). Incidental action observation modulates muscle activity. Experimental Brain Research, 203(2), 427\u2013435. doi:10.1007/s00221-010-2253-z Obhi, S. S., Hogeveen, J., Giacomin, M., & Jordan, C. H. (n.d.). Automatic imitation is reduced in narcissists. Journal of Experimental Psychology: Human Perception and Performance. Obhi, S. S., Matkovich, S., & Chen, R. (2009). Changing the \u201cwhen\u201d and \u201cwhat\u201d of intended actions. Journal of Neurophysiology, 102(5), 2755\u20132762. doi:10.1152/jn.00336.2009 Obhi, S. S., Matkovich, S., & Gilbert, S. J. (2008). Modification of planned actions. Experimental Brain Research, 192(2), 299\u2013299. doi:10.1007/s00221-008-1618-z Press, C., Bird, G., Flach, R., & Heyes, C. (2005). Robotic movement elicits automatic imitation. Cognitive Brain Research, 25(3), 632\u2013640. doi:10.1016/j.cogbrainres.2005.08.020 Santiesteban, I., Banissy, M. J., Catmur, C., & Bird, G. (2012). Enhancing social ability by stimulating right temporoparietal junction. Current Biology, 22(23), 2274\u20132277. Santiesteban, I., White, S., Cook, J., Gilbert, S. J., Heyes, C., & Bird, G. (2012). Training social cognition: from imitation to Theory of Mind. Cognition, 122(2), 228\u201335. doi:10.1016/j.cognition.2011.11.004 Spengler, S., Bird, G., & Brass, M. (2010). Hyperimitation of actions is related to reduced understanding of others\u2019 minds in autism spectrum conditions. Biological Psychiatry, 68(12), 1148\u20131155. doi:10.1016/j.biopsych.2010.09.017 Spengler, S., von Cramon, D. Y., & Brass, M. (2010). Resisting motor mimicry: Control of imitation involves processes central to social cognition in patients with frontal and temporoparietal lesions. Social Neuroscience, 5(4), 401\u2013416. doi:10.1080/17470911003687905 Stroop, J. R. (1935). Studies of interference in serial verbal reactions. Journal of Experimental Psychology, 18, 643\u2013662. Tzelgov, J., Henik, A., & Berger, J. (1992). Controlling Stroop effects by manipulating expectations for color words. Memory & Cognition, 20(6), 727\u2013735. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/1435275 Wang, Y., Ramsey, R., & Hamilton, A. F. D. C. (2011). The control of mimicry by eye contact is mediated by medial prefrontal cortex. The Journal of Neuroscience, 31(33), 12001\u201312010. doi:10.1523/JNEUROSCI.0845-11.2011 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 PeerJ reviewing PDF | (v2013:07:696:1:0:NEW 23 Aug 2013) R ev ie w in g M an us cr ip t Figure Captions Figure 1. (A) Breakdown of the events in a trial of Experiment 1, displaying an incongruent index finger movement trial as an example. (B) Breakdown of the different types of trials in Experiment 1, that were randomized within blocks. Note: the number of baseline trials was equal to the sum total of congruent and incongruent trials 489 490 491 492 493 PeerJ reviewing PDF | (v2013:07:696:1:0:NEW 23 Aug 2013) R ev ie w in g M an us cr ip t Figure 2. Performance data for Experiment 1, with (A) reaction time (RT), and (B) proportion of errors (PE) plotted by experimental condition. Data labels represent the mean (and standard error of the mean) for each bar. The congruence effect was highly significant for both RT and PE (all ps < .001). 494 495 496 497 PeerJ reviewing PDF | (v2013:07:696:1:0:NEW 23 Aug 2013) R ev ie w in g M an us cr ip t Figure 3. Breakdown of the trial events and conditions from Experiment 2. (A) AIT displayed the cue and finger or dot movement simultaneously, whereas (B) the CIT displays the cue, followed by a finger or dot movement that appears during the response preparation period. (C) All of the conditions that took place in the experiment, which were randomized within both the AIT and the CIT blocks. Note: the cue numbers have been exaggerated for display purposes only. Also, whereas Experiment 1 used picture sequences, Experiment 2 immediately moved to the final state of the action, producing an apparent motion effect that has been shown to elicit robust effects in past AIT studies (C. Catmur & Heyes, 2011; Press et al., 2005). 498 499 500 501 502 503 504 505 PeerJ reviewing PDF | (v2013:07:696:1:0:NEW 23 Aug 2013) R ev ie w in g M an us cr ip t Figure 4. Size of the RT interference effect in experiment 2, with (A) the CIT and (B) the AIT interference effects plotted by cue type. Data labels represent the mean (and standard error of the mean) for each bar. The interference effect was significantly higher for the imitative cue in both the CIT and the AIT tasks (ps<.05). 506 507 508 509 PeerJ reviewing PDF | (v2013:07:696:1:0:NEW 23 Aug 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:07:696:1:0:NEW 23 Aug 2013) R ev ie w in g M an us cr ip t Figure 1 cit design PeerJ reviewing PDF | (v2013:07:696:1:0:NEW 23 Aug 2013) R ev ie w in g M an us cr ip t Figure 2 cit results PeerJ reviewing PDF | (v2013:07:696:1:0:NEW 23 Aug 2013) R ev ie w in g M an us cr ip t Figure 3 ait and cit design experiment 2 PeerJ reviewing PDF | (v2013:07:696:1:0:NEW 23 Aug 2013) R ev ie w in g M an us cr ip t Figure 4 experiment 2 interference PeerJ reviewing PDF | (v2013:07:696:1:0:NEW 23 Aug 2013) R ev ie w in g M an us cr ip t reaction time (rt) : Each participant\u2019s average RT for correct responses was entered into a repeated-measures ANOVA with one factor (congruence: congruent, or incongruent). The effect of congruence was highly significant (F1,11=103.06, p<.001, \u03b72=.90). Participants were faster on congruent trials (M=434, SE=9) than they were on incongruent trials (M=553, SE=14; t11=-10.15, p<.001, \u03b72=.68; Figure 2A). proportion of errors (pe) : Errors and missed responses (3.95% of total data) were summed and divided by the number of trials, yielding the proportion of errors (PE) for each participant for each condition. This data was entered into a repeated-measures ANOVA akin to the RT data, and the main effect of congruence was highly significant (F1,11=38.11, p<.001, \u03b72=.78). Participants made a substantially smaller number of errors on congruent (M=0.01, SE=0.01) trials compared to incongruent trials (M=0.14, SE=0.02; t11=-6.17, p<.001, \u03b72=.46; Figure 2B). Discussion The RT interference effect in the well-studied AIT indexes the cost associated with inhibiting other-related processing in favor of executing a self-related task. In study 1, we establish the CIT as a potential method for measuring the cost of self-related motor preparation to the execution of a conflicting imitative response \u2013 i.e. the cost of inhibiting self-related 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 PeerJ reviewing PDF | (v2013:07:696:1:0:NEW 23 Aug 2013) R ev ie w in g M an us cr ip t processing to prioritize processing of the other. Specifically, we found a robust effect in the CIT task: participants were 119ms slower on incongruent relative to congruent trials. While this result provides reason to believe that the CIT might be a useful addition to the study of self-other related processing, it is possible that the results simply represent the cost of action modification generally, rather the cost associated with conflicting self-other representations per se. It is worthwhile pointing out that the size of the CIT effect (119ms) is much greater than we would be anticipate in response to a socially innocuous action modification cue (e.g. 40ms for an auditory tone; Obhi, Matkovich, & Gilbert, 2008), suggesting that the robust effects reported in experiment 1 were at least partially driven by the \u2018specialness\u2019 of the observed action. Regardless, we conducted a second experiment to assess the extent to which the CIT reflects response modification generally, or is specific to controlling co-activated self-other motor representations. To shed more light on how the CIT and AIT might be useful complementary paradigms, we also added a set of AIT blocks to experiment 2. Experiment 2 In experiment 2, we sought to determine whether the CIT indexes self-other control in an imitation context, or simply quantifies the cost associated with modifying a planned response. To this end, participants performed the same task as described in experiment 1, but with the addition of a moving dot condition (spatial control; Cross, Torrisi, Losin, & Iacoboni, 2013; Figure 3). The dot cues were superimposed over the hand stimuli from experiment 1, and on CIT blocks participants were instructed to match what they saw onscreen whenever one of the dots, or one of the fingers, moved during their response preparation period. The difference between \u2018finger incongruent-congruent\u2019 interference and \u2018dot incongruent-congruent\u2019 interference provides an approximation of the cost associated with cancelling a self-related motor plan in favor of an 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 PeerJ reviewing PDF | (v2013:07:696:1:0:NEW 23 Aug 2013) R ev ie w in g M an us cr ip t imitative response, although as with all subtractions, this method may not perfectly isolate imitative processing. Experiment 2 also included a set of AIT blocks. Methods Participants Sixteen participants (9 female) between the ages of 17 and 40 (M=22.44, SD=5.18) completed the study. All of the participants were right-handed, and had normal or corrected-tonormal vision. The study conformed to local ethical guidelines and informed consent was obtained from all participants.",
+    "v2_text": "participants : Twelve participants (9 female) between the ages of 19 and 33 (M=23.33, SD=3.55) completed the study, which was run in accordance with local ethical guidelines. Informed consent was obtained from participants. Three of the 12 participants were left-handed, and all had normal or corrected-to-normal vision. stimuli & apparatus : The experiment was programmed using Superlab v.4.5 (Cedrus Corporation) and run on a Lenovo desktop computer (Lenovo Group Limited). Experimental stimuli were adapted from previous automatic imitation studies (Brass et al., 2000; Cook & Bird, 2012; Hogeveen & Obhi, 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 PeerJ reviewing PDF | (v2013:07:696:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t 2013). The stimuli contained a number cue, followed by an image series depicting either an index or middle finger movement (congruent and incongruent conditions), or a still image of a hand (baseline; Figure 1). Baseline trials made up 50% of the total trials, to ensure that participants were expecting to make the cued response and not simply waiting for the movement of the onscreen hand. The images were rotated orthogonally to the participants\u2019 hand to mitigate the influence of spatial compatibility between the observed action and the appropriate response. On baseline trials, the hand stayed in the same position for the duration of the trial (568ms). On congruent and incongruent trials, the still hand was replaced by a number cue, which was then replaced by the first stage (34ms), second stage (34ms), and final stage (500ms) of an index or middle finger lift. The final event on each trial was a blue screen, which was presented for 3000ms to allow for delayed responses and provide participants with enough time to get situated for the next trial. Procedure The experiment took place in a dimly lit cubicle. Participants were seated at the computer, told they would be holding down the v and b keys of the keyboard with their right index and middle fingers throughout the experiment, and would need to execute an index or middle finger lift as fast and accurately as possible when the number 1 or 2 was presented onscreen, respectively. Further, participants were told: \u201cif the hand onscreen moves, cancel your response and imitate what you see regardless of the movement you were planning to make.\u201d The study began with a practice session containing 24 trials, which observed by the experimenter to ensure that participants understood the task. If the participant made more than 5 mistakes, the practice section was re-run, which only occurred for one participant in the sample. The experiment proper was divided into six blocks of 32 trials, each containing four congruent index finger trials, four congruent middle finger trials, four incongruent index finger trials, four incongruent middle finger trials, eight baseline index finger trials, and eight baseline middle 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 PeerJ reviewing PDF | (v2013:07:696:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t finger trials. Each trial began with a still image of the hand on screen (300, 400, 500, 600, 700, or 800ms), then the response cue was presented, and after a delay period (50, 70, or 90ms) the hand either moved or stayed still (see Stimuli & Apparatus for a description of the hand movement). Results data exclusion : As described in the Methods section, the baseline condition was in place purely to ensure that participants expected to perform the cued response on the majority of trials, and was excluded from further analysis. Reaction time data for index and middle finger actions were collapsed and responses that were 3 SD\u2019s above or below the mean within each experimental condition (i.e. congruent, or incongruent) were excluded from statistical analysis, resulting in the removal of 0.39% of the data. The remaining congruent and incongruent trial data was then subjected to inferential statistical analysis. Reaction Time (RT) Each participant\u2019s average RT for correct responses was entered into a repeated-measures ANOVA with one factor (congruence: congruent, or incongruent). The effect of congruence was highly significant (F1,11=103.06, p<.001). Participants were faster on congruent trials (M=434, SE=9) than they were on incongruent trials (M=553, SE=14; t11=-10.15, p<.001; Figure 2A). Proportion of Errors (PE) Errors and missed responses (3.95% of total data) were summed and divided by the number of trials, yielding the proportion of errors (PE) for each participant for each condition. This data was entered into a repeated-measures ANOVA akin to the RT data, and the main effect of congruence was highly significant (F1,11=38.11, p<.001). Participants made a substantially smaller number of errors on congruent (M=0.01, SE=0.01) trials compared to incongruent trials (M=0.14, SE=0.02; t11=-6.17, p<.001; Figure 2B). Discussion 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 PeerJ reviewing PDF | (v2013:07:696:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t In the present experiment, we report a novel task requiring the online control of competing self- and other-generated motor representations. We have called this task the controlled imitation task (CIT), and suggest that it is a useful complementary task to the well-studied AIT. On each experimental trial, participants had to prepare a cued response, but on 50% of trials they were required to suppress this response and instead prioritize the production of an imitative response that was either congruent or incongruent with the cue. The robust result is that participants were significantly better at imitating an observed movement when it matched the response cue, relative to when it did not match the cue. We feel the CIT provides a valuable method for examining the influence of the self with respect to self-other control in the imitative context. To the extent that the production of a symbolically-cued response reflects an endogenously activated motor representation, the incongruent condition in the CIT can be thought of as requiring suppression of a self-activated action in favor of producing an imitative response, which his activated automatically by the stimulus. In this way, the present results are a perfect complement to the typical AIT data: whereas in the AIT the interference effect reflects the cost of inhibiting an other-generated motor representation in favor of the cued response, in the CIT the interference effect reflects the cost of inhibiting a self-generated motor representation in favor of imitation. There are several interesting issues and possibilities for future work that could be studied with the CIT. Specifically, in light of recent work on the neural basis of self-other control, it provides an excellent and quite well matched experimental paradigm with which to contrast AIT, and investigate task-appropriate shifts in processing towards self or other in the motor domain. For example, one brain region that has been discussed as playing a role in self-other control is right temporoparietal junction (rTPJ), and researchers recently used transcranial direct current stimulation (tDCS) to investigate this possibility. Specifically, Santiesteban and colleagues (2012) applied anodal tDCS to rTPJ for 20 minutes in order to enhance cortical excitability under the 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 PeerJ reviewing PDF | (v2013:07:696:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t electrode site, and found a significantly reduced interference effect in the AIT following stimulation. Their data suggested a causal role of rTPJ in suppressing the automatic tendency to imitate, in order to facilitate task performance which necessitates production of the cued response (Santiesteban, Banissy, et al., 2012). Whereas the size of the AIT interference effect was reduced after rTPJ stimulation, accuracy on a visual perspective-taking task was improved, suggesting that processing of the other person was enhanced or inhibited in a task-appropriate manner. Despite the impressive nature of this finding, the CIT provides a closer other-centric homologue for AIT than visual perspective taking. Thus, if the rTPJ is involved in the task dependent control of self- and other-generated motor representations, we would expect to observe an increased ability to suppress the self-activated motor representation in our task after anodal tDCS to the rTPJ. Thus, the control process in the CIT, at least in the incongruent trials, requires the suppression of a self-related representation whereas the control process in the standard AIT requires the suppression of an other-related representation. Seeing a pattern of results that is consistent with suppression of self-related activity in the CIT would bolster the notion that the rTPJ is involved in task specific control of self-other processing. In the non-social context, another very interesting aspect of the CIT is that it can be clearly contrasted with the truncation paradigm, in which internal generation of action is pitted against external cueing of action in non-imitative contexts (Obhi & Haggard, 2004; Obhi, Matkovich, & Chen, 2009; Obhi, Matkovich, & Gilbert, 2008). In particular, in the truncation paradigm participants begin the trial preparing to make an index finger movement at a time of their own choosing, but are interrupted on a portion of trials by a tone prompting production of the same movement. That, is, in the truncation paradigm, participants prepare to make a self-paced action and are interrupted with a non-social external cue to produce the same action or a different action, whereas in the current study the external cue to action is a movement by another person. Given the dissociation between the ability to inhibit imitation in the AIT and the 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 PeerJ reviewing PDF | (v2013:07:696:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t inhibition of other overlearned responses (Brass et al., 2005; Spengler, von Cramon, et al., 2010), it would be interesting to determine whether the CIT and action modification in the truncation paradigm depend on similar or specialized neural substrates. Such experiments will be essential to understanding the domain-generality or domain-specificity of the functional mechanisms underlying self-other control. In summary, we introduce a new paradigm that we have termed the \u201ccontrolled imitation task\u201d. Using an identical stimulus set to the established AIT, the CIT yields interference effects that shed light on self-other control processes, and that specifically tap into the capacity to suppress, or harness self-activated motor representations to facilitate fast production of an other-activated response (cf. Brass et al., 2009). We have outlined just two possible areas of inquiry where the CIT could a useful paradigm; self-other control, and the more general exploration of internally versus externally triggered action generation. It remains for future work to consider exactly how this paradigm can be employed to further understanding in these areas of study. Acknowledgements SSO is supported by grants from the Natural Sciences & Engineering research council of Canada and the Social Sciences & Humanities research council of Canada. 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 PeerJ reviewing PDF | (v2013:07:696:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t References Brass, M., Bekkering, H., Wohlschl\u00e4ger, A., & Prinz, W. (2000). Compatibility between observed and executed finger movements: Comparing symbolic, spatial, and imitative cues. Brain and Cognition, 44(2), 124\u2013143. doi:10.1006/brcg.2000.1225 Brass, M., Derrfuss, J., & von Cramon, D. Y. (2005). The inhibition of imitative and overlearned responses: A functional double dissociation. Neuropsychologia, 43(1), 89\u201398. doi:10.1016/j.neuropsychologia.2004.06.018 Brass, M., Ruby, P., & Spengler, S. (2009). Inhibition of imitative behaviour and social cognition. Philosophical Transactions of the Royal Society of London. Series B, Biological sciences, 364(1528), 2359\u20132367. doi:10.1098/rstb.2009.0066 Brass, M., Zysset, S., & von Cramon, Y. (2001). The inhibition of imitative response tendencies. NeuroImage, 14, 1416\u20131423. Cavallo, A., Heyes, C., Becchio, C., Bird, G., & Catmur, C. (2013). Timecourse of mirror and counter-mirror effects measured with transcranial magnetic stimulation. Social Cognitive and Affective Neuroscience, in press. doi:10.1093/scan/nst085 Cook, J. L., & Bird, G. (2012). Atypical social modulation of imitation in autism spectrum conditions. Journal of Autism and Developmental Disorders, 42(6), 1045\u20131051. doi:10.1007/s10803-011-1341-7 Fadiga, L., Fogassi, L., Pavesi, G., & Rizzolatti, R. (1995). Motor facilitation during action observation: A magnetic stimulation study. Journal of Neurophysiology, 73(6), 2608\u20132611. Heyes, C. (2011). Automatic imitation. Psychological Bulletin, 137(3), 463\u201383. doi:10.1037/a0022288 Hogeveen, J., & Obhi, S. S. (2011). Altogether now: Activating interdependent self-construal induces hypermotor resonance. Cognitive Neuroscience, 2(2), 74\u201382. doi:10.1080/17588928.2010.533164 Hogeveen, J., & Obhi, S. S. (2013). Automatic imitation is automatic, but less so for narcissists. Experimental brain research. Experimentelle Hirnforschung. Exp\u00e9rimentation c\u00e9r\u00e9brale, 224(4), 613\u201321. doi:10.1007/s00221-012-3339-6 Obhi, S. S., & Haggard, P. (2004). Internally generated and externally triggered actions are physically distinct and independently controlled. Experimental Brain Research., 156(4), 518\u2013523. doi:10.1007/s00221-004-1911-4 Obhi, S. S., & Hogeveen, J. (2010). Incidental action observation modulates muscle activity. Experimental Brain Research, 203(2), 427\u2013435. doi:10.1007/s00221-010-2253-z Obhi, S. S., Hogeveen, J., Giacomin, M., & Jordan, C. H. (n.d.). Automatic imitation is reduced in narcissists. Journal of Experimental Psychology: Human Perception and Performance. Obhi, S. S., Matkovich, S., & Chen, R. (2009). Changing the \u201cwhen\u201d and \u201cwhat\u201d of intended actions. Journal of Neurophysiology, 102(5), 2755\u20132762. doi:10.1152/jn.00336.2009 Obhi, S. S., Matkovich, S., & Gilbert, S. J. (2008). Modification of planned actions. Experimental Brain Research, 192(2), 299\u2013299. doi:10.1007/s00221-008-1618-z Santiesteban, I., Banissy, M. J., Catmur, C., & Bird, G. (2012). Enhancing social ability by stimulating right temporoparietal junction. Current Biology, 22(23), 2274\u20132277. Santiesteban, I., White, S., Cook, J., Gilbert, S. J., Heyes, C., & Bird, G. (2012). Training social cognition: from imitation to Theory of Mind. Cognition, 122(2), 228\u201335. doi:10.1016/j.cognition.2011.11.004 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 PeerJ reviewing PDF | (v2013:07:696:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t Spengler, S., Bird, G., & Brass, M. (2010). Hyperimitation of actions is related to reduced understanding of others\u2019 minds in autism spectrum conditions. Biological Psychiatry, 68(12), 1148\u20131155. doi:10.1016/j.biopsych.2010.09.017 Spengler, S., von Cramon, D. Y., & Brass, M. (2010). Resisting motor mimicry: Control of imitation involves processes central to social cognition in patients with frontal and temporo-parietal lesions. Social Neuroscience, 5(4), 401\u2013416. doi:10.1080/17470911003687905 Stroop, J. R. (1935). Studies of interference in serial verbal reactions. Journal of Experimental Psychology, 18, 643\u2013662. Tzelgov, J., Henik, A., & Berger, J. (1992). Controlling Stroop effects by manipulating expectations for color words. Memory & Cognition, 20(6), 727\u2013735. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/1435275 Wang, Y., Ramsey, R., & Hamilton, A. F. D. C. (2011). The control of mimicry by eye contact is mediated by medial prefrontal cortex. The Journal of Neuroscience, 31(33), 12001\u201312010. doi:10.1523/JNEUROSCI.0845-11.2011 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 PeerJ reviewing PDF | (v2013:07:696:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t Figures Figure 1. (A) Breakdown of the events in a trial of the controlled imitation experiment, displaying an incongruent index finger movement trial as an example. (B) Breakdown of the different types of experimental trials that were randomized within blocks throughout the experiment. Note: the number of baseline trials was equal to the sum total of congruent and incongruent trials. 299 300 301 302 303 304 305 PeerJ reviewing PDF | (v2013:07:696:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t Figure 2. Performance data for the present experiment, with (A) reaction time (RT), and (B) proportion of errors (PE) plotted by experimental condition. Data labels represent the mean (and standard error of the mean) for each bar. The congruence effect was highly significant for both RT and PE (all ps < .001). 306 307 308 309 PeerJ reviewing PDF | (v2013:07:696:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t Figure 1 The experimental design PeerJ reviewing PDF | (v2013:07:696:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t Figure 2 Interference effects and proportion of errors PeerJ reviewing PDF | (v2013:07:696:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t",
+    "url": "https://peerj.com/articles/163/reviews/",
+    "review_1": "Gabriele Sorci \u00b7 Aug 28, 2013 \u00b7 Academic Editor\nACCEPT\nAll the points raised by the three referees have been satisfactorely addressed in the revised version of the manuscript. Thank you for this nice piece of work.",
+    "review_2": "Gabriele Sorci \u00b7 Jun 20, 2013 \u00b7 Academic Editor\nMAJOR REVISIONS\nThe three referees found this manuscript interesting and timely, and I agree with them. The experimental design and the statistical analyses are sound, even though the referees asked for more detailed information, a more accurate interpretation and a clarification of terms.",
+    "review_3": "Alex Richter-Boix \u00b7 Jun 17, 2013\nBasic reporting\nThe manuscript by Steven Brady about local adaptation to relatively new habitat like roadside pools in amphibians is a nice example of the effects and consequences of landscape alteration of humans and the apparently incapacity of many organisms to adjust or adapt correctly to these new environments. In my opinion is a good contribution to conservation ecology from an evolutionary point of view.\nI will like but to suggest to the author to change the title, and remove the word \"fragmented landscape\". I agree with him, that roads (which is the main element analyzed in this manuscript as factor), contribute to landscape fragmentation, but from my point of view, the manuscript is not so much about \"strict\" fragmentation concept (reduction of habitat in small spots,etc.) but about new \"human created\" habitats. The forest pools where frogs are breeding are not reduced their areas, and the authors considered that gene flow among ponds is not affected by roads and traffic. The main factor is that pools closer to roads showed a different chemistry composition that those apart from the roads. I think that maladaptation is a consequence of contamination or alteration of habitats but not of landscape fragmentation.\nExperimental design\nField transplant experiments are nicely complemented with laboratory experiment where different factors like salt concentration can be isolated to test the effects of the main factor which could be involved on the adaptation or not of demes from different environments. The number of demes (10) distributed in five pair (roadside - forest), the number of clutches used from each environment and number of eggs and individuals used in each experiment, as the number of replicas give guarantees for the robustness of the results and acceptance or rejection of hypothesis.\nValidity of the findings\nStatistical analyses are also rightly performed, using in all cases good statistical approaches to analyze the data obtained from the field and laboratory experiment. The control of effect by egg size, or mother size allow the author to control potential maternal effects which could difficult the interpretation of the results obtained.\nThe figures are very compressible and help to readers to get a clear image of the results. I will only suggest to the author to change the scale of some figures, specially the one related with salt concentration treatments. Instead of using same distance between treatments I will like to use a lineal scale in the X-axis to better visualize the response of demes to different salt concentration treatment. Using a lineal scale it would be easier to see if mortality increase is lineal or there is some point of inflexion at which mortality increase in a exponential way.\nAdditional comments\nI will suggest to the author to discuss a little bit about the differences in clutch size between roadside and forest demes. In the results it seems that it is a statistical difference between them, with females from roadside demes producing more eggs than forest demes: around 10% more eggs. Considering that mortality from these demes in higher: 15% in transplant experiments, it can be that the increase of egg deposition in these demes would be an adaptive response to \"compensate\" the higher embryos mortality. I think that this is an important point to be discussed in the discussion section, as a mechanisms that allow to roadside demes to survive and persist in those environments even in depress conditions for longer time before they completely disappear or adapt to the new conditions.\nCite this review as\nRichter-Boix A (2013) Peer Review #1 of \"Microgeographic maladaptive performance and deme depression in response to roads and runoff (v0.1)\". PeerJ https://doi.org/10.7287/peerj.163v0.1/reviews/1",
+    "review_4": "Jos\u00e9 Cano \u00b7 Jun 17, 2013\nBasic reporting\nNo comments\nExperimental design\nI\u2019ve found quite difficult to follow the design of the reciprocal transplant experiment. A scheme with the design in the supplementary file would be very helpful to understand the allocation and levels of replication.\nThe data is not available for review. The author will make them available when accepted for publication.\nFigures and results for survival are summarizing means and SEs for the different origins/demes (roadside vs woodland). I would appreciate to see the results (and figures if it\u2019s not too messy) broken down at the experimental pool level to see whether the lower performance of the roadside origin is consistent across pools.\nIt is important to have more accurate information about the specific timing of egg collection in each deme and the stocking. Any potential differences in breeding time, especially between roadside and woodland pools, may have an impact on the results (implications explained in the \u201cvalidity of the findings\u201d section.\nThe full results of the analyses of the abiotic variables should be presented MANOVA table and descriptive stats (at least in the supplementary file). No doubt the author only found differences in conductance, but it would be informative to see the rest as well. Without knowing the pool first hand, one would expect differences in vegetal cover/radiation and probably temperature between woodland and roadsides.\nInfo about the specific R packages and procedures used to run the GLMMs would be appreciated.\nMany people may not be familiar with Bayesian inference and it would be good clarifying what the posterior mean refers to (eg. line 197) and the mean of what is being reported. Since the posterior mean is reported then is appropriate to give the associated 95% HPDI, not CIs, also explanation about how the Pmcmc is estimated would be helpful to grasp differences with frequentist p-values.\nValidity of the findings\nI find this study very relevant. Studies of microevolutionary (or lack of thereof) intraspecific processes at small scale are much needed. This is a good example of a pertinent well carried study with appropriated well-supported discussion.\nSome minor comments:\n- Differences in the onset of breeding (ca. days) determine growth opportunities and force acceleration/compensation in growth/development rates which, in turn, may have a cost in life-history traits (i.e. condition, size at metamorphosis, depression of immune response). For that reason would be important knowing if there was any significant difference in the times of collection and stocking of the clutches in the experimental pools. Especially any systematic difference that may have occurred between roadside and woodland pools. Perhaps this is not an issue in this study but it would be important to clarify it.\n-Lines 228-230. Perhaps \u201cinherited\u201d is not the best word choice since, as treated extensively in the discussion, there are many non-genetic potential causes behind the observed pattern.\n-Lines 265-270. In addition to accumulation of harmful chemicals may be other components of yolk quality (independent of egg size) differing between origins.\n-Lines 278-283. Assuming equal quality of the eggs of the two origins (woodland vs roadside) I don\u2019t follow this reasoning. The demographic success will rely on the actual number of eggs but not on the number of eggs relative to parent size. If without adjusting for body size there is no difference in fecundity between origins, no differential demographic effects would be expected (everything else being equal).\nCite this review as\nCano JM (2013) Peer Review #2 of \"Microgeographic maladaptive performance and deme depression in response to roads and runoff (v0.1)\". PeerJ https://doi.org/10.7287/peerj.163v0.1/reviews/2",
+    "pdf_1": "https://peerj.com/articles/163v0.2/submission",
+    "pdf_2": "https://peerj.com/articles/163v0.1/submission",
+    "review_5": "Andrew Hendry \u00b7 Jun 4, 2013\nBasic reporting\nPass\nExperimental design\nNeeds some clarification - see comments to author.\nValidity of the findings\nGood - but needs some improvement of interpretation - see comments to author.\nAdditional comments\nThis paper describes an interesting study that tested for local adaptation (woodland versus roadside ponds) by salamanders. The authors perform the usual local adaptation experiments through reciprocal transplants in nature and environmental manipulations (roadside salt) in the lab. They also estimate population sizes in the two environments in nature. Of particular interest, the authors find an unexpected pattern of maladaptation by the roadside salamanders \u2013 they show lower fitness under all conditions. This result appears quite strong and robust and is therefore very interesting and worth publishing.\nComments:\n1. Lines 38-43: Somewhere you need to say what the likely selective factor is \u2013 salt. You can\u2019t leave the reader guessing for so long.\n2. Line 76: Wouldn\u2019t unpaved roads also be salted?\n3. Line 80: But it would introduce a bias in the sense that eggs exposed to salt might be affected in such a way as to cause later reductions in survival even when transferred to non-salt. Thus, it is possible that this 36 hour differential exposure is the reason for the results \u2013 rather than local adaptation, differential migration, or maternal effects.\n4. Line 238: The dispersal would have to also be phenotype biased, which might not be clear based on the statement here. Also, the 36 hours of exposure before collection is another possibility.\n5. Line 240: I am not sure what is meant by \u201crarely reported.\u201d Hereford et al. show many instances where performance is not higher in local environments. (I am also reminded of the Hendry and Gonzalez\u2019s Biology and Philosophy debate about adaptation.) Maybe formally compare values from the current study to those in Hereford. For instance, how many studies in that meta-analysis fall into the category of one type performing better in both environments and, within that category, how strong is your result compared to the others?\n6. For maternal effects to be the cause, the mothers that breed in roadside ponds would also have to have been born in those ponds, right? Or do maternal effects accrue during the short period a female sits in the pond before laying? In short, be more clear on just how the maternal effects are expected to work.\n7. Line 297-304: See Falk et al. (2012) in EER for a possible example.\n8. Roadside females are said to have higher fecundity than woodland females but inspection of the data (Figure 4) shows that this is true only for large females. When making this sort of test, a significant interaction (I can\u2019t find information on whether it is or is not significant) means that one can\u2019t make blanket statements about the difference in egg size \u2013 because it depends on female size. (Note that this is not fixed by \u201cadjusting for body size.\u201d)\nI found the writing to be rather obtuse, with many abstract phrases and terms for which the meaning is ambiguous. The author should strive to improve clarity and to use more straightforward English whenever possible. Here I list a number of those instances \u2013 and these and others should be clarified. In fact, the entire MS should be examined for such instances.\n1. Abstract: define \u201crespond maladaptively\u201d\n2. Line 16: what is meant by \u201ctrajectories\u201d?\n3. Line 19: define \u201cdeme\u201d \u2013 \u201cpopulation\u201d is used more commonly (although I realize deme is the classical pop gen version) and it may not be clear if they mean similar or different things. Moreover, it is confusing later how this term is used. For instance, line 196 (and line 207 and elsewhere) refers to THE roadside deme but there are five roadside ponds, so doesn\u2019t that mean there could be five roadside demes. If by deme, you mean reproductively isolated population, then you don\u2019t know which are demes or not since you don\u2019t have genetic data reported. In short, this is all confusing and needs improvement.\n4. Line 19: Only a specific type of GxE is consistent with local adaptation.\n5. Line 67: Might not be clear to some whether \u201csite selection\u201d is by yourself or by the amphibians. Also it is not clear what is meant by \u201c8 of 10 overlapping pools\u201d \u2013 what happened to the other two and how are they \u201coverlapping\u201d?\n6. Line 85: Not clear what is meant by \u201cweather conditions.\u201d How is this relevant?\n7. Line 131: Confusingly worded as \u201cdissected\u201d for most people means the embryos are cut up, in which case they would survive. I assume you mean individual eggs are separated but you should clarify this.\n8. Line 146: I am not sure it is clear why the pools are \u201cpair-wise\u201d\n9. Line 177: The use of diameter here seems to conflict with the \u201ctwo-dimensional surface area\u201d mentioned earlier.\n10. Line 215: adjusting for body size is important but this phrase wasn\u2019t used in the methods and so I am not sure how the adjustment was done.\n11. Line 224: selection wasn\u2019t measured \u2013 perhaps rephrase\n12. Line 240: Preferentially means they prefer the site but some of the mechanisms here are instead competitive displacement, not preference.\nCite this review as\nHendry A (2013) Peer Review #3 of \"Microgeographic maladaptive performance and deme depression in response to roads and runoff (v0.1)\". PeerJ https://doi.org/10.7287/peerj.163v0.1/reviews/3",
+    "all_reviews": "Review 1: Gabriele Sorci \u00b7 Aug 28, 2013 \u00b7 Academic Editor\nACCEPT\nAll the points raised by the three referees have been satisfactorely addressed in the revised version of the manuscript. Thank you for this nice piece of work.\nReview 2: Gabriele Sorci \u00b7 Jun 20, 2013 \u00b7 Academic Editor\nMAJOR REVISIONS\nThe three referees found this manuscript interesting and timely, and I agree with them. The experimental design and the statistical analyses are sound, even though the referees asked for more detailed information, a more accurate interpretation and a clarification of terms.\nReview 3: Alex Richter-Boix \u00b7 Jun 17, 2013\nBasic reporting\nThe manuscript by Steven Brady about local adaptation to relatively new habitat like roadside pools in amphibians is a nice example of the effects and consequences of landscape alteration of humans and the apparently incapacity of many organisms to adjust or adapt correctly to these new environments. In my opinion is a good contribution to conservation ecology from an evolutionary point of view.\nI will like but to suggest to the author to change the title, and remove the word \"fragmented landscape\". I agree with him, that roads (which is the main element analyzed in this manuscript as factor), contribute to landscape fragmentation, but from my point of view, the manuscript is not so much about \"strict\" fragmentation concept (reduction of habitat in small spots,etc.) but about new \"human created\" habitats. The forest pools where frogs are breeding are not reduced their areas, and the authors considered that gene flow among ponds is not affected by roads and traffic. The main factor is that pools closer to roads showed a different chemistry composition that those apart from the roads. I think that maladaptation is a consequence of contamination or alteration of habitats but not of landscape fragmentation.\nExperimental design\nField transplant experiments are nicely complemented with laboratory experiment where different factors like salt concentration can be isolated to test the effects of the main factor which could be involved on the adaptation or not of demes from different environments. The number of demes (10) distributed in five pair (roadside - forest), the number of clutches used from each environment and number of eggs and individuals used in each experiment, as the number of replicas give guarantees for the robustness of the results and acceptance or rejection of hypothesis.\nValidity of the findings\nStatistical analyses are also rightly performed, using in all cases good statistical approaches to analyze the data obtained from the field and laboratory experiment. The control of effect by egg size, or mother size allow the author to control potential maternal effects which could difficult the interpretation of the results obtained.\nThe figures are very compressible and help to readers to get a clear image of the results. I will only suggest to the author to change the scale of some figures, specially the one related with salt concentration treatments. Instead of using same distance between treatments I will like to use a lineal scale in the X-axis to better visualize the response of demes to different salt concentration treatment. Using a lineal scale it would be easier to see if mortality increase is lineal or there is some point of inflexion at which mortality increase in a exponential way.\nAdditional comments\nI will suggest to the author to discuss a little bit about the differences in clutch size between roadside and forest demes. In the results it seems that it is a statistical difference between them, with females from roadside demes producing more eggs than forest demes: around 10% more eggs. Considering that mortality from these demes in higher: 15% in transplant experiments, it can be that the increase of egg deposition in these demes would be an adaptive response to \"compensate\" the higher embryos mortality. I think that this is an important point to be discussed in the discussion section, as a mechanisms that allow to roadside demes to survive and persist in those environments even in depress conditions for longer time before they completely disappear or adapt to the new conditions.\nCite this review as\nRichter-Boix A (2013) Peer Review #1 of \"Microgeographic maladaptive performance and deme depression in response to roads and runoff (v0.1)\". PeerJ https://doi.org/10.7287/peerj.163v0.1/reviews/1\nReview 4: Jos\u00e9 Cano \u00b7 Jun 17, 2013\nBasic reporting\nNo comments\nExperimental design\nI\u2019ve found quite difficult to follow the design of the reciprocal transplant experiment. A scheme with the design in the supplementary file would be very helpful to understand the allocation and levels of replication.\nThe data is not available for review. The author will make them available when accepted for publication.\nFigures and results for survival are summarizing means and SEs for the different origins/demes (roadside vs woodland). I would appreciate to see the results (and figures if it\u2019s not too messy) broken down at the experimental pool level to see whether the lower performance of the roadside origin is consistent across pools.\nIt is important to have more accurate information about the specific timing of egg collection in each deme and the stocking. Any potential differences in breeding time, especially between roadside and woodland pools, may have an impact on the results (implications explained in the \u201cvalidity of the findings\u201d section.\nThe full results of the analyses of the abiotic variables should be presented MANOVA table and descriptive stats (at least in the supplementary file). No doubt the author only found differences in conductance, but it would be informative to see the rest as well. Without knowing the pool first hand, one would expect differences in vegetal cover/radiation and probably temperature between woodland and roadsides.\nInfo about the specific R packages and procedures used to run the GLMMs would be appreciated.\nMany people may not be familiar with Bayesian inference and it would be good clarifying what the posterior mean refers to (eg. line 197) and the mean of what is being reported. Since the posterior mean is reported then is appropriate to give the associated 95% HPDI, not CIs, also explanation about how the Pmcmc is estimated would be helpful to grasp differences with frequentist p-values.\nValidity of the findings\nI find this study very relevant. Studies of microevolutionary (or lack of thereof) intraspecific processes at small scale are much needed. This is a good example of a pertinent well carried study with appropriated well-supported discussion.\nSome minor comments:\n- Differences in the onset of breeding (ca. days) determine growth opportunities and force acceleration/compensation in growth/development rates which, in turn, may have a cost in life-history traits (i.e. condition, size at metamorphosis, depression of immune response). For that reason would be important knowing if there was any significant difference in the times of collection and stocking of the clutches in the experimental pools. Especially any systematic difference that may have occurred between roadside and woodland pools. Perhaps this is not an issue in this study but it would be important to clarify it.\n-Lines 228-230. Perhaps \u201cinherited\u201d is not the best word choice since, as treated extensively in the discussion, there are many non-genetic potential causes behind the observed pattern.\n-Lines 265-270. In addition to accumulation of harmful chemicals may be other components of yolk quality (independent of egg size) differing between origins.\n-Lines 278-283. Assuming equal quality of the eggs of the two origins (woodland vs roadside) I don\u2019t follow this reasoning. The demographic success will rely on the actual number of eggs but not on the number of eggs relative to parent size. If without adjusting for body size there is no difference in fecundity between origins, no differential demographic effects would be expected (everything else being equal).\nCite this review as\nCano JM (2013) Peer Review #2 of \"Microgeographic maladaptive performance and deme depression in response to roads and runoff (v0.1)\". PeerJ https://doi.org/10.7287/peerj.163v0.1/reviews/2\nReview 5: Andrew Hendry \u00b7 Jun 4, 2013\nBasic reporting\nPass\nExperimental design\nNeeds some clarification - see comments to author.\nValidity of the findings\nGood - but needs some improvement of interpretation - see comments to author.\nAdditional comments\nThis paper describes an interesting study that tested for local adaptation (woodland versus roadside ponds) by salamanders. The authors perform the usual local adaptation experiments through reciprocal transplants in nature and environmental manipulations (roadside salt) in the lab. They also estimate population sizes in the two environments in nature. Of particular interest, the authors find an unexpected pattern of maladaptation by the roadside salamanders \u2013 they show lower fitness under all conditions. This result appears quite strong and robust and is therefore very interesting and worth publishing.\nComments:\n1. Lines 38-43: Somewhere you need to say what the likely selective factor is \u2013 salt. You can\u2019t leave the reader guessing for so long.\n2. Line 76: Wouldn\u2019t unpaved roads also be salted?\n3. Line 80: But it would introduce a bias in the sense that eggs exposed to salt might be affected in such a way as to cause later reductions in survival even when transferred to non-salt. Thus, it is possible that this 36 hour differential exposure is the reason for the results \u2013 rather than local adaptation, differential migration, or maternal effects.\n4. Line 238: The dispersal would have to also be phenotype biased, which might not be clear based on the statement here. Also, the 36 hours of exposure before collection is another possibility.\n5. Line 240: I am not sure what is meant by \u201crarely reported.\u201d Hereford et al. show many instances where performance is not higher in local environments. (I am also reminded of the Hendry and Gonzalez\u2019s Biology and Philosophy debate about adaptation.) Maybe formally compare values from the current study to those in Hereford. For instance, how many studies in that meta-analysis fall into the category of one type performing better in both environments and, within that category, how strong is your result compared to the others?\n6. For maternal effects to be the cause, the mothers that breed in roadside ponds would also have to have been born in those ponds, right? Or do maternal effects accrue during the short period a female sits in the pond before laying? In short, be more clear on just how the maternal effects are expected to work.\n7. Line 297-304: See Falk et al. (2012) in EER for a possible example.\n8. Roadside females are said to have higher fecundity than woodland females but inspection of the data (Figure 4) shows that this is true only for large females. When making this sort of test, a significant interaction (I can\u2019t find information on whether it is or is not significant) means that one can\u2019t make blanket statements about the difference in egg size \u2013 because it depends on female size. (Note that this is not fixed by \u201cadjusting for body size.\u201d)\nI found the writing to be rather obtuse, with many abstract phrases and terms for which the meaning is ambiguous. The author should strive to improve clarity and to use more straightforward English whenever possible. Here I list a number of those instances \u2013 and these and others should be clarified. In fact, the entire MS should be examined for such instances.\n1. Abstract: define \u201crespond maladaptively\u201d\n2. Line 16: what is meant by \u201ctrajectories\u201d?\n3. Line 19: define \u201cdeme\u201d \u2013 \u201cpopulation\u201d is used more commonly (although I realize deme is the classical pop gen version) and it may not be clear if they mean similar or different things. Moreover, it is confusing later how this term is used. For instance, line 196 (and line 207 and elsewhere) refers to THE roadside deme but there are five roadside ponds, so doesn\u2019t that mean there could be five roadside demes. If by deme, you mean reproductively isolated population, then you don\u2019t know which are demes or not since you don\u2019t have genetic data reported. In short, this is all confusing and needs improvement.\n4. Line 19: Only a specific type of GxE is consistent with local adaptation.\n5. Line 67: Might not be clear to some whether \u201csite selection\u201d is by yourself or by the amphibians. Also it is not clear what is meant by \u201c8 of 10 overlapping pools\u201d \u2013 what happened to the other two and how are they \u201coverlapping\u201d?\n6. Line 85: Not clear what is meant by \u201cweather conditions.\u201d How is this relevant?\n7. Line 131: Confusingly worded as \u201cdissected\u201d for most people means the embryos are cut up, in which case they would survive. I assume you mean individual eggs are separated but you should clarify this.\n8. Line 146: I am not sure it is clear why the pools are \u201cpair-wise\u201d\n9. Line 177: The use of diameter here seems to conflict with the \u201ctwo-dimensional surface area\u201d mentioned earlier.\n10. Line 215: adjusting for body size is important but this phrase wasn\u2019t used in the methods and so I am not sure how the adjustment was done.\n11. Line 224: selection wasn\u2019t measured \u2013 perhaps rephrase\n12. Line 240: Preferentially means they prefer the site but some of the mechanisms here are instead competitive displacement, not preference.\nCite this review as\nHendry A (2013) Peer Review #3 of \"Microgeographic maladaptive performance and deme depression in response to roads and runoff (v0.1)\". PeerJ https://doi.org/10.7287/peerj.163v0.1/reviews/3\nReview 6: \n",
+    "LLM_extracted_review": "['Review 1: All the points raised by the three referees have been satisfactorily addressed in the revised version of the manuscript. Thank you for this nice piece of work.\\n\\nReview 2: The three referees found this manuscript interesting and timely, and I agree with them. The experimental design and the statistical analyses are sound, even though the referees asked for more detailed information, a more accurate interpretation and a clarification of terms.\\n\\nReview 3: The manuscript is a good contribution to conservation ecology from an evolutionary point of view. I suggest changing the title and removing the term \"fragmented landscape.\" The manuscript focuses more on new \"human created\" habitats rather than strict fragmentation. The field transplant experiments are well complemented with laboratory experiments, ensuring robustness in results. Statistical analyses are performed correctly, but I suggest changing the scale of some figures for better visualization. Additionally, discussing the differences in clutch size between roadside and forest demes is important, as it may indicate an adaptive response to higher embryo mortality.\\n\\nReview 4: I found it difficult to follow the design of the reciprocal transplant experiment; a scheme would be helpful. The results should be broken down at the experimental pool level to assess consistency across pools. More accurate information about the timing of egg collection is needed, as it may impact results. The full results of abiotic variable analyses should be presented, and clarification on the R packages used for GLMMs would be appreciated. It would also be beneficial to clarify terms related to Bayesian inference for better understanding.\\n\\nReview 5: The study is interesting and tests for local adaptation by salamanders. However, the interpretation of results needs improvement. The manuscript should clarify the likely selective factor (salt) and address potential biases in the study design. The writing is somewhat obtuse and should be clearer. Specific terms and phrases need definitions or rephrasing for better understanding. Additionally, the relationship between maternal effects and demographic success should be clarified, and the significance of differences in egg size should be addressed more carefully.']"
+}

peerj_json_files/PeerJ_Json_119.json ADDED Viewed

	@@ -0,0 +1,18 @@

+{
+    "v1_Abstract": "Despite theoretical understanding and empirical detection of local adaptation in natural environments, our knowledge of such divergence in fragmented habitats remains limited, especially in the context of microgeographic spatial scales and contemporary time scales. I used a combination of reciprocal transplant and common garden exposure experiments to evaluate potential microgeographic divergence in a pool-breeding amphibian occupying a landscape fragmented by roads. As indicated by reduced rates of survival and increased rates of malformation, I found evidence for maladaptation in road adjacent populations. This response is in direct counterpoint to recently described local adaption by a cohabiting species of amphibian. These results suggest that while divergence might commonly follow habitat modification, the direction of its outcome cannot be generalized even in identical habitats. Further, maladaptive responses can be associated with a more generalized depression effect that transcends the local environment. Alongside recent reports, these results suggest that maladaptive responses may be an emerging consequence of human-induced environmental change. Thus future studies should carefully consider the population unit as a key level for inference.",
+    "v1_col_introduction": "introduction  : Identifying informative scales and levels of biological organization for investigation has long been regarded as a critical challenge in ecology, limiting our ability to describe variation in the natural world (Levin 1992). Population level approaches\u2014which are bringing to light profound variation that occurs across small spatial and temporal scales (Pigliucci 2005)\u2014offer a promising advancement on this challenge. Variation observed at this level can reveal dramatic differences among local populations, even across microgeographic spatial scales (Skelly 2004; Urban 2010). For example, spatially heterogeneous selection pressures coupled with heritable phenotypic variation has often been found to result in local adaptation (Hereford 2009; Kawecki & Ebert 2004). Evidenced by a specific form of gene-by-environment interaction (G x E), local adaptation results when the local population has higher fitness than a foreign population measured within the local environment (Kawecki & Ebert 2004). However, gene flow and genetic drift can constrain local adaptation, rendering populations equivalently adapted between environments, or even maladapted to their local environment (Crespi 2000; Falk et al. 2012; Stearns & Sage 1980; Urban 2006). Though evident in wild populations (Moore & Hendry 2009), such empirical examples of population level maladaptation are rare (Spitzer 2006). Rather, most examples of maladaptation in general describe reduced fitness across higher levels of organization, typically in host-parasite or host-pathogen systems, where coevolution can generate reciprocating dynamics of maladaptation and adaptation (Thompson 1999).\nIn human-altered contexts, where novel selection pressures and reduced gene flow can yield adaptive responses (Brady 2012) and/or genetic drift (Delaney et al. 2010), and the rates of phenotypic change are high relative to undisturbed settings (Darimont et al. 2009; Hendry et al. 2008), the potential for population divergence should be elevated. Given this increased potential, population level approaches should be viewed as essential for fulfilling conservation imperatives\n8\n9\n10\n11\n12\n13\n14\n15\n16\n17\n18\n19\n20\n21\n22\n23\n24\n25\n26\n27\n28\n29\n30\n31\nPeerJ reviewing PDF | (v2013:05:486:1:1:NEW 21 Aug 2013)\nR ev ie w in g M an\nus cr ip t\naimed at understanding the long-term fates of populations influenced by human activities. Here, I report the population specific responses of a pool-breeding amphibian (the wood frog, Rana sylvatica = Lithobates sylvaticus) occupying a landscape fragmented by roads. Given the suite of novel selection pressures associated with road fragmentation and runoff\u2014along with previously documented contemporary local adaptation to temperature by the wood frog in this landscape (Skelly 2004)\u2014I expected that road adjacent populations of this species would be locally adapted to the roadside environment. In particular, I expected road runoff contaminants\u2014especially road salt (Brady 2012)\u2014to act as the selective pressure conducing local adaptation in these populations. This outcome seems especially likely in light of recent evidence of local adaptation to roads by another amphibian (the spotted salamander, Ambystoma maculatum) breeding and dwelling in the same locations (Brady 2012).\nI hypothesized 1) that the roadside environment would induce negative effects on the aquatic life stages of the wood frog, and 2) that within a common roadside environment, roadside populations would outperform woodland populations (i.e. the local adaptation hypothesis). I used a fieldbased reciprocal transplant experiment (Figs. 1 and S1) coupled with a series of lab-based exposure experiments to test these hypotheses across a total of 10 populations (five roadside, five woodland). Each of these populations occupied a distinct seasonal pool, which comprises natural, typical breeding habitat. To evaluate demographic consequences of roads in this system, I surveyed wild population size and female reproductive effort. Finally, to investigate potential mechanisms by which road proximity influences wood frog performance, I characterized the environment at each pool by estimating a suite of abiotic variables associated with amphibian distribution and performance (Wellborn et al. 1996). I expected that these characteristics might vary with road proximity, and thus act as potential agents of natural selection in roadside pools.\n32\n33\n34\n35\n36\n37\n38\n39\n40\n41\n42\n43\n44\n45\n46\n47\n48\n49\n50\n51\n52\n53\n54\nPeerJ reviewing PDF | (v2013:05:486:1:1:NEW 21 Aug 2013)\nR ev ie w in g M an\nus cr ip t",
+    "v2_Abstract": "Despite theoretical understanding and empirical detection of local adaptation in natural environments, our knowledge of such divergence in fragmented habitats remains limited, especially in the context of microgeographic spatial scales and contemporary time scales. I evaluated this potential in a pool-breeding amphibian occupying a landscape fragmented by roads. My data show that road adjacent populations respond maladaptively to both roadside and nearby woodland environments, in direct counterpoint to recently described local adaption by a cohabiting species of amphibian. These results indicate that maladaptive responses can be associated with a more generalized depression effect that transcends the local environment, and that while divergence might commonly follow fragmentation, the direction of its outcome cannot be generalized even in identical habitats. Alongside recent reports, these results suggest that maladaptive responses may be an emerging consequence of human-induced environmental change. Thus future studies should carefully consider the deme as a key level for inference.",
+    "v2_col_introduction": "introduction  : Identifying informative scales and levels of biological organization for investigation has long been regarded as a critical challenge in ecology, limiting our ability to describe variation in the natural world (Levin 1992). Deme level approaches\u2014which are bringing to light profound variation that occurs across small spatial and temporal scales (Pigliucci 2005)\u2014offer a promising advancement on this challenge. Variation observed at the deme level can reveal dramatically different trajectories among local populations, even across microgeographic spatial scales (Skelly 2004; Urban 2010). For example, spatially heterogeneous selection pressures coupled with heritable phenotypic variation has often been found to result in local adaptation at the level of the deme (Hereford 2009; Kawecki & Ebert 2004). Evidenced by a gene-by-environment interaction (G x E), local adaptation results when the local deme has higher fitness than a foreign deme measured within the local environment (Kawecki & Ebert 2004). However, gene flow and genetic drift can constrain local adaptation, rendering demes equivalently adapted between environments, or even maladapted to their local environment (Crespi 2000; Falk et al. 2012; Stearns & Sage 1980; Urban 2006). Though evident in wild populations (Moore & Hendry 2009), such empirical examples of deme level maladaptation are rare (Spitzer 2006). Most examples of maladaptation in general describe reduced fitness across higher levels of organization, typically in host-parasite or host-pathogen systems, where coevolution can generate reciprocating dynamics of maladaptation and adaptation (Thompson 1999).\nIn human-altered contexts, where novel selection pressures and reduced gene flow can yield adaptive responses (Brady 2012) and/or genetic drift (Delaney et al. 2010), and the rates of phenotypic change are high relative to undisturbed settings (Darimont et al. 2009; Hendry et al.\n2 2\n10\n11\n12\n13\n14\n15\n16\n17\n18\n19\n20\n21\n22\n23\n24\n25\n26\n27\n28 29\n30\n31\n32\nPeerJ reviewing PDF | (v2013:05:486:0:0:NEW 10 May 2013)\nR ev ie w in g M an\nus cr ip t\n2008), the potential for deme divergence should be elevated. Given this increased potential, deme level approaches should be viewed as essential for fulfilling conservation imperatives aimed at understanding the long-term fates of populations influenced by human activities. Here, I report the deme specific responses of a pool-breeding amphibian (the wood frog, Rana sylvatica = Lithobates sylvaticus) occupying a landscape fragmented by roads. Given the suite of novel selection pressures associated with road fragmentation and runoff\u2014along with previously documented contemporary local adaptation to temperature by the wood frog in this landscape (Skelly 2004)\u2014I expected that road adjacent populations of this species would be locally adapted to the roadside environment. This outcome seems especially likely in light of recent evidence of local adaptation to roads by another amphibian (the spotted salamander, Ambystoma maculatum) breeding and dwelling in the same locations (Brady 2012).\nI hypothesized 1) that the roadside environment would induce negative effects on the aquatic life stages of the wood frog, and 2) that within a common roadside environment, the roadside deme would outperform the woodland deme (i.e. the local adaptation hypothesis). I used a field-based reciprocal transplant experiment (Fig. 1) coupled with a series of lab-based exposure experiments to test these hypotheses across a total of 10 demes (each consisting of a distinct roadside or woodland pool), which comprise natural, typical breeding habitat. To evaluate demographic consequences of roads in this system, I surveyed wild population size and female reproductive effort. Finally, to investigate potential mechanisms by which road proximity influences wood frog performance, I characterized the environment at each pool by estimating a suite of abiotic variables associated with amphibian distribution and performance (Wellborn et al.\n3 3\n33\n34\n35\n36\n37\n38\n39\n40\n41\n42\n43 44\n45\n46\n47\n48\n49\n50\n51\n52\n53\n54\nPeerJ reviewing PDF | (v2013:05:486:0:0:NEW 10 May 2013)\nR ev ie w in g M an\nus cr ip t\n1996). I expected that these characteristics might vary with road proximity, and thus act as potential agents of natural selection in roadside pools.",
+    "v1_text": "materials and methods : Natural history and site selection The wood frog is widely distributed throughout eastern North America and much of Canada, with a range extending from within the Arctic Circle to the southeastern United States. Within the study region, populations of this species breed in late March or early April, when adults migrate from upland terrestrial habitat into ephemeral wetlands to reproduce. The wood frog is an explosive breeder, with the majority of oviposition for a given population occurring in a single night (Berven & Grudzien 1990). Populations in this study region typically breed synchronously within several days of one another, when each female oviposits one egg mass containing approximately 800 embryos per mass. Embryos develop over two-three weeks before hatching, and continue to develop as aquatic larvae throughout spring and early summer until they metamorphose into terrestrial juveniles. My method for selecting populations for this study is previously reported in a reciprocal transplant study that indicated local adaptation to roads in the spotted salamander (Brady 2012). Briefly, I selected the five roadside pools with the highest conductivity (\u00b5S) values (an indicator of road salt runoff) and with a breeding population comprising at least 10 wood frog families. For each roadside pool, a nearby woodland pool was then selected to minimize confounding effects. Specifically, I selected woodland pools based on the criterion of maximizing similarity in the abiotic conditions of canopy cover, pool size, and the presence and type of emergent vegetation. Reciprocal transplants (described below) were conducted within each of these five pool pairs (see Figs. 1 and S1 for overview and schematic). Eight of the 10 pools selected during the aforementioned spotted salamander study (Brady 2012) were also selected here; the remaining two pools did not have breeding populations of the wood frog and thus two different pools were selected for the present study. Inference from these overlapping populations provides an 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 PeerJ reviewing PDF | (v2013:05:486:1:1:NEW 21 Aug 2013) R ev ie w in g M an us cr ip t especially strong basis for comparison of responses between two amphibians occupying identical habitats. When referring to the full complement of roadside or woodland populations, I use the term \u2018deme.\u2019 Thus, \u2018roadside deme\u2019 refers collectively to the five roadside populations while \u2018woodland deme\u2019 refers collectively to the five woodland populations. The work described here was conducted with permitting from the State of Connecticut Department of Environmental Protection and under protocols from the Yale University Institutional Animal Care and Use Committee (Projects 2006-11024 and 2009-11024). Reciprocal transplant experiment In spring 2008, I collected from each of 10 pools (five roadside, five woodland) a subset of embryos from 10 egg masses less than 36 hours old. To ensure collection within this timeframe, pools were monitored daily as the breeding season approached. Roadside pools were located < 10 m from a paved road while woodland pools were located > 200 m from the nearest paved road and at least 100 m from the nearest dirt road (Fig. 1). Dirt roads in this region typically see low traffic use and do not receive winter deicing salts. While roadside egg masses appeared somewhat more compact than woodland egg masses (Karraker & Gibbs 2011), there were no conspicuous differences in the embryos themselves between sites, and selection of embryos from within each egg mass was haphazard. Thus, any potential carryover effect of the < 36 hour exposure did not bias the selection of the subset of embryos employed in the experiment. The consequences of such potential effects are elaborated upon the Discussion. Each egg mass was collected whole into a plastic 710 ml container. From each egg mass, I separated out two clusters of embryos. Each cluster was photographed and stocked into one of five experimental enclosures in the origin pool and the transplant pool (Fig. S1). (Details of the experimental enclosures are described elsewhere (Brady 2012) and an image can be found in Figure S1.) For a subset of egg masses, this procedure occurred in the field. However, heavy rains impeded photography; remaining egg 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 PeerJ reviewing PDF | (v2013:05:486:1:1:NEW 21 Aug 2013) R ev ie w in g M an us cr ip t masses were instead transported to the lab on ice for processing, stored overnight in an 8 \u00b0C incubator, and returned for stocking the following day. Each pool contained five experimental blocks. I targeted stocking approximately 50 embryos from each of two egg masses per enclosure, yielding one unique pairing of egg masses (hereafter \u201cclutch pair\u201d) that was replicated across, but not within pool pairs. This design was chosen to maximize family level diversity while maintaining an additional level of replication at the clutch pair level, while at the same time balancing logistics of resources and spatial constraints within pools. Enclosure assignment was haphazard for each clutch pair. In total across 10 pools, I stocked 100 enclosures containing approximately 100 embryos each. Breeding began first at the two most southerly pools on 2 April 2008. 8 April 2008 was the last day that breeding began in a new pool. Given the proximity between each pair of pools (i.e. the complement of a woodland-roadside pairing), breeding began within a one-day period across all pairs. Stocking of enclosures occurred in a pairwise manner between each roadside and woodland pond. In this way, there were no systematic differences in the timing of collection or stocking. I used a dissecting stereoscopic microscope to estimate the developmental stage of each egg mass upon collection from the field. For each embryo cluster (n = 200) I estimated egg size from photographs. Specifically, I used ImageJ (Rasband 1997-2012) to measure the two-dimensional surface area of 15 embryos per cluster represented by a best-fit ellipse (Brady 2012). At the conclusion of the experiment, when all eggs had either hatched and reached feeding stage or died, I estimated survival, developmental stage (Gosner 1960), and snout-vent length (SVL) at the container level. One enclosure was removed from analysis due to predator intrusion. Chronic road salt exposure experiment Of the 10 egg masses collected, nine were haphazardly assigned into groups of three (hereafter \u201cclutch triad\u201d). From each egg mass, three separate clusters\u2014each of approximately 30 embryos 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 PeerJ reviewing PDF | (v2013:05:486:1:1:NEW 21 Aug 2013) R ev ie w in g M an us cr ip t \u2014were separated out and haphazardly assigned by clutch triad to each of three road salt treatments. Each cluster (n = 270) was photographed and embryo size was estimated as described above. Each experimental unit comprised one clutch triad, the unique composition of which was maintained across treatments providing an additional random effect. In total, each of 10 pools contributed representatives from nine egg masses, grouped in clutches of three, and exposed to three treatments for a total of 90 experimental units (3 clutch triads per treatment x 3 treatments x 10 pools = 90). The three treatments (low, medium, high) reflected the range of specific conductance found in the study region. All treatments comprised aged and conditioned tap water. Road salt comprising at least 95% NaCl obtained from the Connecticut Department of Transportation was added to the medium and high treatments to achieve specific conductance values of 1,000 \u00b5S and 4,000 \u00b5S respectively, which approximate the mean and maximum values I detected in roadside pools in the study region. No road salt was added to the low treatment, which had a natural specific conductance of approximately 175 \u00b5S. Embryos were reared in 11.3 l plastic containers (42 x 28 x 18 cm) filled with approximately 8 l of assigned treatment water. Each container was fitted with mesh hardware cloth to suspend embryos in the water column off the bottom of the container. Containers were housed on tables in an outdoor facility under 50% shade cloth. Concentrations of roadsalt were maintained in response to evaporation and rain by adding lower or higher concentration treatment water respectively, and through routine, complete water changes. When all embryos hatched or died, I estimated survival, developmental stage, SVL, and prevalence of axial malformations (Bantle et al. 1991). Acute road salt exposure experiment Following assignment of embryo clusters to reciprocal transplant and chronic exposure experiments described above, egg masses were retained within collection containers and placed haphazardly within in 150 l plastic wading pools that served as incubators, located in an outdoor 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 PeerJ reviewing PDF | (v2013:05:486:1:1:NEW 21 Aug 2013) R ev ie w in g M an us cr ip t enclosure covered with 50% shade cloth. When embryos reached developmental stage 18, I separated out approximately 45 embryos from each egg mass and placed them into a common garden environment consisting of a 150 l wading pool filled with aged, conditioned tap water. Embryos were grouped by origin such that each wading pool contained embryos from each of the 10 egg masses representing a single natural pool. Upon approaching feeding stage [Gosner developmental stage 25 (Gosner 1960)], larvae were fed a mixture of 3:1 ground rabbit chow and fish flakes at a daily ration of 10% average body mass. When all individuals appeared to reach feeding stage, I exposed wood frog larvae to a series of five concentrations of road salt, ranging from 0.5 to 10.0 g/l. These concentrations were selected to represent previously documented acute exposure assays of the wood frog (Sanzo & Hecnar 2006). From each pool, 10 larvae were randomly stocked into each treatment and replicated across four experimental units per treatment (10 pools x 5 concentrations x 4 replicates = 200 experimental units). A 5.1 l container (33 x 20 x 11 cm) filled with 4 l of treatment water comprised each experimental unit. Larvae in each experimental unit were given a four-day food ration at the start of the experiment. After 96 hours of exposure, I evaluated larval survival. In order to standardize developmental stage of larvae in the experiment, exposure was conducted in two separate, consecutive 96-hour rounds\u2014the first comprising six pools (representing three pool pairs) and the second comprising four pair-wise pools (representing two pool pairs). Population size and female fecundity In spring 2008 following the conclusion of natural oviposition, I conducted visual counts of wood frog egg masses in each of the 10 pools to estimate population size of breeding females. In spring 2010, I collected 56 pairs of inbound wood frog adults from a subset of six pools (three roadside, three woodland) and haphazardly assigned male-female combinations grouped by origin pool. Pairs were placed in 11.3 l plastic containers filled with 2.5 l of aged spring water, equipped with 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 PeerJ reviewing PDF | (v2013:05:486:1:1:NEW 21 Aug 2013) R ev ie w in g M an us cr ip t a small branch and one white oak leaf (Quercus alba) to provide a substrate for oviposition. Containers were positioned on a slight angle to create a depth cline, and adults were allowed to amplex and oviposit. Following oviposition, egg masses were photographed and fecundity was estimated from each photo using ImageJ (Rasband 1997-2012)to manually count the total number of eggs oviposited per female. Characterizing the environment In each pool, I measured seven environmental characteristics associated with amphibian distribution and performance (Wellborn et al. 1996). Specific conductance, dissolved oxygen, pH, and wetland depth were measured once during the experiment, while temperature was measured every thirty minutes using deployed temperature loggers. Global site factor\u2014a measure of solar radiation reaching the pool\u2014was calculated from hemispherical photographs, while wetland area was estimated from visual rangefinder measurements. Specific methods are reported elsewhere (Brady 2012). I also collected water samples at eight pools (four roadside, four woodland) to assay the concentration of chloride ions using liquid chromatography. Statistical analyses All statistical analyses were conducted in R V. 2.15.0 (R Development Core Team 2012). I employed AIC based selection procedures to compose a suite of mixed effects models to evaluate performance variables across the G x E interaction both for transplant and exposure experiments. I analyzed survival and malformation each as bivariate responses of successes and failures. Growth and developmental rates were derived as exponential functions of change in size and stage, respectively, in relation to the number of days elapsed (e.g. [ln(final size)-ln(initial size)]/period). Initial size was defined as embryo diameter derived from estimates of embryo area, while final size comprised hatchling snout-vent length (SVL). These derived rates were 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 PeerJ reviewing PDF | (v2013:05:486:1:1:NEW 21 Aug 2013) R ev ie w in g M an us cr ip t characterized by Gaussian distributions in the statistical models. All models of performance variables were composed with and without embryo size as a covariate in order to estimate the potential influence of egg size mediated maternal effects (Laugen et al. 2002). Each response variable was evaluated across a suite of models that differed in random effects structure (see Tables S1-S3 in Supporting Information). For each response variable, I evaluated for inference the most parsimonious model with respect to both the significance of interacting main effects and the random effect structure as indicated by lowest AIC score (see Tables S4-S6). I also used mixed effects models to evaluate the influence of pool type on embryo size and fecundity. I used a combination of MCMC randomization methods and log-likelihood approaches to conduct inference (Bolker et al. 2009). I used MANOVA to evaluate the suite of abiotic variables characterizing the roadside and woodland environment. I used a standard linear model to evaluate the influence of pool type on population size, which was scaled to pool area and log-transformed to meet model assumptions. The full details regarding these analyses are described in the Supplementary Text. All data for this study are available at: to be completed after manuscript is accepted for publication. results : Within the field experiment, I found that regardless of the pool environment in which wood frog embryos were reared, the average survival of the roadside deme was lower than that of the woodland deme (Posterior mean = -0.964; 95% HPD = -1.492 \u2013 -0.454; Pmcmc = 0.001). Specifically, survival of embryos originating from the roadside deme averaged 73%, as compared to 86% for the woodland deme (Fig. 2). Thus, across both environments, survival of the roadside deme was 15% lower than that of the woodland deme. I found analogous outcomes when wood frog embryos and larvae were exposed to road salt, a predominant contaminant in roadside pools. Exposure to road salt throughout embryonic development resulted in an increased proportion of 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 PeerJ reviewing PDF | (v2013:05:486:1:1:NEW 21 Aug 2013) R ev ie w in g M an us cr ip t hatchlings with axial malformations (\u03c7\u00b2 = 11.95, df = 8, 2, P = 0.003). At the highest concentration of road salt\u2014corresponding to the maximum detected in roadside pools in this region\u201449% of roadside hatchlings developed axial malformations, as compared to 33% for woodland hatchlings (Fig. 3a); thus, axial malformations were on average 50% more prevalent in the roadside deme. In an acute exposure experiment across a series of five concentrations of roadsalt, survival by the roadside deme averaged just 56% compared to 72% for the woodland deme (Fig. 3b), yielding a 22% relative reduction in survival (Posterior mean = -7.569; 95% HPD = -14.517 \u2013 -0.508; Pmcmc = 0.038). No responses were influenced by embryo size (all P > 0.22, tables S4 and S5). Similarly, embryo size did not differ across deme (MCMC mean = 0.0017; 95% HPD = -0.0006 \u2013 0.0041; Pmcmc = 0.127). Despite a suite of negative consequences of road adjacency and road salt exposure (Fig. 3; see also Supplementary Text and Figs. S1 and S2), I found no difference in the population size of breeding females between roadside and woodland pools (F = 0.033, df = 8, 1, P = 0.861). However, female size and origin interacted such that the number of eggs provisioned by roadside as compared to woodland females increased at a steeper rate with respect to body size (Posterior mean = 21.856 95% HPD = 3.763 \u2013 39.380; Pmcmc = 0.017). Across all populations, female length averaged 53.4 mm SVL. At this size, females from the roadside deme lay 10.5% more eggs than females from the woodland deme (Fig. 4). With regard to abiotic pool conditions, only specific conductance varied with respect to environment (Table S7), reflecting the input of chloride ions from road runoff. There was weak evidence that dissolved oxygen varied between the two environments (T = -2.19, df =8, 1, P = 0.060), with roadside pools averaging 2.38 mg l-1 (\u00b1 0.108 SE) as compared to 2.632 mg l-1 (\u00b1 0.043 SE) in woodland pools. Roadside pools contained an average chloride concentration of 147 mg l-1 as compared to 3.4 mg l-1 in woodland pools, contributing to the 26-fold increase in specific conductance found in roadside pools (Fig. 1 bar graph). 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 PeerJ reviewing PDF | (v2013:05:486:1:1:NEW 21 Aug 2013) R ev ie w in g M an us cr ip t acknowledgements : I am grateful to the Leopold Schepp Foundation for their support of my work through scholarship. I thank D. Skelly, S. Alonzo, P. Turner, M. Urban, A. Hendry, and R. Calsbeek for key guidance and project advice. I am grateful to A. Brady and S. Bolden for extensive support. I thank the reviewers for their constructive and keen feedback on the original manuscript. I thank D. Cholewiak, D. Risch, and D. Sigourney for helpful discussions. S. Attwood, J. Burmeister, G. Antonioli, J. Richardson, M. Manickam, M. Rogalski, H. McMillan all assisted in the field. Z., M., and A. Ladzinkski and S. and R. Brady greatly assisted with infrastructure. J. Bushey provided chemical analysis. The Connecticut Department of Transportation provided road salt for exposure assays. discussion : Despite strong negative effects associated with the roadside environment, I found no evidence of local adaptation, but instead found evidence for maladaptive performance. Specifically, compared to the woodland deme, the roadside deme survived at lower rates, both in the roadside environment and when experimentally exposed to road salt; similarly, the roadside deme accrued more malformations following road salt exposure. Strikingly, the presence of a maladaptive response was not limited to these roadside contexts\u2014even when reared in woodland pools, survival of the roadside deme remained lower than that of the woodland deme (Fig. 2). This suggests that the maladaptive performance induced by the roadside environment further spurs a generalized depression effect on the roadside deme, which manifests independently of the immediate negative influences of both the roadside environment and road salt exposure. Because adult wood frogs exhibit high site fidelity (Berven 1990) and have shown phenotypic differentiation across microgeographic environmental gradients (Skelly 2004), I assumed that the embryos I collected represent a lineage of wood frogs specific to the collection pool (i.e. \u2018population\u2019). While I later discuss the implications of relaxing this assumption, for now I assume that the offspring entered into this experiment represent an independent population potentially connected by some level of dispersal and gene flow. With this in mind, the pattern of maladaptive performance seen here suggests that negative effects associated with exposure to the roadside environment are transmitted from parent to offspring prior to birth. In this way, embryos may be predisposed to reduced performance prior to any direct encounter with the roadside environment. Alternatively, maladaptive performance may have been caused by early embryonic exposure to roadside pools. Specifically, the first 36 hours following natural oviposition but preceding collection\u2014during which time embryos were exposed to their natal environment\u2014may have had strong negative effects on embryos. Though such early exposure alone could explain the putative 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 PeerJ reviewing PDF | (v2013:05:486:1:1:NEW 21 Aug 2013) R ev ie w in g M an us cr ip t maladaptive response, preliminary evidence suggests that early exposure has no such carryover effect in this system (S. P. Brady unpublished data). In addition to any potential effect caused by early exposure, the overall pattern of parentally mediated maladaptive performance described here is consistent with several potential mechanisms, including maladaptation, non-genetic inherited effects (e.g. maternal effects), and demographic processes such as phenotype-biased habitat-oriented dispersal. Maladaptation can be defined most generally as a deviation from an adaptive peak caused by genetic processes (Crespi 2000). As an empirical phenomenon, maladaptation is less common than local adaptation (Crespi 2000; Spitzer 2006). Yet even in reciprocal transplant studies where local adaptation was hypothesized and evaluated, maladaptation frequently occurs. In a review of this literature, Hereford (2009) found the frequency of maladaptation to be 0.29, as compared to the frequency of local adaptation, which was 0.71. Where it has been described, maladaptation at the population level is typically the consequence of migration load (Hanski et al. 2011), resulting when asymmetrical gene flow from non-adapted source populations yields genotypes of sub-optimal fitness in the recipient population (Garcia-Ramos & Kirkpatrick 1997). In the present study however, given the relative survival advantage of the woodland deme of wood frogs, regardless of environment, it seems unlikely that asymmetrical gene flow (i.e. from the woodland deme to the roadside deme) is limiting the local adaptation capacity of the roadside deme. If this were the case, we would expect similar phenotypes between populations grown out within the roadside environment. Likewise, phenotypic homogeneity would be expected if we view maladaptation through the lens of the adaptive phenotypic landscapes framework (Simpson 1944), which predicts that maladaptation is a ubiquitous outcome of dynamic environments because the response to selection always lags one generation behind selection itself; therefore, the adaptive high point is constantly shifting before phenotypes can reach their adaptive peak (Crespi 2000). 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 PeerJ reviewing PDF | (v2013:05:486:1:1:NEW 21 Aug 2013) R ev ie w in g M an us cr ip t However, this process alone does not explain the maladaptation seen here. Specifically, if we assume that the roadside deme did not differ from the woodland deme prior to the recent influence of roads, then we would again expect comparable performance declines between populations when exposed to the roadside environment. In addition to the potential role for genetic based maladaptation, non-genetic processes may be responsible for the observed maladaptive performance of the roadside deme. For example, maternal effects, broadly defined as \u201ca direct effect of a parent\u2019s phenotype on the phenotype of its offspring\u201d (Bernardo 1996) may be associated with the roadside deme\u2019s reduced ability to survive. In this system, if we assume that parents were born in a roadside pool, maternal effects could originate from their exposure to roadside water during their period of aquatic development, which spans newly laid embryos through metamorphosis. Similarly, aspects of the terrestrial roadside environment might propagate negative maternal effects. Yet, I found no evidence that egg size\u2014a maternally mediated trait known to influence amphibian performance (reviewed by Mousseau & Fox 1998)\u2014varied across deme or had any effect on performance variables. Still, maternal effects in this system may be mediated by mechanisms independent of size. For example, nutrient availability and/or allocation of resources to egg quality may differ between these two environments. Alternatively, roadside embryos might accrue chemical contaminants as in utero eggs through maternal transfer. This possibility is supported by outcomes from studies in which coal combustion waste contaminants such as strontium and selenium are maternally transferred, negatively affecting offspring survival and development in two species of toads (Hopkins et al. 2006; Metts et al. 2012). Demographic processes might also lead to maladaptive performance of the roadside deme. For example, if we assume that roadside adults face higher mortality risk\u2014either through carryover 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 PeerJ reviewing PDF | (v2013:05:486:1:1:NEW 21 Aug 2013) R ev ie w in g M an us cr ip t effects from the aquatic environment, or through terrestrial road pressures such as roadkill\u2014then the average age of roadside breeders should be lower, a condition associated with poorer offspring provisioning in wood frogs (Berven 1988). On the other hand, demographic processes might also explain how the roadside deme can persist in spite of the negative effects of the roadside environment and given a predisposition for lower survival. Specifically, across demes, I detected no difference in population size of breeding females. As one possible explanation of this demographic pattern, I found that on average, and adjusted for mean body size, females from the roadside deme lay 10.5% more eggs than females from the woodland deme. Thus while average survival is 15% lower for roadside embryos, the differential in numerical reproductive output per female is less substantial. For example, based on the overall average female size, roadside wood frogs lay 890 eggs per clutch while woodland wood frogs lay 795 eggs per clutch. If we apply the deme respective survival rates (73% versus 86% respectively), we can estimate that for the average sized female, 684 embryos survive to feeding stage in the woodland deme as compared to 650 embryos in the roadside deme. This amounts to just a 5% decrease in reproductive success per female of the roadside deme. Thus, differential reproductive efforts may potentially offset average reductions in offspring survival, mitigating the detriment to demographic success. This increase in reproductive effort may serve to buffer the population against stochastic effects in the face of intense of selection. Ultimately, given the interacting effects of female size and deme (Fig. 4), the adaptive significance of this increased reproductive effort will depend upon the distribution of female size. Similarly, the potential for a tradeoff between quality and quantity of eggs will affect the utility of this apparent strategy. A second demographic possibility is that roadside pools may serve as sink habitats, wherein roadside populations are augmented through immigration from woodland populations, bolstering the chance of persistence. However, here again, if this were true, we might expect that the 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 PeerJ reviewing PDF | (v2013:05:486:1:1:NEW 21 Aug 2013) R ev ie w in g M an us cr ip t migration of woodland individuals into roadside pools would yield homogenized performance responses between the roadside and woodland deme, which is inconsistent with the observed pattern. Still, source-sink dynamics could occur if we assume that putative migrants arriving to breed in roadside pools are themselves depressed in some way. For example, younger or otherwise misfit adults may breed more frequently in the suboptimal conditions found in roadside as compared to woodland pools, potentially through competitive exclusions or the process of \u2018matching habitat choice\u2019 (Edelaar et al. 2008). As an alternative explanation of persistence, it is possible that the negative effects on the aquatic stages of roadside wood frogs might be reversed at later life history stages. For example, the surviving individuals from the roadside deme may produce inferior embryos and larvae, but superior adults. It is surprising to consider that the maladaptive response of the roadside deme was unabated in woodland pools, where survival is typically high. While this generalized depression effect (hereafter \u2018deme depression\u2019) is not well described in the literature (but see Falk et al. 2012), there are conceptual analogues worth considering. As one such example, inbreeding depression can arise in the context of reduced effective population size (Newman & Pilson 1997) that could result from intense selection (Robertson 1961) in the roadside environment. However, inbreeding depression is predicted to be minimal in the context of intense maladaptation (Ronce et al. 2009). Still, roads can act as barriers to migration, which can potentially spur inbreeding depression. However, the types of roads in this system do not appear to limit gene flow between wood frog populations in surrounding landscapes separated by similar or larger distances of space (Richardson 2012). A similar argument for gene flow could therefore be made against genetic drift as a mechanism driving these patterns. As an alternative possibility, genetic mutations or epigenetic modifications could lead to maladaptive responses and overall deme depression. This 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 PeerJ reviewing PDF | (v2013:05:486:1:1:NEW 21 Aug 2013) R ev ie w in g M an us cr ip t may be especially likely in harsh roadside environments, where mutagens such as heavy metals are known to accumulate from road runoff (Trombulak & Frissell 2000). Regardless of mechanism, the maladaptive performance of the roadside deme of wood frogs occurs in direct contrast to the local adaptation of the roadside deme of spotted salamanders breeding and dwelling in the very same pools (Brady 2012). That deme divergence has been detected for two different species occupying identical habitats suggests that population differentiation may be a common consequence of landscape modification. That these outcomes occurred in opposing directions highlights the complex nature of population level responses across species\u2014even when they occupy identical habitats\u2014suggesting that more generalized approaches may be inadequate, and reinforcing the imperative for population specific investigations. Though microgeographic population outcomes have received little attention, there is reason to believe that maladaptive responses may be an emerging and prominent consequence of human environmental modification. For example, phenotypes of species harvested for human consumption have undergone rapid and substantial change (Darimont et al. 2009) in directions believed to oppose adaptation to the increased environmental variation predicted to occur with climate change (Koons 2009). Similarly, hatchery-reared fish have become locally adapted to the hatchery environment at the cost of maladaptation to the wild (Christie et al. 2012). It is worth noting how the conservation implications of these cited examples of maladaptation differ from that of the maladaptive responses reported here. Wherein harvest regimes and hatcheries induce maladaptation through selection as a consequence of a specific management strategy, roads induce maladaptive performance as a collateral effect of global change and for a species that is a bystander to human actions, not a direct target. Thus, even indirect human activities in relatively undeveloped settings can induce maladaptive responses in wild populations. 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 PeerJ reviewing PDF | (v2013:05:486:1:1:NEW 21 Aug 2013) R ev ie w in g M an us cr ip t",
+    "v2_text": "materials and methods : Natural history and site selection The wood frog is widely distributed throughout eastern North America and much of Canada, with a range extending from within the Arctic Circle to the southeastern United States. Within the study region, populations of this species breed in late March or early April, when adults migrate from upland terrestrial habitat into ephemeral wetlands to reproduce. In this study region, females oviposit one egg mass containing approximately 800 embryos per mass. Embryos develop over two-three weeks before hatching, and continue to develop as aquatic larvae throughout spring and early summer until they metamorphose into terrestrial juveniles. Site selection is previously described and resulted in eight of 10 overlapping pools (Brady 2012). The work described here was conducted with permitting from the State of Connecticut Department of Environmental Protection and under protocols from the Yale University Institutional Animal Care and Use Committee (Projects 2006-11024 and 2009-11024). Reciprocal transplant experiment In spring 2008, I collected from each of 10 pools (five roadside, five woodland) a subset of embryos from 10 egg masses less than 36 hours old. To ensure collection within this timeframe, pools were monitored daily as the breeding season approached. Roadside pools were located < 10 m from a paved road while woodland pools were located > 200 m from the nearest paved road. While roadside egg masses appeared somewhat more compact than woodland egg masses 4 4 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 PeerJ reviewing PDF | (v2013:05:486:0:0:NEW 10 May 2013) R ev ie w in g M an us cr ip t (Karraker & Gibbs 2011), there were no conspicuous differences in the embryos themselves between sites, and selection of embryos from within each egg mass was haphazard. Thus, any potential effect of the < 36 hour exposure did not bias the subset of embryos employed in the experiment. Each egg mass was collected whole into a plastic 710 ml container. From each egg mass, I dissected two clusters of embryos. Each cluster was photographed and stocked into one of five experimental enclosures in the origin pool and the transplant pool [enclosure details described elsewhere (Brady 2012)]. For a subset of egg masses, this procedure occurred in the field until weather conditions proved limiting; remaining egg masses were transported to the lab on ice for processing, stored overnight in an 8 \u00b0C incubator, and returned for stocking the following day. Each pool contained five experimental blocks. I targeted stocking approximately 50 embryos from each of two egg masses per enclosure, yielding one unique pairing of egg masses (hereafter \u201cclutch pair\u201d) that was replicated across, but not within pool pairs. This design was chosen to maximize family level diversity while maintaining an additional level of replication at the clutch pair level, while at the same time balancing logistics of resources and spatial constraints within pools. Enclosure assignment was haphazard for each clutch pair. In total across 10 pools, I stocked 100 enclosures containing approximately 100 embryos each. I used a dissecting stereoscopic microscope to estimate the developmental stage of each egg mass upon collection from the field. For each embryo cluster (n = 200) I estimated egg size from photographs. Specifically, I used ImageJ (Rasband 1997-2012)to measure the two-dimensional surface area of 15 embryos per cluster represented by a best-fit ellipse (Brady 2012). At the conclusion of the experiment, when all eggs had either hatched and reached feeding stage or died, I estimated survival, developmental stage (Gosner 1960), and snout-vent length (SVL) at the container level. One enclosure was removed from analysis due to predator intrusion. 5 5 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 PeerJ reviewing PDF | (v2013:05:486:0:0:NEW 10 May 2013) R ev ie w in g M an us cr ip t Chronic road salt exposure experiment Of the 10 egg masses collected, nine were haphazardly assigned into groups of three (hereafter \u201cclutch triad\u201d). From each egg mass, three separate clusters\u2014each of approximately 30 embryos \u2014were dissected out and haphazardly assigned by clutch triad to each of three road salt treatments. Each cluster (n = 270) was photographed and embryo size was estimated as described above. Each experimental unit comprised one clutch triad, the unique composition of which was maintained across treatments providing an additional random effect. In total, each of 10 pools contributed representatives from nine egg masses, grouped in clutches of three, and exposed to three treatments for a total of 90 experimental units (3 clutch triads per treatment x 3 treatments x 10 pools = 90). The three treatments (low, medium, high) reflected the range of specific conductance found in the study region. All treatments comprised aged and conditioned tap water. Road salt comprising at least 95% NaCl obtained from the Connecticut Department of Transportation was added to the medium and high treatments to achieve specific conductance values of 1,000 \u00b5S and 4,000 \u00b5S respectively, which approximate the mean and maximum values I detected in roadside pools in the study region. No road salt was added to the low treatment, which had a natural specific conductance of approximately 175 \u00b5S. Embryos were reared in 11.3 l plastic containers (42 x 28 x 18 cm) filled with approximately 8 l of assigned treatment water. Each container was fitted with mesh hardware cloth to suspend embryos in the water column off the bottom of the container. Containers were housed on tables in an outdoor facility under 50% shade cloth. Concentrations of roadsalt were maintained in response to evaporation and rain by adding lower or higher concentration treatment water respectively, and through routine, complete 6 6 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 PeerJ reviewing PDF | (v2013:05:486:0:0:NEW 10 May 2013) R ev ie w in g M an us cr ip t water changes. When all embryos hatched or died, I estimated survival, developmental stage, SVL, and prevalence of axial malformations (Bantle et al. 1991). Acute road salt exposure experiment Following assignment of embryo clusters to reciprocal transplant and chronic exposure experiments described above, egg masses were retained within collection containers and placed haphazardly within in 150 l plastic wading pools that served as incubators, located in an outdoor enclosure covered with 50% shade cloth. When embryos reached developmental stage 18, I dissected approximately 45 embryos from each egg mass and placed them into a common garden environment consisting of a 150 l wading pool filled with aged, conditioned tap water. Embryos were grouped by origin such that each wading pool contained embryos from each of the 10 egg masses representing a single natural pool. Upon approaching feeding stage [Gosner developmental stage 25 (Gosner 1960)], larvae were fed a mixture of 3:1 ground rabbit chow and fish flakes at a daily ration of 10% average body mass. When all individuals appeared to reach feeding stage, I exposed wood frog larvae to a series of five concentrations of road salt, ranging from 0.5 to 10.0 g/l. These concentrations were selected to represent previously documented acute exposure assays of the wood frog (Sanzo & Hecnar 2006). From each pool, 10 larvae were randomly stocked into each treatment and replicated across four experimental units per treatment (10 pools x 5 concentrations x 4 replicates = 200 experimental units). A 5.1 l container (33 x 20 x 11 cm) filled with 4 l of treatment water comprised each experimental unit. Larvae in each experimental unit were given a four-day food ration at the start of the experiment. After 96 hours of exposure, I evaluated larval survival. In order to standardize developmental stage of larvae in 7 7 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 PeerJ reviewing PDF | (v2013:05:486:0:0:NEW 10 May 2013) R ev ie w in g M an us cr ip t the experiment, exposure was conducted in two separate, consecutive 96-hour rounds\u2014the first comprising six pair-wise pools and the second comprising four pair-wise pools. Population size and female fecundity In spring 2008 following the conclusion of natural oviposition, I conducted visual counts of wood frog egg masses in each of the 10 pools to estimate population size of breeding females. In spring 2010, I collected 56 pairs of inbound wood frog adults from a subset of six pools (three roadside, three woodland) and haphazardly assigned male-female combinations grouped by origin pool. Pairs were placed in 11.3 l plastic containers filled with 2.5 l of aged spring water, equipped with a small branch and one white oak leaf (Quercus alba) to provide a substrate for oviposition. Containers were positioned on a slight angle to create a depth cline, and adults were allowed to amplex and oviposit. Following oviposition, egg masses were photographed and fecundity was estimated from each photo using ImageJ (Rasband 1997-2012)to manually count the total number of eggs oviposited per female. Characterizing the environment In each pool, I measured seven environmental characteristics associated with amphibian distribution and performance (Wellborn et al. 1996). Specific conductance, dissolved oxygen, pH, and wetland depth were measured once during the experiment, while temperature was measured every thirty minutes using deployed temperature loggers. Global site factor\u2014a measure of solar radiation reaching the pool\u2014was calculated from hemispherical photographs, while wetland area was estimated from visual rangefinder measurements. Specific methods are 8 8 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 PeerJ reviewing PDF | (v2013:05:486:0:0:NEW 10 May 2013) R ev ie w in g M an us cr ip t reported elsewhere (Brady 2012). I also collected water samples at eight pools (four roadside, four woodland) to assay the concentration of chloride ions using liquid chromatography. Statistical analyses All statistical analyses were conducted in R V. 2.15.0 (R Development Core Team 2012). I employed AIC based selection procedures to compose a suite of mixed effects models to evaluate performance variables across the G x E interaction both for transplant and exposure experiments. I analyzed survival and malformation each as bivariate responses of successes and failures. Growth and developmental rates were derived as exponential functions of change in size and stage, respectively, in relation to the number of days elapsed (e.g. [ln(final size)-ln(initial size)]/period). Initial size was defined as embryo diameter calculated from estimates of embryo area, while final size comprised hatchling snout-vent length (SVL). These derived rates were characterized by Gaussian distributions in the statistical models. All models of performance variables were composed with and without embryo size as a covariate in order to estimate the potential influence of egg size mediated maternal effects (Laugen et al. 2002). Each response variable was evaluated across a suite of models that differed in random effects structure (see Tables S1-S3 in Supporting Information). For each response variable, I evaluated for inference the most parsimonious model with respect to both the significance of interacting main effects and the random effect structure as indicated by lowest AIC score (see Tables S4-S6). I also used mixed effects models to evaluate the influence of pool type on embryo size and fecundity. I used a combination of MCMC randomization methods and log-likelihood approaches to conduct inference (Bolker et al. 2009). I used MANOVA to evaluate the suite of abiotic variables characterizing the roadside and woodland environment. I used a standard linear model to 9 9 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 PeerJ reviewing PDF | (v2013:05:486:0:0:NEW 10 May 2013) R ev ie w in g M an us cr ip t evaluate the influence of pool type on population size, which was scaled to pool area and log-transformed to meet model assumptions. All data for this study are available at: to be completed after manuscript is accepted for publication results : Within the field experiment, I found that regardless of the pool environment in which wood frog embryos were reared, the average survival of the roadside deme was lower than that of the woodland deme (Posterior mean = -0.964; 95% CI = -1.492 \u2013 -0.454; Pmcmc = 0.001). Specifically, survival of embryos originating from the roadside deme averaged 73%, as compared to 86% for the woodland deme (Fig. 2). Thus, across both environments, survival of the roadside deme was 15% lower than that of the woodland deme. I found analogous outcomes when wood frog embryos and larvae were exposed to road salt, a predominant contaminant in roadside pools. Exposure to road salt throughout embryonic development resulted in an increased proportion of hatchlings with axial malformations (\u03c7\u00b2 = 11.95, df = 8, 2, P = 0.003). At the highest concentration of road salt\u2014corresponding to the maximum detected in roadside pools in this region\u201449% of roadside hatchlings developed axial malformations, as compared to 33% for woodland hatchlings (Fig. 3a); thus, axial malformations were on average 50% more prevalent in the roadside deme. In an acute exposure experiment across a series of five concentrations of roadsalt, survival by the roadside deme averaged just 56% compared to 72% for the woodland deme (Fig. 3b), yielding a 22% relative reduction in survival (Posterior mean = -7.569; 95% CI = -14.517 \u2013 -0.508; Pmcmc = 0.038). No responses were influenced by embryo size (all P > 0.22, tables S4 and S5). Similarly, embryo size did not differ across deme (MCMC mean = 0.0017; 95% HPD = -0.0006 \u2013 0.0041; Pmcmc = 0.127). Despite a suite of negative 10 10 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 PeerJ reviewing PDF | (v2013:05:486:0:0:NEW 10 May 2013) R ev ie w in g M an us cr ip t consequences of road adjacency and road salt exposure (Fig. 3; see also Supplementary Text and Figs. S1 and S2), I found no difference in the population size of breeding females between roadside and woodland pools (F = 0.033, df = 8, 1, P = 0.861). However, adjusted for body size, females from the roadside deme lay 10.5% more eggs than females from the woodland deme (MCMC mean = 21.856 95% HPD = 3.763 \u2013 39.380; Pmcmc = 0.017; Fig. 4). With regard to abiotic pool conditions, only specific conductance varied with respect to environment, reflecting the input of chloride ions from road runoff. Roadside pools contained an average chloride concentration of 147 mg l-1 as compared to 3.4 mg l-1 in woodland pools, contributing to the 26-fold increase in specific conductance found in roadside pools (Fig. 1 bar graph). acknowledgements : I am grateful to the Leopold Schepp Foundation for their support of my work through scholarship. I thank D. Skelly, S. Alonzo, P. Turner, M. Urban, A. Hendry, and R. Calsbeek for key guidance and project advice. I am grateful to A. Brady and S. Bolden for extensive support. S. Attwood, J. Burmeister, G. Antonioli, J. Richardson, M. Manickam, M. Rogalski, H. McMillan all assisted in the field. Z., M., and A. Ladzinkski and S. and R. Brady greatly assisted with infrastructure. J. Bushey provided chemical analysis. The Connecticut Department of Transportation provided road salt for exposure assays. discussion : Despite strong selection on roadside wood frogs, I found no evidence of local adaptation, but instead found evidence for maladaptive performance. Compared to the woodland deme, the roadside deme survived at lower rates, both in the roadside environment and when experimentally exposed to road salt; similarly, the roadside deme accrued more malformations following road salt exposure. Taken together, these results indicate that the negative consequences of roads are inherited from parent to offspring, predisposing embryos to reduced performance. Strikingly, these maladaptive responses are not limited to the roadside context. Even when reared in woodland pools, survival of the roadside deme remains lower than that of the woodland deme (Fig. 2). This suggests that the maladaptive performance induced by the roadside environment further spurs a generalized depression effect on the roadside deme, which occurs independently of the immediate negative influence of the roadside environment. 11 11 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 PeerJ reviewing PDF | (v2013:05:486:0:0:NEW 10 May 2013) R ev ie w in g M an us cr ip t The overall pattern of inherited maladaptive performance described here is consistent with several potential mechanisms, including maladaptation, non-genetic inherited effects (e.g. maternal effects), and demographic processes such as habitat-oriented dispersal. Maladaptation can be defined most generally as a deviation from an adaptive peak caused by genetic processes (Crespi 2000). As an empirical phenomenon, maladaptation has rarely been reported among wild populations (Crespi 2000; Spitzer 2006). Where it has been described, maladaptation at the deme level is typically the consequence of migration load (Hanski et al. 2011), resulting when asymmetrical gene flow from non-adapted source populations yields genotypes of sub-optimal fitness in the recipient population (Garcia-Ramos & Kirkpatrick 1997). However, given the relative survival advantage of the woodland deme of wood frogs, regardless of environment, it seems unlikely that asymmetrical gene flow (i.e. from the woodland deme to the roadside deme) is limiting the local adaptation capacity of the roadside deme. If this were the case, we would expect similar phenotypes between demes grown out within the roadside environment. Likewise, phenotypic homogeneity would be expected if we view maladaptation through the lens of the adaptive phenotypic landscapes framework (Simpson 1944), which predicts that maladaptation is a ubiquitous outcome of dynamic environments because the response to selection always lags one generation behind selection itself; therefore, the adaptive high point is constantly shifting before phenotypes can reach their adaptive peak (Crespi 2000). However, this process alone does not explain the maladaptation seen here. Specifically, if we assume that the roadside deme did not differ from the woodland deme prior to the recent influence of roads, then we would again expect comparable performance declines between demes when exposed to the roadside environment. 12 12 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 PeerJ reviewing PDF | (v2013:05:486:0:0:NEW 10 May 2013) R ev ie w in g M an us cr ip t In addition to the potential role for genetic based maladaptation, non-genetic processes may be responsible for the observed maladaptive performance of the roadside deme. For example, maternal effects, broadly defined as \u201ca direct effect of a parent\u2019s phenotype on the phenotype of its offspring\u201d (Bernardo 1996) may be associated with the roadside deme\u2019s reduced ability to survive. Yet, I found no evidence that egg size\u2014a maternally mediated trait known to influence amphibian performance (reviewed by Mousseau & Fox 1998)\u2014varied across deme or had any effect on performance variables. Still, maternal effects in this system may be mediated chemically, rather than by size, wherein roadside embryos might accrue contaminants through maternal transfer. This possibility is supported by outcomes from studies in which coal combustion waste contaminants such as strontium and selenium are maternally transferred, negatively affecting offspring survival and development in two species of toads (Hopkins et al. 2006; Metts et al. 2012). Demographic processes might also lead to maladaptive performance of the roadside deme. For example, if we assume that roadside adults face higher mortality risk\u2014either through carryover effects from the aquatic environment, or through terrestrial road pressures such as roadkill\u2014then the average age of roadside breeders should be lower, a condition associated with poorer offspring provisioning in wood frogs (Berven 1988). On the other hand, demographic processes might also explain how the roadside deme can persist in spite of the negative effects of the roadside environment and given a predisposition for lower survival. Specifically, across demes, I detected no difference in population size of breeding females. As one possible explanation of this pattern, I found that on average, and adjusted for body size, females from the roadside deme lay 10.5% more eggs than females from the woodland deme. Thus, differential reproductive efforts 13 13 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 PeerJ reviewing PDF | (v2013:05:486:0:0:NEW 10 May 2013) R ev ie w in g M an us cr ip t may potentially offset average reductions in offspring survival, mitigating the detriment to demographic success. A second possibility is that roadside pools may serve as sink habitats, wherein roadside populations are augmented through immigration from the woodland deme, bolstering the chance of persistence. However, here again, if this were true, we might expect that the migration of woodland individuals into roadside pools would yield homogenized performance responses between the roadside and woodland deme, which is inconsistent with the observed pattern. Still, source-sink dynamics could occur if we assume that putative migrants arriving to breed in roadside pools are themselves depressed in some way. For example, younger or otherwise misfit adults may preferentially breed in the suboptimal conditions found in roadside pools, potentially through competitive exclusions or the process of \u2018matching habitat choice\u2019 (Edelaar et al. 2008). As an alternative explanation of persistence, it is possible that the negative effects on the aquatic stages of roadside wood frogs might be reversed at later life history stages. For example, the surviving individuals from the roadside deme may produce inferior embryos and larvae, but superior adults. It is surprising to consider that the maladaptive response of the roadside deme was unabated in woodland pools, where survival is typically high. While this generalized depression effect (hereafter \u2018deme depression\u2019) is not described in the literature, there are conceptual analogues worth considering. As one such example, inbreeding depression can arise in the context of reduced effective population size (Newman & Pilson 1997) that could result from intense selection (Robertson 1961) in the roadside environment. However, inbreeding depression is predicted to be minimal in the context of intense maladaptation (Ronce et al. 2009). Still, roads can act as barriers to migration, which can potentially spur inbreeding depression. However, the 14 14 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 PeerJ reviewing PDF | (v2013:05:486:0:0:NEW 10 May 2013) R ev ie w in g M an us cr ip t types of roads in this system do not appear to limit gene flow between wood frog demes in surrounding landscapes separated by similar or larger distances of space (Richardson 2012). A similar argument for gene flow could therefore be made against genetic drift as a mechanism driving these patterns. As an alternative possibility, genetic mutations or epigenetic modifications could lead to maladaptive responses and overall deme depression. This may be especially likely in harsh roadside environments, where mutagens such as heavy metals are known to accumulate from road runoff (Trombulak & Frissell 2000). Regardless of mechanism, the maladaptive performance of the roadside deme of wood frogs occurs in direct contrast to the local adaptation of the roadside deme of spotted salamanders breeding and dwelling in the very same pools (Brady 2012). That deme divergence has been detected for two different species occupying identical habitats suggests that population differentiation may be a common consequence of landscape modification. That these outcomes occurred in opposing directions highlights the complex nature of deme level responses across species\u2014even when they occupy identical habitats\u2014suggesting that more generalized approaches may be inadequate, and reinforcing the imperative for deme specific investigations. Though microgeographic deme outcomes have received little attention, there is reason to believe that maladaptive responses may be an emerging and prominent consequence of human environmental modification. For example, phenotypes of species harvested for human consumption have undergone rapid and substantial change (Darimont et al. 2009) in directions believed to oppose adaptation to the increased environmental variation predicted to occur with climate change (Koons 2009). Similarly, hatchery-reared fish have become locally adapted to the 15 15 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 PeerJ reviewing PDF | (v2013:05:486:0:0:NEW 10 May 2013) R ev ie w in g M an us cr ip t hatchery environment at the cost of maladaptation to the wild (Christie et al. 2012). It is worth noting how the conservation implications of these cited examples of maladaptation differ from that of the maladaptive responses reported here. Wherein harvest regimes and hatcheries induce maladaptation through selection as a consequence of a specific management strategy, roads induce maladaptive performance as a collateral effect of global change and for a species that is a bystander to human actions, not a direct target. Thus, even indirect human activities in relatively undeveloped settings can induce maladaptive responses in wild populations.",
+    "url": "https://peerj.com/articles/164/reviews/",
+    "review_1": "Harry Hochheiser \u00b7 Aug 29, 2013 \u00b7 Academic Editor\nACCEPT\nThank you for your prompt responses to the reviewers' comments.",
+    "review_2": "Reviewer 1 \u00b7 Aug 25, 2013\nBasic reporting\nThe authors have addressed my previous comments in a satisfactory manner.\nExperimental design\nThe authors have addressed my previous comments in a satisfactory manner.\nValidity of the findings\nThe authors have addressed my previous comments in a satisfactory manner.\nAdditional comments\nThe authors have addressed my previous comments in a satisfactory manner.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Using cloud-based mobile technology for assessment of competencies among medical students (v0.2)\". PeerJ https://doi.org/10.7287/peerj.164v0.2/reviews/1",
+    "review_3": "Harry Hochheiser \u00b7 Jul 23, 2013 \u00b7 Academic Editor\nMAJOR REVISIONS\nThe reviewers seem to be in agreement that additional detail regarding the stastical design and analyses would be helpful for understanding the experiment and its results. Both reviewers provided suggestions for additional detail that would make this paper both clearer and more informative. Please consider these comments in your revisions.",
+    "review_4": "Reviewer 1 \u00b7 Jul 23, 2013\nBasic reporting\nThe authors have met standards.\nExperimental design\nThe authors hypothesized that JIT would: 1) facilitate the\ndirect observation and provision of feedback to trainees on their clinical competencies; 2) generally be accepted by faculty; 3) provide a means for recording the observations of trainee performance, and 4) possess adequate reliability and validity.\n\nTo evaluate these hypotheses, the authors collected following data: the specific training problems and competencies observed and assessed by the evaluators, the grades associated with the observation and descriptive data from faculty on the use of JIT.\n\nThe experimental design appears adequate, however I do wish more qualitative data regarding general acceptance by faculty would have been collected an analyzed.\n\nThe explanation of the statistical methods is inadequate and fails to meet standards. For instance, the authors fail to describe adequate methods regarding the statistical computation of inter-rater reliability. This outcome can be assessed through a number of statistical tests, and the specific method should be described. In addition, correlation coefficient was computed and presented, but the method was not described (e.g. Spearman's or Pearson's).\nValidity of the findings\nThe findings generally support determination of the four hypotheses described by the authors. I would like to see more data to support hypothesis #2 (acceptance by faculty) and a more thorough description of the statistical methodology in order to assess the conclusion of hypothesis #4 (possess adequate reliability and validity).\n\nIt would be nice for the authors to identify one primary outcome measure for the study (which of the four stated hypotheses was the major focus of this study?).\n\nIt appears a very specific number (17) of evaluators were selected to test for reliability and validity. How did the authors arrive at this number of evaluators? Did they compute a sample size based on expected reliability and minimum reliability? What was the method by which they computed the sample size of 17 evaluators for the reliability testing? Was the group of 17 evaluators a convenience sample or were they a specific target group? What was the demographic composition of the group (it appears some residents were used as evaluators, but the number is never mentioned). There does not appear to be sufficient operational detail regarding the specifics of this aspect of the study for a reader to reproduce the results.\nAdditional comments\nThis paper reports the use of a tool for enabling educators to assess medical student performance at the point of care using a mobile-responsive web-based application. While the application itself is not novel, the implementation and execution of the concept appears elegant.\n\nSome areas for future work: 1) integration into institution-centric learning management systems, 2) incorporation of reflective learning opportunities for students to self-assess their own performance, 3) ability to provide audio or video-based feedback for learners using the mobile interface, 4) ability for evaluators to use their own devices for logging feedback rather than using the student's device.\n\nA qualitative analysis of user experience and feedback from learners about this system would be useful to gauge difficulties with the system and/or ways to scale implementation of this tool widely throughout the institution.\n\nThe manuscript should be revised to provide a more detailed description of the statistical methods used in the paper. More details about the reliability and validity testing should also be included.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Using cloud-based mobile technology for assessment of competencies among medical students (v0.1)\". PeerJ https://doi.org/10.7287/peerj.164v0.1/reviews/1",
+    "pdf_1": "https://peerj.com/articles/164v0.2/submission",
+    "pdf_2": "https://peerj.com/articles/164v0.1/submission",
+    "review_5": "Reviewer 2 \u00b7 Jul 15, 2013\nBasic reporting\nThe authors describe a cloud-based application for assessment of competencies in medical students. The application was implemented and feasibility evaluations were done.\n\nIn the Introduction, provide some background description of CEX with some references since readers may not be familiar with CEX. An illustrative example of a CEX will also be useful.\n\nMinor points:\nUse lower case \u201cc\u201d in Cloud-based\nUse \u201capplication\u201d instead of the abbreviation \u201capp\u201d\nExperimental design\nIn the Materials & Methods section:\n\nMention for what platforms the application is available (iOS, Android, etc.)\n\nWhat cut-offs were used to determine the three grades that were generated by the algorithm?\n\nGive details about how CEX assessment time was measured.\n\nGive details of the instrument/questions that were used to assess the faculty\u2019s satisfaction with the app.\n\nProvide the statistic that was used to measure inter-rater reliability and the statistical software used to do the analysis.\n\nProvide details of the statistic and the statistical software used for analyzing the correlation between the \u201cgateway\u201d performance assessment examinations and the CEX assessments.\nValidity of the findings\nIn the Results section:\n\nIt will be useful to provide more detailed results of grades, time and satisfaction for the 3567 assessments. Bar charts for each of the 3 variables that show the percent breakdown will be useful.\n\nFor interrater reliability provide 2-sided confidence intervals along with the statistic for faculty alone, residents alone and faculty and residents combined.\nAdditional comments\nAre there tools (e.g. paper based instruments) that have been described in the literature for assisting in CEX assessments? If so, it would be useful to include brief descriptions of them in the Discussion section.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"Using cloud-based mobile technology for assessment of competencies among medical students (v0.1)\". PeerJ https://doi.org/10.7287/peerj.164v0.1/reviews/2",
+    "all_reviews": "Review 1: Harry Hochheiser \u00b7 Aug 29, 2013 \u00b7 Academic Editor\nACCEPT\nThank you for your prompt responses to the reviewers' comments.\nReview 2: Reviewer 1 \u00b7 Aug 25, 2013\nBasic reporting\nThe authors have addressed my previous comments in a satisfactory manner.\nExperimental design\nThe authors have addressed my previous comments in a satisfactory manner.\nValidity of the findings\nThe authors have addressed my previous comments in a satisfactory manner.\nAdditional comments\nThe authors have addressed my previous comments in a satisfactory manner.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Using cloud-based mobile technology for assessment of competencies among medical students (v0.2)\". PeerJ https://doi.org/10.7287/peerj.164v0.2/reviews/1\nReview 3: Harry Hochheiser \u00b7 Jul 23, 2013 \u00b7 Academic Editor\nMAJOR REVISIONS\nThe reviewers seem to be in agreement that additional detail regarding the stastical design and analyses would be helpful for understanding the experiment and its results. Both reviewers provided suggestions for additional detail that would make this paper both clearer and more informative. Please consider these comments in your revisions.\nReview 4: Reviewer 1 \u00b7 Jul 23, 2013\nBasic reporting\nThe authors have met standards.\nExperimental design\nThe authors hypothesized that JIT would: 1) facilitate the\ndirect observation and provision of feedback to trainees on their clinical competencies; 2) generally be accepted by faculty; 3) provide a means for recording the observations of trainee performance, and 4) possess adequate reliability and validity.\n\nTo evaluate these hypotheses, the authors collected following data: the specific training problems and competencies observed and assessed by the evaluators, the grades associated with the observation and descriptive data from faculty on the use of JIT.\n\nThe experimental design appears adequate, however I do wish more qualitative data regarding general acceptance by faculty would have been collected an analyzed.\n\nThe explanation of the statistical methods is inadequate and fails to meet standards. For instance, the authors fail to describe adequate methods regarding the statistical computation of inter-rater reliability. This outcome can be assessed through a number of statistical tests, and the specific method should be described. In addition, correlation coefficient was computed and presented, but the method was not described (e.g. Spearman's or Pearson's).\nValidity of the findings\nThe findings generally support determination of the four hypotheses described by the authors. I would like to see more data to support hypothesis #2 (acceptance by faculty) and a more thorough description of the statistical methodology in order to assess the conclusion of hypothesis #4 (possess adequate reliability and validity).\n\nIt would be nice for the authors to identify one primary outcome measure for the study (which of the four stated hypotheses was the major focus of this study?).\n\nIt appears a very specific number (17) of evaluators were selected to test for reliability and validity. How did the authors arrive at this number of evaluators? Did they compute a sample size based on expected reliability and minimum reliability? What was the method by which they computed the sample size of 17 evaluators for the reliability testing? Was the group of 17 evaluators a convenience sample or were they a specific target group? What was the demographic composition of the group (it appears some residents were used as evaluators, but the number is never mentioned). There does not appear to be sufficient operational detail regarding the specifics of this aspect of the study for a reader to reproduce the results.\nAdditional comments\nThis paper reports the use of a tool for enabling educators to assess medical student performance at the point of care using a mobile-responsive web-based application. While the application itself is not novel, the implementation and execution of the concept appears elegant.\n\nSome areas for future work: 1) integration into institution-centric learning management systems, 2) incorporation of reflective learning opportunities for students to self-assess their own performance, 3) ability to provide audio or video-based feedback for learners using the mobile interface, 4) ability for evaluators to use their own devices for logging feedback rather than using the student's device.\n\nA qualitative analysis of user experience and feedback from learners about this system would be useful to gauge difficulties with the system and/or ways to scale implementation of this tool widely throughout the institution.\n\nThe manuscript should be revised to provide a more detailed description of the statistical methods used in the paper. More details about the reliability and validity testing should also be included.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Using cloud-based mobile technology for assessment of competencies among medical students (v0.1)\". PeerJ https://doi.org/10.7287/peerj.164v0.1/reviews/1\nReview 5: Reviewer 2 \u00b7 Jul 15, 2013\nBasic reporting\nThe authors describe a cloud-based application for assessment of competencies in medical students. The application was implemented and feasibility evaluations were done.\n\nIn the Introduction, provide some background description of CEX with some references since readers may not be familiar with CEX. An illustrative example of a CEX will also be useful.\n\nMinor points:\nUse lower case \u201cc\u201d in Cloud-based\nUse \u201capplication\u201d instead of the abbreviation \u201capp\u201d\nExperimental design\nIn the Materials & Methods section:\n\nMention for what platforms the application is available (iOS, Android, etc.)\n\nWhat cut-offs were used to determine the three grades that were generated by the algorithm?\n\nGive details about how CEX assessment time was measured.\n\nGive details of the instrument/questions that were used to assess the faculty\u2019s satisfaction with the app.\n\nProvide the statistic that was used to measure inter-rater reliability and the statistical software used to do the analysis.\n\nProvide details of the statistic and the statistical software used for analyzing the correlation between the \u201cgateway\u201d performance assessment examinations and the CEX assessments.\nValidity of the findings\nIn the Results section:\n\nIt will be useful to provide more detailed results of grades, time and satisfaction for the 3567 assessments. Bar charts for each of the 3 variables that show the percent breakdown will be useful.\n\nFor interrater reliability provide 2-sided confidence intervals along with the statistic for faculty alone, residents alone and faculty and residents combined.\nAdditional comments\nAre there tools (e.g. paper based instruments) that have been described in the literature for assisting in CEX assessments? If so, it would be useful to include brief descriptions of them in the Discussion section.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"Using cloud-based mobile technology for assessment of competencies among medical students (v0.1)\". PeerJ https://doi.org/10.7287/peerj.164v0.1/reviews/2\nReview 6: \n",
+    "LLM_extracted_review": "['Review 1: The authors have addressed my previous comments in a satisfactory manner.\\n\\nReview 2: The authors have addressed my previous comments in a satisfactory manner.\\n\\nReview 3: The reviewers seem to be in agreement that additional detail regarding the statistical design and analyses would be helpful for understanding the experiment and its results.\\n\\nReview 4: The experimental design appears adequate, however I do wish more qualitative data regarding general acceptance by faculty would have been collected and analyzed. The explanation of the statistical methods is inadequate and fails to meet standards. I would like to see more data to support hypothesis #2 (acceptance by faculty) and a more thorough description of the statistical methodology in order to assess the conclusion of hypothesis #4 (possess adequate reliability and validity). It would be nice for the authors to identify one primary outcome measure for the study. There does not appear to be sufficient operational detail regarding the specifics of this aspect of the study for a reader to reproduce the results. A qualitative analysis of user experience and feedback from learners about this system would be useful to gauge difficulties with the system and/or ways to scale implementation of this tool widely throughout the institution. The manuscript should be revised to provide a more detailed description of the statistical methods used in the paper.\\n\\nReview 5: In the Introduction, provide some background description of CEX with some references since readers may not be familiar with CEX. Mention for what platforms the application is available (iOS, Android, etc.). What cut-offs were used to determine the three grades that were generated by the algorithm? Give details about how CEX assessment time was measured. Provide the statistic that was used to measure inter-rater reliability and the statistical software used to do the analysis. It will be useful to provide more detailed results of grades, time and satisfaction for the 3567 assessments. For interrater reliability provide 2-sided confidence intervals along with the statistic for faculty alone, residents alone and faculty and residents combined. If there are tools that have been described in the literature for assisting in CEX assessments, it would be useful to include brief descriptions of them in the Discussion section.']"
+}

peerj_json_files/PeerJ_Json_12.json ADDED Viewed

	@@ -0,0 +1,14 @@

+{
+    "v1_Abstract": "Copepods as feed promote better growth and development in marine fish larvae than rotifers. However, unlike rotifers, copepods contain several minerals such as iodine (I), at potentially toxic levels. Iodine is an essential trace element and both under and over supply of I can inhibit the production of the I containing thyroid hormones. It is unknown whether marine fish larvae require copepod levels of I or if mechanisms are present that prevent I toxicity. In this study, larval Atlantic cod (Gadus morhua) were fed rotifers enriched to intermediate (26 mg I kg dry weight; MI group) or copepod (129 mg I kg DW; HI group) I levels and compared to cod larvae fed control rotifers (0.6 mg I kg DW). Larval I concentrations were increased by 3 (MI) and 7 (HI) fold compared to controls during the rotifer feeding period. No differences in growth were observed, but the HI diet increased thyroid follicle colloid to epithelium ratios, and effected the essential element concentrations of larvae compared to the other groups. The thyroid follicle morphology in the HI larvae is typical of colloid goitre, a condition resulting from excessive I intake, even though whole body I levels were below those found previously in copepod fed cod larvae. This is the first observation of dietary induced I toxicity in fish, and suggests I toxicity may be determined to a greater extent by bioavailability and nutrient interactions than by total body I concentrations in fish larvae. Rotifers with 0.6 mg I kg DW appeared sufficient to prevent gross signs of I deficiency in cod larvae reared with continuous water exchange, while modelling of cod larvae versus rotifer I levels suggests that optimum I levels in rotifers for cod larvae is 3.5 mg I kg DW.",
+    "v1_text": "acknowledgements : This work was financed by the Norwegian Research Council (project no. 185006/S40). Thank you to technical staff at IMR Austevoll and NIFES for fish husbandry, sampling help and skilled analytical assistance, especially Stig Ove Utskot and Berit Solli. Thank you to Karin Pittman for allowing access to equipment for thyroid follicle morphology quantification. Contributions of the authors. S. P., K. H., A. N. and T.H designed the project, T. H. conducted animal trials, S. P., K. H. and T. H. sampled. S. P., O.S. and S.H. analysed samples. S. P. statistically analysed data. All authors contributed to manuscript writing and revision. Conflicts of interest. None. 16 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 Pre Prin ts Pre Prin ts 2 cod larvae rearing : The experiment was performed at the Institute of Marine Research (IMR), Austevoll Research Station, Norway. This study was carried out within the Norwegian animal welfare act guidelines (code 750.000) at an approved facility. As this trial was assumed to be a nutrition trial based on all available studies up to the date of the trial, no specific permit was required under the guidelines. Naturally spawned and fertilised Atlantic cod eggs were obtained from in house second generation brood stock. Prior to incubation, eggs were disinfected with 200 mg L-1 glutaraldehyde for 9 min at 6\u00b0C and eggs were incubated with a standard protocol as described in Penglase et al. (2010). Upon 100% hatching (16 dpf, 99 day degrees), larval density in the incubators was measured via tube sampling and ranged from 2000-3000 larvae L-1. Three days post 100% hatching (dph), larvae (65 \u00b1 2 \u00b5g DW fish-1, n = 2 where n is a pool of 428 or 520 fish, and 4.9 \u00b1 0.2 mm fish-1, n = 30 (mean \u00b1 SD), measured 5 dph) were transferred into the experimental tanks. Larvae were stocked at an estimated density of 50 000 larvae (120 larvae L -1) in each of the nine 500 L (400 L water volume) experimental tanks (n=3), using volumes of larvae taken from incubators based on the initial larval density measurements. The larval tanks, including colour, material, water inlets, filter size, cleaning procedures and algal additions were as described previously (Penglase et al., 2010). Water inflow to each tank (temperature 8.0\u00b0C, salinity 34.8 \u00b1 0.2\u2030, 20-\u00b5m sand/lamella filtered, degassed, from 160 m depth) started at 0.8 L min-1 at larval transfer and increased over time to reach 4 L min -1 by 39 dph. The water temperature in larval tanks at transfer was 8.0\u00b0C (3 dph), and gradually increased and then maintained at 11.5\u00b0C from 27 dph. Oxygen saturation (75\u2013102%) and temperature were measured once daily in the outlet pipe of each tank. Dim light was provided continuously. Rotifers (Brachionus plicatilis. \u2018Cayman\u2019, adult lorica length 184 \u00b1 10 \u00b5m, width 134 \u00b1 11 \u00b5m) were batch cultured in 500 L tanks and washed as previously described (Penglase et al., 2011) (section 2.2.5 and 2.2.6) with the exception that algae paste (Chlorella sp., Docosa, SV12, Japan) was used as the culture feed. After washing, rotifers were enriched with either a control or treatment diet. The control enrichment was 250 mg Ori-green (Skretting, Norway) million-1 rotifers. The treatment enrichment (I+rotifers) was as per controls, but in addition 60 mg L-1 of sodium iodide (NaI; VWR, Belgium art. no. 27915.297) was added to the water at the start of rotifer enrichment. Ori-green was prepared to manufacturer\u2019s directions, while NaI 5 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 Pre Prin ts Pre Prin ts was dissolved in cold tap water, prior to addition to rotifer enrichment tanks. Rotifers were enriched for 2 h at densities between 1000-2000 mL-1 in water (as for rotifer culturing) with continual aeration and oxygenation (oxygen saturation was kept above 80%). After 1.5 h of this enrichment, an antibacterial (Pyceze, Novartis, Switzerland) was added to both control and treatment enrichment tanks at a rate of 0.2 ml L-1. Pyceze was used to lower rotifer bacterial numbers and thus control for any antibacterial effect of I enrichment. After enrichment, rotifers were washed, concentrated to 2000-4500 rotifers mL-1, transferred to storage tanks with aeration, and cooled rapidly (<10 min) to 8.5\u00b0C. To maintain I concentrations in the I+rotifers, 60 mg I L-1 (as NaI) was added to the treatment rotifer holding tank. Rotifers samples for element analysis were collected from rotifer storage tanks on 4 separate days during the larvae feeding trial. Rotifers were collected on 62 \u00b5m mesh, washed for 5 min with 12\u00b0C saltwater, placed in 25 mL containers and stored at -20\u00b0C. The feeding trial started at 4 dph, using rotifers as prepared in section 2.2. Larvae were fed control or I+rotifers (HI+rotifers) or a mixture of both (80:20, control:I+, MI+rotifers). Each tank received increasing quantities of rotifers with time, starting from 3.5 million rotifers tank-1 day-1 at 4 dph increasing to 6 million rotifers tank-1 day-1 by 39 dph. The same quantity of rotifers was fed to each tank, and larvae were assumed to be fed to satiation. The rotifers were fed daily to larvae in two batch feedings of equal rotifer quantity at 10:00 and 15:00. Rotifers were poured gently into larval tanks in a circular motion to ensure even rotifer distribution and minimal larvae disturbance. Control larvae and HI+larvae were fed only control or treatment rotifers respectively. The MI+larvae were fed I+ and control rotifers at 10:00, and only control rotifers at 15:00 resulting in the overall feeding ration consisting of 80:20 control: I+rotifers. For later analysis of skeletal deformities, fish were reared on identical diets from 40 to 124 dph. Larvae were co-fed Artemia (OK performance cysts, INVE, Belgium) and rotifers from 40-44 dph. Both Artemia and rotifers were enriched with Ori-green as per directions. Fish were fed only Artemia from 45-68 dph, co-fed Artemia and formulated diet (AgloNorse-EX1, Trofi, Troms\u00f8, Norway) from 69-91 dph. Only formulated feed was fed from 92 dph (EX1; 92-94 dph, EX1 and 2; 95-103 dph, EX2; 104-115 dph, EX3; 116-124 dph). Formulated feed was administered continuously for 24 hrs day-1 by belt feeders. Flow rate was increased from 4 L min-1 at 30 dph to 8 L min-1 at 108 dph, while water current speed in tanks was minimized by letting water enter through a 32 mm diameter inlet tube. Along with the increased water 6 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 Pre Prin ts Pre Prin ts flow rate, oxygen saturation (64-96%) was maintained by removal of fingerling cod when required. Larvae were sampled for weight and length at 5, 9, 19, 30, and 124 dph, element and thyroid hormone analysis at 5, 9, 19 and 30 dph, thyroid follicle morphology at 19, 30 and 37 dph and for analysis of skeletal deformities at 124 dph. All fish were euthanised with an overdose of tricaine methane sulfonate (MS 222) upon sampling. Larvae sampled for weight, minerals and thyroid hormones were collected on mesh (62 \u00b5m), briefly rinsed with ddH20, and patted dry from underneath with paper towel. Larvae were then placed in pre weighed tubes and frozen immediately in liquid nitrogen and stored at -80\u00b0C until analysis. All tubes were then reweighed to determine sample wet weights. Tubes sampled for weight determination were thawed and larvae were counted (n=20-100) to determine wet weight per larvae. Dry weight was determined from dry matter, which in turn was determined from tubes weighed before and after lyophilising. The standard length of the larvae was measured according to Hamre et al. (2008a) on 10 larvae tank-1. Larval densities in tanks were measured at 30 dph as described by Penglase et al. (2010). Larvae for thyroid follicle (3 fish tank-1) were selectively sampled to be similar in length, thus representing similar levels of morphological development (S\u00e6le and Pittman, 2010). Larvae were placed in individual tubes containing 1 ml of 4% paraformaldehyde in PBS buffer at pH 7.2. Samples were left overnight and then transferred to separate tubes containing 70% ethanol, where they remained at 4\u00b0C until embedding. For analysis of skeletal deformities, cod juveniles (124 dph, n=50 per tank) were measured for length and weight, frozen flat and subsequently stored in individual labelled plastic bags at -20\u00b0C until analysis. Survival in one HI+ tank decreased to zero prior to this sampling, so data represents the mean \u00b1 SD n=2 for the HI+fish at 124 dph. Samples for analysis of total I were digested under alkaline conditions using tetra methyl ammonium hydroxide ((CH3)4NOH, Tamapure-AA, Tama chemicals, Japan) and then analysed with ICP-MS (Agilent 7500 series, USA) as described by Julshamn, Dahl and Eckhoff (2001) with cod muscle (BCR-422, Belgium) used as the standard reference material. Samples for the analysis of other elements were prepared by wet digestion with nitric acid (65 7 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 Pre Prin ts Pre Prin ts % HNO3 Suprapur\u00ae, Merck, Germany) and hydrogen peroxide (30 % H2O2, Merck, Germany), in a microwave (Ethos 1600, Milestone, USA) as described by Julshamn et al. (2004). The samples were then analysed with ICP-MS along with blanks and standard reference material as described previously (Julshamn et al., 2004), with modifications to the mass of Mn (Mass 55) and Pb (Mass 208) measured. The standard reference materials used (NIST-SRM 1566, Oyster tissue, USA; TORT-2, NRC, lobster hepatopancreas, Canada) had similar concentrations of minerals as the samples analysed. Thyroid hormones extraction from larvae was carried out according to Einarsdottir et al. (2006) with some modifications. Approximately 500 mg WW of larvae (440 to 630 mg) was homogenised (Precellys 24, Bertin technologies, France) in 1 ml of ice-cold methanol (Sigma-Aldrich, Germany), and then stored over night at -20oC. Samples were then centrifuged (30 min, 3000 rpm, 4 \u00baC) and the supernatant removed from the pellet. This extraction procedure was then repeated on the pellet twice more. Nitrogen was used to dry the supernatant of methanol. Lipids were removed from samples by modified Folch extraction. To eliminate any lipid in samples, the dried extracts were dissolved in barbital buffer (0.1 M pH 8.6), methanol and chloroform (1:1:2). The aqueous phase containing T4 and T3 was transferred to a fresh tube, evaporated with nitrogen and frozen at -20o C until use. To estimate extraction efficiency, \u22481000 cpm of [125I]-rT3 (NEX109, Perkin Elmer, USA) were added to the homogenates after homogenisation. The extraction efficiency ranged between 71 to 81%. The T4 and T3 content were determined by radioimmunoassay using an external standard curve according to Einarsdottir et al. (2006), and further corrected for the extraction efficiency of each sample. Larval T3 contents were also reanalysed by a DELFIA\u00ae T3 Kit (PerkinElmer, Turku, Finland) according to manufacturer\u2019s instructions. Larvae were dehydrated in an increasing gradient of ethanol and embedded in Technovit 7100 as per directions (Heraeus Kulzer, Wehrheim, Germany). The resin blocks were then sectioned into 5 \u03bcm thick slices, and every second section was placed on a slide and stained with toluidine blue. The follicle number within the pharyngeal region was counted and the area of epithelium and colloid were measured for each larvae (2 larvae tank -1 at 19 and 30 dph, 1 larvae tank-1 at 37 dph) at 200\u00d7 magnification using a microscope and computer assisted program CAST-grid version 2 (Olympus, Albertslund, Denmark) according to Saele et al. (2003). The colloid and epithelium volume were calculated using area and width of sections, 8 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 Pre Prin ts Pre Prin ts and skipped sections were assumed to have the same volume as that measured on the preceding section. Larval sections were also scanned for evidence of thyroid follicles in the kidneys, as has been observed for some other fish species such as common carp (Cyprinus carpio) (Geven et al., 2007), but none were observed. The radiographic imaging and analysis of skeletal deformities was performed as described previously (Penglase et al., 2010) with skeletal deformities and degree of deformities classified according to Baeverfjord et al. (undated). Briefly, radiographic images were visually examined for any skeletal pathology, and deviations from normal were recorded. Deviations were classified in axial deviations, vertebral deviations and head deformities, and further classified into sub categories and degrees of severity. Larval survival was adjusted using a linear individual probability timeline for each tank to calculate the probable survival of sampled larvae had they remained in the tanks until density measurements at 30 dph, using the following equation; Estimated survival of sampled larvae at time point Y = 100 \u2013 (((100-S)/T1)*(T1-T2))/100)*X) where S is the measured survival % at 30 dph, T1 equals the time period in days from the start (5 dph) and end (30 dph) survival measurements (25 d), T2 equals the number of days post 100% survival (trial start) and X equals the number of larvae sampled at time point Y. This equation was used to produce two numbers, one for the sampling at 9 (T2 = 4) and one at 19 (T2 = 14) dph, and along with the number of larvae sampled on day 30, were added to the larvae measured in tanks from density measurements taken after sampling at 30 dph. Specific growth rates (SGR) of cod larvae were calculated with the following equation SGR = (e^((lnW1 \u2212 lnW0)/(t2-t1))-1)\u00d7100 where W0 and W1 are the initial and final dry weights (tank means) respectively, and t2 \u2013 t1 is the time interval in days between age t1 and t2 (Ricker, 1958). Fulton\u2019s condition factor (FC) was calculated using FC = Weight (g)*100/Length(cm)3. The I concentration ratio (CR) between larvae and feed was CR = Larval I content (mg kg-1 DW)/rotifer I content (mg kg-1 DW). Statistica software (Statsoft Inc., 2008, Tulsa, USA, Ver.9) was used for statistical analysis of data except when GraphPad Prism (GraphPad Software, San Diego, CA, USA, Ver. 5) was 9 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 Pre Prin ts Pre Prin ts used to model fit the I concentration of cod larvae versus rotifers, and on data from 124 dph, as the loss of one replicate in the HI larvae group prevented ANOVA analysis between all three groups. Data analysed with Statistica were checked for homogeneity of variances using Levene\u2019s test before having significance tested with one-way ANOVA followed with Fisher\u2019s least significant difference (LSD) homogeneity post-hoc test at each time point. Data with significantly different variance between treatments according to Levene\u2019s test (p<0.05) was log transformed before analysis. As the density of cod larvae has a large effect on growth (Koedijk et al., 2010), growth data during the larval stage was analysed with ANCOVA with the final larval density in tanks included as a continuous predictor. Data from 124 dph were analysed using regression, and tested against the null hypothesis that rotifer I content had no effect on outcome. Differences among means were considered significant at p<0.05. 10 277 278 279 280 281 282 283 284 285 286 287 Pre Prin ts Pre Prin ts 3 Results 3 cod larvae growth : No statistically significant differences in cod larvae length (p > 0.08) or dry weight (p > 0.25) occurred between treatments, although high variation between tanks may have masked effects (Fig. 1). On average, cod larvae grew from 4.9 \u00b1 0.2 mm to 6.8 \u00b1 0.5 mm in length, and 0.065 \u00b1 0.002 mg to 0.27 \u00b1 0.07 mg fish-1 in dry weight from 5 to 30 dph (Fig. 1), representing a specific growth rate of 6.3% day-1 during this period. There were no statistical differences (p>0.39) in the cod larval survival adjusted for sampled larvae (see section 2.6) between treatments at 30 dph which were 28 \u00b1 2, 39 \u00b1 20 and 30 \u00b1 15 % for controls, MI and HI groups respectively, while the average for all groups was 32 \u00b1 13 %. Survivals based solely on densities in tanks at 30 dph without taking into account sampled larvae were 12 \u00b1 2, 19 \u00b1 12 and 12 \u00b1 10 % for controls, MI and HI groups respectively. Control rotifers contained 0.60 \u00b1 0.33 mg I kg-1 DW while HI+rotifers contained 129 \u00b1 101 mg I kg-1 DW (Table 1). Whole body I levels in cod larvae were significantly different between groups (p<0.01, Fig. 2). Cod larvae (5 dph) had a starting concentration of 4.0 \u00b1 0.3 mg I kg-1 DW. Between 9 and 30 dph average I concentrations were 1.6 \u00b1 0.3 mg I kg-1 DW for control larvae, while MI larvae had 3 fold higher levels (4.9 \u00b1 2.4 mg I kg-1 DW), and HI larvae 7 fold higher levels (11.0 \u00b1 3.3 mg I kg-1 DW) than controls. Other element concentrations were also affected by treatment in both rotifers and cod larvae. HI+rotifers contained less Fe and Mn than controls (Table 1), while HI larvae contained more Mn, Fe and Cu than controls and MI larvae at one or more time points (Fig. 2b-c, e). Both HI and MI larvae contained higher levels of Co than controls (Fig. 2d), while larval Zn and Se concentrations were unaffected by treatment (Fig. 2f-g). Rotifer macro mineral concentrations were unaffected by treatment (Table 1), but increased levels of Ca and Mg, and lower levels of P and K were observed in HI larvae in comparison to controls and/or MI larvae during the rotifer feeding period (Fig. 3). The rate of increase in cod larvae I concentrations decreased as dietary I levels (rotifer I concentration) increased, and thus the I concentration ratio between cod larvae and rotifers displayed a negative trend (Fig 4; p<0.01). The age of the cod larvae did not effect their I concentration ratio (p = 0.96). The model predicts that the ratio of I in cod larvae versus rotifers equals 1 when rotifers have 3.5 mg I kg-1 DW (Fig. 4). 11 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 Pre Prin ts Pre Prin ts There were no differences in thyroid hormone levels or ratios between treatments (Fig. 5). Data was normalised to aid interpretation, as T3 results were higher than obtained previously (Penglase et al., 2010) and the high result was validated by the analysis of T3 with a second method (see methods). The total volume of thyroid follicle was 1.3 fold lower and the total epithelium volume was 1.4 fold lower per fish in HI versus MI larvae, but not controls, at 30 dph (Fig. 6a, c). The thyroid follicle colloid to epithelium ratio was higher in HI larvae than controls at 19 (1.7 fold) and 37 (1.8 fold) dph, while MI larvae did not differ from controls (Fig. 6d). No statistically significant differences were observed between groups in colloid volume or total number of thyroid follicles (Fig. 6b, e). Images of thyroid follicle sections from control and HI larvae at 37 dph are shown in figure 7. There were no significant differences in the weights (average; 2.50 \u00b1 0.19 g), lengths (6.14 \u00b1 0.16 cm) or condition factors (1.05 \u00b1 0.03) between groups at 124 dph (Table 2, n=400). Neck axis angle became closer to normal (180 \u00b1 3 degrees) with increasing I levels in rotifers, but there were no differences in any of the other skeletal deformity measurements (Table 2). 12 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 Pre Prin ts Pre Prin ts Discussion The hypothesis of this study was that commercially reared cod larvae fed rotifers would benefit from increased dietary I. The bases for this hypothesis were, one; rotifer I concentrations are often at the lower end or below juvenile/adult fish requirements (NRC, 2011), two; rotifers have 6 - 600 fold lower concentrations of I than copepods (Hamre et al., 2008b), the natural feed of cod larvae (Thompson and Harrop, 1991), three; cod larvae have better growth and development when fed copepods versus rotifers (Busch et al., 2010; Koedijk et al., 2010), four; I levels in copepod fed cod larvae are higher than rotifer fed cod larvae (Busch et al., 2010), and five; increased growth and/or survival has been observed in larval stages of several marine fish species fed or reared in environments with increased levels of bioavailable I (Hamre et al., 2008a; Witt et al., 2009; Ribeiro et al., 2011; Ribeiro et al., 2012). However, in contrast to the hypothesis, the increased thyroid follicle colloid to epithelium (C/E) ratio observed in this study indicates that I toxicity occurred in cod larvae fed rotifers with 129 mg I kg-1 DW. This observation purely in relation to the I level is not unexpected. A high C/E ratio in thyroid follicles is a classic symptom of I induced toxicity termed I (Baker, 2004) or colloid goitre, and occurs in mice at dietary I concentrations 10 fold higher than requirements with increasing severity developing with increasing I ingestion rates (Yang et al., 2006). What is interesting is that despite copepods containing 50 \u2013 350 mg I kg-1 DW (Solbakken et al., 2002; Hamre et al., 2008b), cod larvae have either similar or lower thyroid follicle C/E ratios when fed copepods compared to rotifers (Gr\u00f8tan, 2005). Furthermore, I concentrations observed in cod larvae fed natural zooplankton (29 mg I kg-1 DW at 27 dph; (Busch et al., 2010)) were 2.2 fold higher than the highest level observed in the current study (HI larvae, 30 dph; 13 \u00b1 4 mg I kg-1 DW). Thus it appears that high I concentrations in copepods do not induce morphological changes in thyroid follicles consistent with I toxicity, but do appear to be effectively transferred from copepods to fish larvae upon consumption. It is possible that copepods do not induce I toxicity in fish larvae due to nutrient interactions. For example, I toxicity can be prevented by the simultaneous presence of the bromine anion (Br-) in animals ranging from chicks (Gallus gallus) (Baker, Parr and Augspurger, 2003) to Artemia (S. Penglase et al., unpublished data). The exact mechanism for this Br-/I- interaction is still unknown, but it has been demonstrated that Br- decreases iodide accumulation in the thyroid follicles and increases I excretion via the kidneys in rats (Pavelka, 2004). Although the bromide concentrations in whole copepods and rotifers are unknown, we speculate that copepods have relatively high levels of bromide reflecting the high levels found in other 13 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 Pre Prin ts Pre Prin ts marine organisms, and this bromide helps prevent I toxicity in fish larvae. In the marine ecosystem, bromine is naturally found at similar high concentrations as I in seaweed (Romaris-Hortas, Moreda-Pineiro and Bermejo-Barrera, 2009), adult fish (Arafa et al., 2000; Wan et al., 2010), and as part of the hard chitin structures of crustaceans such as crabs (Cribb et al., 2009; Schofield et al., 2009) and copepods (Perry, Grime and Watt, 1988). Thyroid hormone levels and ratios were similar between cod larvae groups, and this is probably due to the compensatory changes observed in the thyroid follicles. For example pathological changes of over 70% in thyroid gland morphology have been observed in dogs (Canis lupus familiaris) with little change in circulating TH levels (Graham, Refsal and Nachreiner, 2007), and significant changes in fish thyroid follicle morphology with little change in thyroid hormone levels have also been reported for fish (Hawkyard et al., 2011; Morris et al., 2011). Few differences were found between the control and MI larvae groups, with the exception of whole body I concentrations. The increased whole body I content of cod larvae demonstrates the effective transfer of I from the rotifers to the cod larvae. The current study differs to previous studies exploring the uptake of I in fish larvae, as it attempted to ensure the ingestion of graded levels of I by maintaining the I concentration in the prey organism up until the point of feeding. Srivastava et al. (2012) found that I leaches rapidly from rotifers after enrichment with sodium iodide. Previous studies have found no difference in the I concentration of cod larvae fed control or I supplemented rotifers (Hamre et al., 2008a); S. Penglase et al., unpublished data), and this is probably a consequence of I leaching from rotifers in the minimum 1.5 to 2 h period between rotifer enrichment and feeding of the rotifers to cod larvae in these studies. In the current study, cod larvae iodine level increases were proportionally smaller for each increase in dietary I levels; control fish were 2.7 fold higher, while MI were 5 fold lower and HI larvae were 12 fold lower in I than their respective diets. Body stores of minerals are a good indicator of status (Baker, 1986), and the decreasing level of I retention in cod larvae relative to feed I levels indicates that requirements were met at levels lower than those fed to MI larvae. Modelling of the ratio between cod larvae and rotifer I concentrations predicts that based on a ratio of 1:1, rotifer I concentrations of 3.5 mg kg-1 DW meet cod larvae requirements. Both food and water contribute to the I status of adult, juvenile (Lall, 2002) and larval fish (Witt et al., 2009; Ribeiro et al., 2011). Alongside the I content in the continuously 14 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 Pre Prin ts Pre Prin ts exchanging seawater (88 \u00b5g I L-1, Moren, Sloth and Hamre, 2008), the control rotifers in the current study with 0.6 mg I kg-1 DW appeared to prevent any gross signs of I deficiency in cod larvae. This is at the lower end of the 0.6 \u2013 1.1 mg I kg-1 DW recommended by the national research council (NRC, 2011) as the I requirements of juvenile/adult fish. The reason that symptoms of severe I deficiency such as classic goitre have been observed in other fish studies is probably due to water parameters. Clear signs of I deficiency (goitre, decreased growth and/or decreased survival) occurred in fish larvae reared in either recirculation systems (Ribeiro et al., 2011; Ribeiro et al., 2012) or well water (Witt et al., 2009). Nitrate (NO3-) is goitrogenic as it competitively blocks iodide uptake by the sodium iodide symporter (Tonacchera et al., 2004), and NO3- at levels commonly found in recirculation systems causes goitre in sharks (Crow et al., 1998; Morris et al., 2011). Furthermore, ozone (O3) used as a disinfectant in recirculation systems readily oxidises bioavailable I, iodide (I-) to iodate (IO3-) (Sherrill, Whitaker and Wong, 2004). Dissolve iodate is presumed to have low bioavailablity for fish (Sherrill, Whitaker and Wong, 2004), and higher levels of iodate compared to iodide were correlated to poor growth and survival in well water reared pacific threadfin larvae (Witt et al., 2009). Thus in recirculation systems, the presence of high levels of goitrogens (NO3-) and low levels of dissolved bioavailable I (I-) may increase the dietary I requirements of fish larvae over those reared with continuous water exchange where nitrate and its precursors are continuously removed and iodide continuously replaced, such as in the current study. Along with thyroid follicle morphology, dietary I also influenced the mineral composition of cod larvae. While most of the tested mineral concentrations in MI larvae were similar to controls, HI larvae had 10 to 25% higher levels of Ca, Mg, Mn, Fe, Co and Cu and around 10% lower levels of P and K than controls at one or more time points within the rotifer feeding period. For most of the minerals, differences in levels were observed by the first sampling point (9 dph; 4 days of feeding on rotifers). The differences cannot be explained by the feed; the HI rotifers had \u2248 10% less Mn and Fe, and no statistical differences were observed in Ca, Mg, K, P, Cu or Co concentrations. Hamre et al. (2008a) found that cod larvae fed increased levels of both I and Se had an 8% increase in whole body copper levels, similar to this study. Nguyen et al. (2008) found increased or decreased copper levels (20%) in red sea bream (Pagrus major) larvae depending on whether they were fed rotifers enriched with Mn alone or alongside Zn. While it is known that I deficiency can alter mineral distribution and homeostasis of Cu, Mn, Fe and Zn (Giray et al., 2003), to our knowledge this 15 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 Pre Prin ts Pre Prin ts is the first data demonstrating that I oversupply can also effect mineral homeostasis. Although there were few differences in growth or skeletal deformities observed between treatments at 124 dph, there was a small but statistically significant improvement in the neck axis angle in the HI compared to the control and MI cod groups (Table 2), and this may be linked to the differences in cod mineral concentrations in the larval stage. Conclusion Iodine enriched rotifers increased the levels of I in cod larvae, although as I levels in rotifers increased the increases in cod larvae I levels became proportionally smaller. Few differences occurred between cod larvae reared on control diets with 0.6 mg I kg -1 DW and those reared on diets with 26 mg I kg-1 DW, while the I concentration ratio between cod larvae and rotifers suggests cod larvae have an I requirement of 3.5 mg I kg -1 rotifers DW. Rotifers with copepod levels of I (129 mg I kg-1 DW) changed cod larvae whole body concentration of many essential minerals and induced changes in thyroid follicles morphology consistent with colloid goitre. The data presents one of the first observations of dietary induced I toxicity in fish, and suggests that I toxicity in fish larvae may be determined to a greater extent by I bioavailability and nutrient interactions than by body burdens of I. l ength (data set l; mm fish -1 , left y axis) and dry weight (data set w; mg fish -1 , right y axis) of : cod larvae fed control (\u25a1), MI ( \u25cb ) or HI ( \u25cf ) rotifers, from 5 to 30 dph. At 5 dph, data are mean \u00b1 SD of 2 analytical parallels for dry weight and mean \u00b1 SD (n=30) for length. Data at all other dph are mean \u00b1 SD (n=3) where n represents the average of 10 larvae tank -1 measured for length, and a group of 47 to 520 larvae group weighed then counted to determine individual mass. Pre Prin ts Pre Prin ts Figure 2 Essential micro element concentration in whole cod larvae Essential micro element concentrations (mg kg -1 DW) in whole cod larvae fed either control (\u25a1), MI ( \u25cb ) or HI ( \u25cf ) rotifers, from 5 to 30 dph. Letters denote statistically significant differences in mineral concentrations between treatments at a given day (one-way ANOVA; p<0.05). Data are mean \u00b1 SD (n=3), except at 5 dph where data are mean \u00b1 SD of analytical parallels. Pre Prin ts Pre Prin ts Figure 3 Essential micro element concentration in whole cod larvae Pre Prin ts Pre Prin ts Figure 4 Essential micro element concentration in whole cod larvae Pre Prin ts Pre Prin ts Figure 5 Essential micro element concentration in whole cod larvae Pre Prin ts Pre Prin ts Figure 6 Essential micro element concentration in whole cod larvae Pre Prin ts Pre Prin ts Figure 7 Essential micro element concentration in whole cod larvae Pre Prin ts Pre Prin ts Figure 8 Essential micro element concentration in whole cod larvae Pre Prin ts Pre Prin ts Figure 9 Essential macro element concentrations in whole cod larvae Essential macro mineral concentrations (g kg -1 DW) in whole cod larvae fed either control (\u25a1), MI ( \u25cb ) or HI ( \u25cf ) rotifers, from 5 to 30 dph. Letters denote statistically significant differences in mineral concentrations between treatments at a given day (one-way ANOVA, p<0.05). Data are mean \u00b1 SD (n=3), except at 5 dph where data are from a single analysis. Pre Prin ts Pre Prin ts Figure 10 Essential macro element concentrations in whole cod larvae Pre Prin ts Pre Prin ts Figure 11 Essential macro element concentrations in whole cod larvae Pre Prin ts Pre Prin ts Figure 12 Essential macro element concentrations in whole cod larvae Pre Prin ts Pre Prin ts Figure 13 Cod larvae iodine concentration in relation to their feed Ratio of iodine concentration (mg kg -1 DW) in cod larvae versus their diet (rotifers iodine levels (mg kg -1 DW)). X axis is log transformed. Line represents best fit model (Morrison Ki, R 2 = 0.94). Data are mean \u00b1 SD (n=9). Pre Prin ts Pre Prin ts Figure 14 Cod larvae thyroid hormone levels and ratios Normalised mean thyroid hormone levels (NML) in cod larvae fed either control (\u25a1), MI ( \u25cb ) or HI ( \u25cf ) rotifers. Graph A is tri-iodothyronine (T 3 ), Graph B is thyroxine (T 4 ), while graph C is the ratio between the NML of T 3 /T 4 . Data are mean \u00b1 SD (n=3) for all data points except controls at 30 dph which has an outlier removed in graph B and C (n=2). Pre Prin ts Pre Prin ts Figure 15 Cod larvae thyroid hormone levels and ratios Pre Prin ts Pre Prin ts Figure 16 Cod larvae thyroid hormone levels and ratios Pre Prin ts Pre Prin ts Figure 17 Cod larvae thyroid follicle morphology Thyroid follicle morphology in cod larvae fed either control (\u25a1), MI ( \u25cb ) or HI ( \u25cf ) rotifers. Graph A shows the total number of follicles per fish, Graph B is the total thyroid follicle volume per fish, Graphs C and D show the volume of colloid or epithelium per fish, Graph E shows the ratio between the colloid and epithelium volumes. Letters denote statistically significant differences between treatments per time point (one-way ANOVA, p<0.05). Data are mean \u00b1 SD (n=3) where n consists of the average measurements from two fish per tank at 19 and 30 dph, and one fish per tank at 37 dph. Pre Prin ts Pre Prin ts Figure 18 Cod larvae thyroid follicle morphology Pre Prin ts Pre Prin ts Figure 19 Cod larvae thyroid follicle morphology Pre Prin ts Pre Prin ts Figure 20 Cod larvae thyroid follicle morphology Pre Prin ts Pre Prin ts Figure 21 Cod larvae thyroid follicle morphology Pre Prin ts Pre Prin ts Figure 22 Thyroid follicle sections from cod larvae Thyroid follicle section from c od larvae (37 dph) fed either control (A) or HI ( B ) rotifers. Sections are stained with toulidine blue. C; thyroid follicle colloid, E; example of thyroid follicle epithelium. Scale bars are 100 \u00b5m. Pre Prin ts Pre Prin ts Figure 23 Thyroid follicle sections from cod larvae Pre Prin ts Pre Prin ts",
+    "v2_text": "acknowledgements : Thank you to technical staff at IMR Austevoll and NIFES for fish husbandry, sampling help and skilled analytical assistance, especially Stig Ove Utskot and Berit Solli. Thank you to Karin Pittman for allowing access to equipment for thyroid follicle morphology quantification. Contributions of the authors. S. P., K. H., A. N. and T.H. designed the project, T. H. conducted animal trials, S. P., K. H. and T. H. sampled. S. P., O.S. and S.H. analysed samples. S. P. statistically analysed data. All authors contributed to manuscript writing and revision. 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 Pre Prin ts Pre Prin ts 2 cod larvae rearing : The experiment was performed at the Institute of Marine Research (IMR), Austevoll Research Station, Norway. This study was carried out within the Norwegian animal welfare act guidelines (code 750.000) at an approved facility. As this trial was assumed to be a nutrition trial based on all available studies up to the date of the trial, no specific permit was required under the guidelines. Naturally spawned and fertilised Atlantic cod eggs were obtained from in house second generation brood stock. Prior to incubation, eggs were disinfected with 200 mg L-1 glutaraldehyde for 9 min at 6\u00b0C and eggs were incubated with a standard protocol as described in Penglase et al. (2010). Upon 100% hatching (16 dpf, 99 day degrees), larval density in the incubators was measured via tube sampling and ranged from 2000-3000 larvae L-1. Three days post 100% hatching (dph), larvae (65 \u00b1 2 \u00b5g DW fish-1, n = 2 where n is a pool of 428 or 520 fish, and 4.9 \u00b1 0.2 mm fish-1, n = 30 (mean \u00b1 SD), measured 5 dph) were transferred into the experimental tanks. Larvae were stocked at an estimated density of 50 000 larvae (120 larvae L -1) in each of the nine 500 L (400 L water volume) experimental tanks (n=3), using volumes of larvae taken from incubators based on the initial larval density measurements. The larval tanks, including colour, material, water inlets, filter size, cleaning procedures and algal additions were as described previously (Penglase et al., 2010). Water inflow to each tank (temperature 8.0\u00b0C, salinity 34.8 \u00b1 0.2 \u2030, 20-\u00b5m sand/lamella filtered, degassed, from 160 m depth) started at 0.8 L min-1 at larval transfer and increased over time to reach 4 L min -1 by 39 dph. The water temperature in larval tanks at transfer was 8.0\u00b0C (3 dph), and gradually increased and then maintained at 11.5\u00b0C from 27 dph. Oxygen saturation (75\u2013 102%) and temperature were measured once daily in the outlet pipe of each tank. Dim light was provided continuously. Rotifers (Brachionus plicatilis. \u2018Cayman\u2019, adult lorica length 184 \u00b1 10 \u00b5m, width 134 \u00b1 11 \u00b5m) were batch cultured in 500 L tanks and washed as previously described (Penglase et al., 2011) (section 2.2.5 and 2.2.6) with the exception that algae paste (Chlorella sp., Docosa, SV12, Japan) was used as the culture feed. After washing, rotifers were enriched with either a control or treatment diet. The control enrichment was 250 mg Ori-green (Skretting, Norway) million-1 rotifers. The treatment enrichment (I+rotifers) was as per controls, but in addition 60 mg L-1 of sodium iodide (NaI; VWR, Belgium art. no. 27915.297) was added to the water at the start of rotifer enrichment. Ori-green was prepared to manufacturer\u2019s directions, while NaI 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 Pre Prin ts Pre Prin ts was dissolved in cold tap water, prior to addition to rotifer enrichment tanks. Rotifers were enriched for 2 h at densities between 1000-2000 mL-1 in water (as for rotifer culturing) with continual aeration and oxygenation (oxygen saturation was kept above 80%). After 1.5 h of this enrichment, an antibacterial (Pyceze, Novartis, Switzerland) was added to both control and treatment enrichment tanks at a rate of 0.2 ml L-1. Pyceze was used to lower rotifer bacterial numbers and thus control for any antibacterial effect of I enrichment. After enrichment, rotifers were washed, concentrated to 2000-4500 rotifers mL-1, transferred to storage tanks with aeration, and cooled rapidly (<10 min) to 8.5\u00b0C. To maintain I concentrations in the I+rotifers, 60 mg I L-1 (as NaI) was added to the treatment rotifer holding tank. Rotifers samples for element analysis were collected from rotifer storage tanks on 4 separate days during the larvae feeding trial. Rotifers were collected on 62 \u00b5m mesh, washed for 5 min with 12\u00b0C saltwater, placed in 25 mL containers and stored at -20\u00b0C. The feeding trial started at 4 dph, using rotifers as prepared in section 2.2. Larvae were fed control or I+rotifers (HI+rotifers) or a mixture of both (80:20, control:I+, MI+rotifers). Each tank received increasing quantities of rotifers with time, starting from 3.5 million rotifers tank-1 day-1 at 4 dph increasing to 6 million rotifers tank-1 day-1 by 39 dph. The same quantity of rotifers was fed to each tank, and larvae were assumed to be fed to satiation. The rotifers were fed daily to larvae in two batch feedings of equal rotifer quantity at 10:00 and 15:00. Rotifers were poured gently into larval tanks in a circular motion to ensure even rotifer distribution and minimal larvae disturbance. Control larvae and HI+larvae were fed only control or treatment rotifers respectively. The MI+larvae were fed I+ and control rotifers at 10:00, and only control rotifers at 15:00 resulting in the overall feeding ration consisting of 80:20 control: I+rotifers. For later analysis of skeletal deformities, fish were reared on identical diets from 40 to 124 dph. Larvae were co-fed Artemia (OK performance cysts, INVE, Belgium) and rotifers from 40-44 dph. Both Artemia and rotifers were enriched with Ori-green as per directions. Fish were fed only Artemia from 45-68 dph, co-fed Artemia and formulated diet (AgloNorse-EX1, Trofi, Troms\u00f8, Norway) from 69-91 dph. Only formulated feed was fed from 92 dph (EX1; 92-94 dph, EX1 and 2; 95-103 dph, EX2; 104-115 dph, EX3; 116-124 dph). Formulated feed was administered continuously for 24 hrs day-1 by belt feeders. Flow rate was increased from 4.0 L min-1 at 30 dph to 8 L min-1 at 108 dph, while water current speed in tanks was minimized by letting water enter through a 32 mm diameter inlet tube. Along with the 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 Pre Prin ts Pre Prin ts increased water flow rate, oxygen saturation (64-96%) was maintained by removal of fingerling cod when required. Larvae were sampled for weight and length at 5, 9, 19, 30, and 124 dph, element and thyroid hormone analysis at 5, 9, 19 and 30 dph, thyroid follicle morphology at 19, 30 and 37 dph and for analysis of skeletal deformities at 124 dph. All fish were euthanised with an overdose of tricaine methane sulfonate (MS 222) upon sampling. Larvae sampled for weight, minerals and thyroid hormones were collected on mesh (62 \u00b5m), briefly rinsed with ddH20, and patted dry from underneath with paper towel. Larvae were then placed in pre weighed tubes and frozen immediately in liquid nitrogen and stored at -80\u00b0C until analysis. All tubes were then reweighed to determine sample wet weights. Tubes sampled for weight determination were thawed and larvae were counted (n=20-100) to determine wet weight per larvae. Dry weight was determined from dry matter, which in turn was determined from tubes weighed before and after lyophilising. The standard length of the larvae was measured according to Hamre et al. (2008a) on 10 larvae tank-1. Larval densities in tanks were measured at 30 dph as described by Penglase et al. (2010). Larvae for thyroid follicle (3 fish tank-1) were selectively sampled to be similar in length, thus representing similar levels of morphological development (S\u00e6le and Pittman, 2010). Larvae were placed in individual tubes containing 1 ml of 4% paraformaldehyde in PBS buffer at pH 7.2. Samples were left overnight and then transferred to separate tubes containing 70% ethanol, where they remained at 4\u00b0C until embedding. For analysis of skeletal deformities, cod juveniles (124 dph, n=50 per tank) were measured for length and weight, frozen flat and subsequently stored in individual labelled plastic bags at -20\u00b0C until analysis. Survival in one HI+ tank decreased to zero prior to this sampling, so data represents the mean \u00b1 SD n=2 for the HI+fish at 124 dph. Samples for analysis of total I were digested under alkaline conditions using tetra methyl ammonium hydroxide ((CH3)4NOH, Tamapure-AA, Tama chemicals, Japan) and then analysed with ICP-MS (Agilent 7500 series, USA) as described by Julshamn, Dahl and Eckhoff (2001) with cod muscle (BCR-422, Belgium) used as the standard reference material. Samples for the analysis of other elements were prepared by wet digestion with nitric acid (65 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 Pre Prin ts Pre Prin ts % HNO3 Suprapur\u00ae, Merck, Germany) and hydrogen peroxide (30 % H2O2, Merck, Germany), in a microwave (Ethos 1600, Milestone, USA) as described by Julshamn et al. (2004). The samples were then analysed with ICP-MS along with blanks and standard reference material as described previously (Julshamn et al., 2004), with modifications to the mass of Mn (Mass 55) and Pb (Mass 208) measured. The standard reference materials used (NIST-SRM 1566, Oyster tissue, USA; TORT-2, NRC, lobster hepatopancreas, Canada) had similar concentrations of minerals as the samples analysed. Thyroid hormones extraction from larvae was carried out according to Einarsdottir et al. (2006) with some modifications. Approximately 500 mg WW of larvae (440 to 630 mg) was homogenised (Precellys 24, Bertin technologies, France) in 1 ml of ice-cold methanol (Sigma-Aldrich, Germany), and then stored over night at -20oC. Samples were then centrifuged (30 min, 3000 rpm, 4 \u00baC) and the supernatant removed from the pellet. This extraction procedure was then repeated on the pellet twice more. Nitrogen was used to dry the supernatant of methanol. Lipids were removed from samples by modified Folch extraction. To eliminate any lipid in samples, the dried extracts were dissolved in barbital buffer (0.1 M pH 8.6), methanol and chloroform (1:1:2). The aqueous phase containing T4 and T3 was transferred to a fresh tube, evaporated with nitrogen and frozen at -20o C until use. To estimate extraction efficiency, \u22481000 cpm of [125I]-rT3 (NEX109, Perkin Elmer, USA) were added to the homogenates after homogenisation. The extraction efficiency ranged between 71 to 81%. The T4 and T3 content were determined by radioimmunoassay using an external standard curve according to Einarsdottir et al. (2006), and further corrected for the extraction efficiency of each sample. Larval T3 contents were also reanalysed by a DELFIA\u00ae T3 Kit (PerkinElmer, Turku, Finland) according to manufacturer\u2019s instructions. Larvae were dehydrated in an increasing gradient of ethanol and embedded in Technovit 7100 as per directions (Heraeus Kulzer, Wehrheim, Germany). The resin blocks were then sectioned into 5 \u03bcm thick slices, and every second section was placed on a slide and stained with toluidine blue. The follicle number was counted and the area of epithelium and colloid were measured for each larvae (2 larvae tank-1 at 19 and 30 dph, 1 larvae tank-1 at 37 dph) at 200\u00d7 magnification using a microscope and computer assisted program CAST-grid version 2 (Olympus, Albertslund, Denmark) according to Saele et al. (2003). The colloid and 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 Pre Prin ts Pre Prin ts epithelium volume were calculated using area and width of sections, and skipped sections were assumed to have the same volume as that measured on the preceding section. The radiographic imaging and analysis of skeletal deformities was performed as described previously (Penglase et al., 2010). Larval survival was adjusted for sampled fish using a linear individual probability timeline for each tank to calculate the probable survival of sampled larvae had they not been sampled at each sampling point using the following equation; Estimated survival of sampled larvae at time point Y = 100 \u2013 (((100-S)/T1)*(T1-T2))/100)*X) where S is the measured survival % at 30 dph, T1 equals the time period in days from the start (5 dph) and end (30 dph) survival measurements (= 25 d), T2 equals the number of days post 100% survival (trial start) and X equals the number of larvae sampled at time point Y. This equation was used to produce two numbers, one for the sampling at 9 (T2 = 4) and one at 19 (T2 = 14) dph, and along with the number of larvae sampled on day 30, were added to the larvae measured in tanks from density measurements taken after sampling at 30 dph. Specific growth rates (SGR) of cod larvae were calculated with the following equation SGR = (e^((lnW1 \u2212 lnW0)/(t2-t1))-1)\u00d7100 where W0 and W1 are the initial and final dry weights (tank means) respectively, and t2 \u2013 t1 is the time interval in days between age t1 and t2 (Ricker, 1958). Fulton\u2019s condition factor (FC) was calculated using FC = Weight (g)*100/Length(cm)3. The I concentration ratio (CR) between larvae and feed was CR = Larval I content (mg kg-1 DW)/rotifer I content (mg kg-1 DW). Statistica software (Statsoft Inc., 2008, Tulsa, USA, Ver.9) was used for statistical analysis of data except when GraphPad Prism (GraphPad Software, San Diego, CA, USA, Ver. 5) was used to model fit the I concentration of cod larvae versus rotifers, and on data from 124 dph, as the loss of one replicate in the HI larvae group prevented ANOVA analysis between all three groups. Data analysed with Statistica were checked for homogeneity of variances using Levene\u2019s test before having significance tested with one-way ANOVA followed with Fisher\u2019s least significant difference (LSD) homogeneity post-hoc test at each time point. Data with significantly different variance between treatments according to Levene\u2019s test (p<0.05) was 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 Pre Prin ts Pre Prin ts log transformed before analysis. As the density of cod larvae has a large effect on growth (Koedijk et al., 2010), growth data during the larval stage was analysed with ANCOVA with the final larval density in tanks included as a continuous predictor. Data from 124 dph were analysed using regression, and tested against the null hypothesis that rotifer I content had no effect on outcome. Differences among means were considered significant at p<0.05. 254 255 256 257 258 Pre Prin ts Pre Prin ts 3 Results 3 cod larvae growth : No statistically significant differences in cod larvae length (p > 0.08) or dry weight (p > 0.25) occurred between treatments, although high variation between tanks may have masked effects (Fig. 1). On average, cod larvae grew from 4.9 \u00b1 0.2 mm to 6.8 \u00b1 0.5 mm in length, and 0.065 \u00b1 0.002 mg to 0.27 \u00b1 0.07 mg fish-1 in dry weight from 5 to 30 dph (Fig. 1), representing a specific growth rate of 6.3% day-1 during this period. There were no statistical differences (p>0.39) in the cod larval survival adjusted for sampled larvae (see section 2.6) between treatments at 30 dph which were 28 \u00b1 2, 39 \u00b1 20 and 30 \u00b1 15 % for controls, MI and HI groups respectively, while the average for all groups was 32 \u00b1 13 %. Survivals based solely on densities in tanks at 30 dph without taking into account sampled larvae were 12 \u00b1 2, 19 \u00b1 12 and 12 \u00b1 10 % for controls, MI and HI groups respectively. Control rotifers contained 0.60 \u00b1 0.33 mg I kg-1 DW while HI+rotifers contained 129 \u00b1 101 mg I kg-1 DW (Table 1). Whole body I levels in cod larvae were significantly different between groups (p<0.01, Fig. 2). Cod larvae (5 dph) had a starting concentration of 4.0 \u00b1 0.3 mg I kg-1 DW. Between 9 and 30 dph average I concentrations were 1.6 \u00b1 0.3 mg I kg-1 DW for control larvae, while MI larvae had 3 fold higher levels (4.9 \u00b1 2.4 mg I kg-1 DW), and HI larvae 7 fold higher levels (11.0 \u00b1 3.3 mg I kg-1 DW) than controls. Other element concentrations were also affected by treatment in both rotifers and cod larvae. HI+rotifers contained less Fe and Mn than controls (Table 1), while HI larvae contained more Mn, Fe and Cu than controls and MI larvae at one or more time points (Fig. 2b,c,e). Both HI and MI larvae contained higher levels of Co than controls (Fig. 2d), while larval Zn and Se concentrations were unaffected by treatment (Fig. 2f,g). Rotifer macro mineral concentrations were unaffected by treatment (Table 1), but increased levels of Ca and Mg, and lower levels of P and K were observed in HI larvae in comparison to controls and/or MI larvae during the rotifer feeding period (Fig. 3). The rate of increase in cod larvae I concentrations decreased as dietary I levels (rotifer I concentration) increased, and thus the I concentration ratio between cod larvae and rotifers displayed a negative trend (Fig 4; p<0.01). The age of the cod larvae did not affect their I concentration ratio (p = 0.96). The model predicts that the ratio of I in cod larvae versus rotifers equals 1 when rotifers have 3.5 mg I kg-1 DW (Fig. 4). 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 Pre Prin ts Pre Prin ts There were no differences in thyroid hormone levels or ratios between treatments (Fig. 5). Data was normalised to aid interpretation, as T3 results were higher than obtained previously (Penglase et al., 2010) and the high result was validated by the analysis of T3 with a second method (see methods). The total volume of thyroid follicle was 1.3 fold lower and the total epithelium volume was 1.4 fold lower per fish in HI versus MI larvae, but not controls, at 30 dph (Fig. 6a,c). The thyroid follicle colloid to epithelium ratio was higher in HI larvae than controls at 19 (1.7 fold) and 37 (1.8 fold) dph, while MI larvae did not differ from controls (Fig. 6d). No statistically significant differences were observed between groups in colloid volume or total number of thyroid follicles (Fig. 6b,e). Images of thyroid follicle sections from control and HI larvae at 37 dph are shown in figure 7. There were no significant differences in the weights (average; 2.50 \u00b1 0.19 g), lengths (6.14 \u00b1 0.16 cm) or condition factors (1.05 \u00b1 0.03) between groups at 124 dph (Table 2, n=400). Neck axis angle became closer to normal (180 \u00b1 3 degrees) with increasing I levels in rotifers, but there were no differences in any of the other skeletal deformity measurements (Table 2). 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 Pre Prin ts Pre Prin ts Discussion The hypothesis of this study was that commercially reared cod larvae fed rotifers would benefit from increased dietary I. The bases for this hypothesis were, one; rotifer I concentrations are often at the lower end or below juvenile/adult fish requirements (NRC, 2011), two; rotifers have 6 - 600 fold lower concentrations of I than copepods (Hamre et al., 2008b), the natural feed of cod larvae (Thompson and Harrop, 1991), three; cod larvae have better growth and development when fed copepods versus rotifers (Busch et al., 2010; Koedijk et al., 2010), four; I levels in copepod fed cod larvae are higher than rotifer fed cod larvae (Busch et al., 2010), and five; increased growth and/or survival has been observed in larval stages of several marine fish species fed or reared in environments with increased levels of bioavailable I (Hamre et al., 2008a; Witt et al., 2009; Ribeiro et al., 2011; Ribeiro et al., 2012). However, in contrast to the hypothesis, the increased thyroid follicle colloid to epithelium (C/E) ratio observed in this study indicates that I toxicity occurred in cod larvae fed rotifers with 129 mg I kg-1 DW. This observation purely in relation to the I level is not unexpected. A high C/E ratio in thyroid follicles is a classic symptom of I induced toxicity termed I (Baker, 2004) or colloid goitre, and occurs in mice at dietary I concentrations 10 fold higher than requirements with increasing severity developing with increasing I ingestion rates (Yang et al., 2006). What is interesting is that despite copepods containing 50 \u2013 350 mg I kg-1 DW (Solbakken et al., 2002; Hamre et al., 2008b), cod larvae have either similar or lower thyroid follicle C/E ratios when fed copepods compared to rotifers (Gr\u00f8tan, 2005). Furthermore, I concentrations observed in cod larvae fed natural zooplankton (29 mg I kg-1 DW at 27 dph; (Busch et al., 2010)) were 2.2 fold higher than the highest level observed in the current study (HI larvae, 30 dph; 13 \u00b1 4 mg I kg-1 DW). Thus it appears that high I concentrations in copepods do not induce morphological changes in thyroid follicles consistent with I toxicity, but do appear to be effectively transferred from copepods to fish larvae upon consumption. It is possible that copepods do not induce I toxicity in fish larvae due to nutrient interactions. For example, I toxicity can be prevented by the simultaneous presence of the bromine anion (Br-) in animals ranging from chicks (Gallus gallus) (Baker, Parr and Augspurger, 2003) to Artemia (S. Penglase et al., unpublished data). The exact mechanism for this Br-/I- interaction is still unknown, but it has been demonstrated that Br- decreases iodide accumulation in the thyroid follicles and increases I excretion via the kidneys in rats (Pavelka, 2004). Although the bromide concentrations in whole copepods and rotifers are unknown, we speculate that copepods have relatively high levels of bromide reflecting the high levels found in other 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 Pre Prin ts Pre Prin ts marine organisms, and this bromide helps prevent I toxicity in fish larvae. In the marine ecosystem, bromine is naturally found at similar high concentrations as I in seaweed (Romaris-Hortas, Moreda-Pineiro and Bermejo-Barrera, 2009), adult fish (Arafa et al., 2000; Wan et al., 2010), and as part of the hard chitin structures of crustaceans such as crabs (Cribb et al., 2009; Schofield et al., 2009) and copepods (Perry, Grime and Watt, 1988). Thyroid hormone levels and ratios were similar between cod larvae groups, and this is probably due to the compensatory changes observed in the thyroid follicles. For example pathological changes of over 70% in thyroid gland morphology have been observed in dogs (Canis lupus familiaris) with little change in circulating TH levels (Graham, Refsal and Nachreiner, 2007), and significant changes in fish thyroid follicle morphology with little change in thyroid hormone levels have also been reported for fish (Hawkyard et al., 2011; Morris et al., 2011). Few differences were found between the control and MI larvae groups, with the exception of whole body I concentrations. The increased whole body I content of cod larvae demonstrates the effective transfer of I from the rotifers to the cod larvae. The current study differs to previous studies exploring the uptake of I in fish larvae, as it attempted to ensure the ingestion of graded levels of I by maintaining the I concentration in the prey organism up until the point of feeding. Srivastava et al. (2012) found that I leaches rapidly from rotifers after enrichment with sodium iodide. Previous studies have found no difference in the I concentration of cod larvae fed control or I supplemented rotifers (Hamre et al., 2008a; S. Penglase et al., unpublished data), and this is probably a consequence of I leaching from rotifers in the minimum 1.5 to 2 h period between rotifer enrichment and feeding of the rotifers to cod larvae in these studies. In the current study, cod larvae iodine level increases were proportionally smaller for each increase in dietary I levels; control fish were 2.7 fold higher, while MI were 5 fold lower and HI larvae were 12 fold lower in I than their respective diets. Body stores of minerals are a good indicator of status (Baker, 1986), and the decreasing level of I retention in cod larvae relative to feed I levels indicates that requirements were met at levels lower than those fed to MI larvae. Modelling of the ratio between cod larvae and rotifer I concentrations predicts that based on a ratio of 1:1, rotifer I concentrations of 3.5 mg kg-1 DW meet cod larvae requirements. Both food and water contribute to the I status of adult, juvenile (Lall, 2002) and larval fish (Witt et al., 2009; Ribeiro et al., 2011). Alongside the I content in the continuously 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 Pre Prin ts Pre Prin ts exchanging seawater (88 \u00b5g I L-1, Moren, Sloth and Hamre, 2008), the control rotifers in the current study with 0.6 mg I kg-1 DW appeared to prevent any gross signs of I deficiency in cod larvae. This is at the lower end of the 0.6 \u2013 1.1 mg I kg-1 DW recommended by the national research council (NRC, 2011) as the I requirements of juvenile/adult fish. The reason that symptoms of severe I deficiency such as classic goitre have been observed in other fish studies is probably due to water parameters. Clear signs of I deficiency occurred in fish larvae reared in either recirculation systems (Ribeiro et al., 2011; Ribeiro et al., 2012) or well water (Witt et al., 2009). Nitrate (NO3-) is goitrogenic as it competitively blocks iodide uptake by the sodium iodide symporter (Tonacchera et al., 2004), and NO3- at levels commonly found in recirculation systems causes goitre in sharks (Crow et al., 1998; Morris et al., 2011). Furthermore, ozone (O3) used as a disinfectant in recirculation systems readily oxidises bioavailable I, iodide (I-) to iodate (IO3-) (Sherrill, Whitaker and Wong, 2004). Dissolve iodate is presumed to have low bioavailablity for fish (Sherrill, Whitaker and Wong, 2004), and higher levels of iodate compared to iodide were correlated to poor growth and survival in well water reared pacific threadfin larvae (Witt et al., 2009). Thus in recirculation systems, the presence of high levels of goitrogens (NO3-) and low levels of dissolved bioavailable I (I-) may increase the dietary I requirements of fish larvae over those reared with continuous water exchange where nitrate and its precursors are continuously removed and iodide continuously replaced, such as in the current study. Along with thyroid follicle morphology, dietary I also influenced the mineral composition of cod larvae. While most of the tested mineral concentrations in MI larvae were similar to controls, HI larvae had 10 to 25% higher levels of Ca, Mg, Mn, Fe, Co and Cu and around 10% lower levels of P and K than controls at one or more time points within the rotifer feeding period. For most of the minerals, differences in levels were observed by the first sampling point (9 dph; 4 days of feeding on rotifers). The differences cannot be explained by the feed; the HI rotifers had \u2248 10% less Mn and Fe, and no statistical differences were observed in Ca, Mg, K, P, Cu or Co concentrations. Hamre et al. (2008a) found that cod larvae fed increased levels of both I and Se had an 8% increase in whole body copper levels, similar to this study. Nguyen et al. (2008) found increased or decreased copper levels (20%) in red sea bream (Pagrus major) larvae depending on whether they were fed rotifers enriched with manganese alone or alongside zinc. While it is known that I deficiency can alter mineral distribution and homeostasis of Cu, Mn, Fe and Zn (Giray et al., 2003), to our knowledge this is the first data demonstrating that I oversupply can also effect mineral homeostasis. Although 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 Pre Prin ts Pre Prin ts there were few differences in growth or skeletal deformities observed between treatments at 124 dph, there was a small but statistically significant improvement in the neck axis angle in the HI compared to the control and MI cod groups (Table 2), and this may be linked to the differences in cod mineral concentrations in the larval stage. Conclusion Iodine enriched rotifers increased the levels of I in cod larvae, although as I levels in rotifers increased the increases in cod larvae I levels became proportionally smaller. Few differences occurred between cod larvae reared on control diets with 0.6 mg I kg -1 DW and those reared on diets with 26 mg I kg-1 DW, while the I concentration ratio between cod larvae and rotifers suggests cod larvae have an I requirement of 3.5 mg I kg -1 rotifers DW. Rotifers with copepod levels of I (129 mg I kg-1 DW) changed cod larvae whole body concentration of many essential minerals and induced changes in thyroid follicles morphology consistent with colloid goitre. The data presents one of the first observations of dietary induced I toxicity in fish, and suggests that I toxicity in fish larvae may be determined to a greater extent by I bioavailability and nutrient interactions than by body burdens of I.",
+    "url": "https://peerj.com/articles/22/reviews/",
+    "review_1": "David Solomon \u00b7 Jan 8, 2013 \u00b7 Academic Editor\nACCEPT\nThank you for addressing the feedback from the reviewers. I look forward to seeing your manuscript in print.",
+    "review_2": "David Solomon \u00b7 Dec 21, 2012 \u00b7 Academic Editor\nMINOR REVISIONS\nThank you for submitting this impressive study for publication in PeerJ. I feel with a few revisions it will be an excellent addition to the literature.\n\nI apologize for the delay in providing you with feedback. I had difficulty finding reviewers who had the necessary expertise necessary and were able to review the manuscript.\n\nThe two reviewers provide a wealth of useful feedback I hope you will find helpful in revising your manuscript. There are a few what I feel to be key points that I would appreciate if you could address in a revision of the paper. Along with the four items below, please consider carefully the other feedback provided by the reviewers using your own judgment on what to implement in the revision.\n\n1. Although your study is fairly theoretical, it would seem it would have some important implications for educational practice with second language learners. As noted by both reviewers, the paper would be significantly enhanced if the discussion provided some specific suggestions on how what has been learned from your study could be applied in educational settings or in better preparing second language learners for entering these settings.\n\n2. While the research questions and hypotheses are discussed 133-155 as noted by the second reviewer, they are not stated as clearly as they could be. Just as suggestion, along with discussing them it might be helpful to state them specifically 1, 2,3 etc.\n\n3. As noted by the second reviewer, it is unclear how the 113 subjects for the second part of the study were selected. Please clarify how they were selected from the 582 students who participated in the first phase of the study.\n\n4. Please check the document carefully for typographical and grammatical errors.",
+    "review_3": "Sonia Crandall \u00b7 Dec 21, 2012\nBasic reporting\nThe purpose of the current study is to explore whether students\u2019 verbal working memory influences their academic achievement. The theoretical framework for this study focused on the challenges faced by students in an environment where they are learning in a second language. The authors make a compelling case for the study.\nExperimental design\nThe authors designed an elaborate study using both self-report and observed measures. The design is novel and interesting with a comprehensive statistical analysis plan that addresses multiple confounding variables. This human subjects approved study has been rigorously conducted.\nValidity of the findings\nThe findings are well justified. The authors provide interesting data displays that contribute to overall understanding of the results explained within the text. Study limitations are adequately addressed. Conclusions are sound, clearly reported and are connected to the original question.\nAdditional comments\nOverall this is an excellently conducted research study. However, the educational significance is missing. What are medical educators supposed to do with these results? Nearly all, if not all, medical schools have L2 learners and would benefit from the authors\u2019 recommendations for supporting students\u2019 learning. Educational research must support practice.\n\nNeeded revisions:\nThere are numerous acronyms to keep in one\u2019s mind while reading the paper. Is it possible to include an acronym table that can be referred to easily?\n\nAbstract: Please explain what is meant by tertiary medical students. The readership is international and many may not be familiar with this term.\n\nTypos:\nLine 295: The word data implies plural; thus, the verb should match: was should be were.\n\nLine 398: Cronbach is spelled incorrectly.\nCite this review as\nCrandall S (2013) Peer Review #1 of \"Poorer verbal working memory for a second language selectively impacts academic achievement in university medical students (v0.1)\". PeerJ https://doi.org/10.7287/peerj.22v0.1/reviews/1",
+    "review_4": "Reviewer 2 \u00b7 Dec 16, 2012\nBasic reporting\nIn the abstract, information like independent variable(s), dependent variable(s), expected results and actual results all should be included. However, the authors didn\u2019t include briefly stated results in the abstract. In addition, as an expected factor which may influence the results of the study, the age of English language acquisition was left behind in the abstract too.\n\nThe paper also emphasizes the role of age of language acquisition in speech perception, mentioning several differences in the ability to process speech between monolinguals and late bilinguals. It would be better if the authors show expected and actual results about the effects of age of second language acquisition in the abstract of the paper.\nIn the introduction, the article should make more clarification about what the theoretical and empirical significance of the current study is and how the study can contribute to the understanding of the impact of verbal working memory on academic attainment, particularly for English as second language users.\n\nThe hypothesis/hypotheses of the study should be more clearly defined. It seemed in the abstract that the authors hypothesized poorer working memory under high noisy contexts significantly reduced the academic performance for L2. But the paper neither clearly stated the hypothesis/hypotheses or research question(s) in the discussion, nor did it explain why and how other constructs such as English proficiency (e.g., AoAoE, PEP) related to the hypothesis/hypotheses or research question(s).\n\nBesides, the paper discussed in great details about why English proficiency would not be a confounding variable in the relationship between vWM and academic performance because all international students have to take standard tests in order to get the admission. However, this is a fair weak argument since passing stringent measures of English proficiency prior to enrolment, such as the IELTS or the TOEFL, would not guarantee that international students would performance as well as their local counterparts in academic achievements. In addition, in later examination, the paper extended to consider English proficiency as an indicator of academic achievements in the research design. It could be very confusing for readers to identify the role of English proficiency in this article.\n\nMoreover, the evidence exhibited in the article does not sufficiently support the relationship between the \u201cability to process speech and to recall academic material.\u201d Only one study (e.g., Ljung et al., 2010) from line 118 to 124 was cited for illustration. Given that this relationship is crucial for developing hypothesis/hypotheses, more relevant research from the extant literature would be better to enhance the argument.\nExperimental design\n1. There\u2019s no reliability report for the Perceived Stress Scale and the Index of learning Styles Questionnaire used in the research design.\n\n2. Stress and learning styles of the medical undergraduate students were measured as two potential factors which may affect their academic achievements. However, the authors failed to justify the reason why these two factors may or may not cause an impact. Except for two studies mentioned in line 165 to 169, there were insufficient literature reviews and explications to support this point.\n\n3. Participants: Line 182 to 189 showed that 103 students were recruited into the complete research project. However, the authors didn\u2019t clearly mention where they recruited the 54 participants for the year second study. Were they from the 103 participants of the first year study? Or were they from the 582 students who returned the surveys and different from the 103 students? Why there was a half loss of the sample? Would the sample loss cause any impact on the expected effect size?\n\n4. The criterion that distinguishes \u201clocal\u201d and \u201cinternational\u201d cannot rule out the possibility that an increasing number of students in \u201cmigrant families\u201d who hold the permanent residency but not a L1 student or cannot speak English fluently and idiomatically. This may constitute a confound effect in the comparison between L1 and L2.\n\n5. The authors didn\u2019t rationalize and explain clearly the investigation of students\u2019 academic achievements for both first and second year of the course. It is not clarified how they collected the second year\u2019s data.\n\n6. In terms of the limitations of the study, the within-subject experimental design may result in participants\u2019 fatigue particularly when each participant in the study was required to listen and verbally repeat 60 sentences in total under three noise backgrounds (high, moderate and low). The author explained that high noise group would not be the first condition in order to putting the participants in a difficult situation at the very beginning. However, there is still a possibility that the participants who should have performed well scored low in the last group because of fatigue.\n\n7. The use of everyday clinically-used sentences may pose a threat to ecological validity. Given that students are taught academic contents, which are different semantically with everyday language (e.g., academic lecture use more long sentences and professional vocabulary), in undergraduate medical courses, they may not be able to fully emerge in experimental settings.\nValidity of the findings\n1. From line 367 to 374, when conducting the multiple regressions, the authors used five items significantly correlated to SNR50 that pertained to English proficiency and/or usage, such as AoAoE, PEP, MSE. However, there were insufficient justifications and supportive literature review to explain why the authors did this subdividing and how did these five items pertained to English proficiency and usage.\n2. In line 395, the authors did the principle component analysis for the five items pertaining to English proficiency and usage to rule out the threat of multicollinearity. What\u2019s the VIF and tolerance for the new scale renamed \u201cEnglish Language Skills\u201d (ELS)? Have the multicollinearity problem been solved successfully?\n3. The authors indicated that the new ELS was an approximation of the students\u2019 overall English proficiency. How did they draw out this conclusion? Is there any empirical evidence or related literature to support this point?\n4. In the discussion section, the authors classified the background noise into 2 groups: energetic and informational, which a) did not have literature to support the categorization; b) were ambiguous in the way that how the categorization fit into the hypothesis and the design of the experiment. According to the article, the author mentioned the use of headphone to play noise in line 232 and \u201cbabble noise\u201d in line 219, but it is not clear what category do these belong to.\nAdditional comments\nThe article includes quite a few grammatical and stylistic mistakes. For example, when listing previous literatures, \u201cet al.\u201d has been used as \u201cet al\u201d without the period in a few places. In line 489, it is not necessary to list all three authors and put on \u201cet al.\u201d at the end. When mentioning \u201clocal participants\u201d and \u201cinternational participants\u201d, the capitalization of the letter \"L\" in the word \"local\" should be consistent in the paper. For example, in line 137, \"L\" is not capitalized, while it is in line 457. The letter \"I\" in the word \"international\" is another case. In line 516, it is better to explain what are \u201cPSE; the SNR50\u201d used for (are they examples or illustrations). Also, like in the line of 235, the first line of the paragraph should be indented. From the line of 479 to 485, the format of the text needs to be fixed.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"Poorer verbal working memory for a second language selectively impacts academic achievement in university medical students (v0.1)\". PeerJ https://doi.org/10.7287/peerj.22v0.1/reviews/2",
+    "pdf_1": "https://peerj.com/articles/22v0.2/submission",
+    "pdf_2": "https://peerj.com/articles/22v0.1/submission",
+    "all_reviews": "Review 1: David Solomon \u00b7 Jan 8, 2013 \u00b7 Academic Editor\nACCEPT\nThank you for addressing the feedback from the reviewers. I look forward to seeing your manuscript in print.\nReview 2: David Solomon \u00b7 Dec 21, 2012 \u00b7 Academic Editor\nMINOR REVISIONS\nThank you for submitting this impressive study for publication in PeerJ. I feel with a few revisions it will be an excellent addition to the literature.\n\nI apologize for the delay in providing you with feedback. I had difficulty finding reviewers who had the necessary expertise necessary and were able to review the manuscript.\n\nThe two reviewers provide a wealth of useful feedback I hope you will find helpful in revising your manuscript. There are a few what I feel to be key points that I would appreciate if you could address in a revision of the paper. Along with the four items below, please consider carefully the other feedback provided by the reviewers using your own judgment on what to implement in the revision.\n\n1. Although your study is fairly theoretical, it would seem it would have some important implications for educational practice with second language learners. As noted by both reviewers, the paper would be significantly enhanced if the discussion provided some specific suggestions on how what has been learned from your study could be applied in educational settings or in better preparing second language learners for entering these settings.\n\n2. While the research questions and hypotheses are discussed 133-155 as noted by the second reviewer, they are not stated as clearly as they could be. Just as suggestion, along with discussing them it might be helpful to state them specifically 1, 2,3 etc.\n\n3. As noted by the second reviewer, it is unclear how the 113 subjects for the second part of the study were selected. Please clarify how they were selected from the 582 students who participated in the first phase of the study.\n\n4. Please check the document carefully for typographical and grammatical errors.\nReview 3: Sonia Crandall \u00b7 Dec 21, 2012\nBasic reporting\nThe purpose of the current study is to explore whether students\u2019 verbal working memory influences their academic achievement. The theoretical framework for this study focused on the challenges faced by students in an environment where they are learning in a second language. The authors make a compelling case for the study.\nExperimental design\nThe authors designed an elaborate study using both self-report and observed measures. The design is novel and interesting with a comprehensive statistical analysis plan that addresses multiple confounding variables. This human subjects approved study has been rigorously conducted.\nValidity of the findings\nThe findings are well justified. The authors provide interesting data displays that contribute to overall understanding of the results explained within the text. Study limitations are adequately addressed. Conclusions are sound, clearly reported and are connected to the original question.\nAdditional comments\nOverall this is an excellently conducted research study. However, the educational significance is missing. What are medical educators supposed to do with these results? Nearly all, if not all, medical schools have L2 learners and would benefit from the authors\u2019 recommendations for supporting students\u2019 learning. Educational research must support practice.\n\nNeeded revisions:\nThere are numerous acronyms to keep in one\u2019s mind while reading the paper. Is it possible to include an acronym table that can be referred to easily?\n\nAbstract: Please explain what is meant by tertiary medical students. The readership is international and many may not be familiar with this term.\n\nTypos:\nLine 295: The word data implies plural; thus, the verb should match: was should be were.\n\nLine 398: Cronbach is spelled incorrectly.\nCite this review as\nCrandall S (2013) Peer Review #1 of \"Poorer verbal working memory for a second language selectively impacts academic achievement in university medical students (v0.1)\". PeerJ https://doi.org/10.7287/peerj.22v0.1/reviews/1\nReview 4: Reviewer 2 \u00b7 Dec 16, 2012\nBasic reporting\nIn the abstract, information like independent variable(s), dependent variable(s), expected results and actual results all should be included. However, the authors didn\u2019t include briefly stated results in the abstract. In addition, as an expected factor which may influence the results of the study, the age of English language acquisition was left behind in the abstract too.\n\nThe paper also emphasizes the role of age of language acquisition in speech perception, mentioning several differences in the ability to process speech between monolinguals and late bilinguals. It would be better if the authors show expected and actual results about the effects of age of second language acquisition in the abstract of the paper.\nIn the introduction, the article should make more clarification about what the theoretical and empirical significance of the current study is and how the study can contribute to the understanding of the impact of verbal working memory on academic attainment, particularly for English as second language users.\n\nThe hypothesis/hypotheses of the study should be more clearly defined. It seemed in the abstract that the authors hypothesized poorer working memory under high noisy contexts significantly reduced the academic performance for L2. But the paper neither clearly stated the hypothesis/hypotheses or research question(s) in the discussion, nor did it explain why and how other constructs such as English proficiency (e.g., AoAoE, PEP) related to the hypothesis/hypotheses or research question(s).\n\nBesides, the paper discussed in great details about why English proficiency would not be a confounding variable in the relationship between vWM and academic performance because all international students have to take standard tests in order to get the admission. However, this is a fair weak argument since passing stringent measures of English proficiency prior to enrolment, such as the IELTS or the TOEFL, would not guarantee that international students would performance as well as their local counterparts in academic achievements. In addition, in later examination, the paper extended to consider English proficiency as an indicator of academic achievements in the research design. It could be very confusing for readers to identify the role of English proficiency in this article.\n\nMoreover, the evidence exhibited in the article does not sufficiently support the relationship between the \u201cability to process speech and to recall academic material.\u201d Only one study (e.g., Ljung et al., 2010) from line 118 to 124 was cited for illustration. Given that this relationship is crucial for developing hypothesis/hypotheses, more relevant research from the extant literature would be better to enhance the argument.\nExperimental design\n1. There\u2019s no reliability report for the Perceived Stress Scale and the Index of learning Styles Questionnaire used in the research design.\n\n2. Stress and learning styles of the medical undergraduate students were measured as two potential factors which may affect their academic achievements. However, the authors failed to justify the reason why these two factors may or may not cause an impact. Except for two studies mentioned in line 165 to 169, there were insufficient literature reviews and explications to support this point.\n\n3. Participants: Line 182 to 189 showed that 103 students were recruited into the complete research project. However, the authors didn\u2019t clearly mention where they recruited the 54 participants for the year second study. Were they from the 103 participants of the first year study? Or were they from the 582 students who returned the surveys and different from the 103 students? Why there was a half loss of the sample? Would the sample loss cause any impact on the expected effect size?\n\n4. The criterion that distinguishes \u201clocal\u201d and \u201cinternational\u201d cannot rule out the possibility that an increasing number of students in \u201cmigrant families\u201d who hold the permanent residency but not a L1 student or cannot speak English fluently and idiomatically. This may constitute a confound effect in the comparison between L1 and L2.\n\n5. The authors didn\u2019t rationalize and explain clearly the investigation of students\u2019 academic achievements for both first and second year of the course. It is not clarified how they collected the second year\u2019s data.\n\n6. In terms of the limitations of the study, the within-subject experimental design may result in participants\u2019 fatigue particularly when each participant in the study was required to listen and verbally repeat 60 sentences in total under three noise backgrounds (high, moderate and low). The author explained that high noise group would not be the first condition in order to putting the participants in a difficult situation at the very beginning. However, there is still a possibility that the participants who should have performed well scored low in the last group because of fatigue.\n\n7. The use of everyday clinically-used sentences may pose a threat to ecological validity. Given that students are taught academic contents, which are different semantically with everyday language (e.g., academic lecture use more long sentences and professional vocabulary), in undergraduate medical courses, they may not be able to fully emerge in experimental settings.\nValidity of the findings\n1. From line 367 to 374, when conducting the multiple regressions, the authors used five items significantly correlated to SNR50 that pertained to English proficiency and/or usage, such as AoAoE, PEP, MSE. However, there were insufficient justifications and supportive literature review to explain why the authors did this subdividing and how did these five items pertained to English proficiency and usage.\n2. In line 395, the authors did the principle component analysis for the five items pertaining to English proficiency and usage to rule out the threat of multicollinearity. What\u2019s the VIF and tolerance for the new scale renamed \u201cEnglish Language Skills\u201d (ELS)? Have the multicollinearity problem been solved successfully?\n3. The authors indicated that the new ELS was an approximation of the students\u2019 overall English proficiency. How did they draw out this conclusion? Is there any empirical evidence or related literature to support this point?\n4. In the discussion section, the authors classified the background noise into 2 groups: energetic and informational, which a) did not have literature to support the categorization; b) were ambiguous in the way that how the categorization fit into the hypothesis and the design of the experiment. According to the article, the author mentioned the use of headphone to play noise in line 232 and \u201cbabble noise\u201d in line 219, but it is not clear what category do these belong to.\nAdditional comments\nThe article includes quite a few grammatical and stylistic mistakes. For example, when listing previous literatures, \u201cet al.\u201d has been used as \u201cet al\u201d without the period in a few places. In line 489, it is not necessary to list all three authors and put on \u201cet al.\u201d at the end. When mentioning \u201clocal participants\u201d and \u201cinternational participants\u201d, the capitalization of the letter \"L\" in the word \"local\" should be consistent in the paper. For example, in line 137, \"L\" is not capitalized, while it is in line 457. The letter \"I\" in the word \"international\" is another case. In line 516, it is better to explain what are \u201cPSE; the SNR50\u201d used for (are they examples or illustrations). Also, like in the line of 235, the first line of the paragraph should be indented. From the line of 479 to 485, the format of the text needs to be fixed.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"Poorer verbal working memory for a second language selectively impacts academic achievement in university medical students (v0.1)\". PeerJ https://doi.org/10.7287/peerj.22v0.1/reviews/2\nReview 5: nan\nReview 6: \n",
+    "LLM_extracted_review": "['Review 1: Thank you for addressing the feedback from the reviewers. I look forward to seeing your manuscript in print.\\n\\nReview 2: \\n1. The paper would be significantly enhanced if the discussion provided specific suggestions on how the study\\'s findings could be applied in educational settings for second language learners.\\n2. The research questions and hypotheses are not stated as clearly as they could be; it might be helpful to list them specifically.\\n3. It is unclear how the 113 subjects for the second part of the study were selected; clarification is needed.\\n4. The document should be checked for typographical and grammatical errors.\\n\\nReview 3: Overall, this is an excellently conducted research study. However, the educational significance is missing; recommendations for supporting students\u2019 learning should be included. \\n- An acronym table is needed for clarity.\\n- The term \"tertiary medical students\" should be explained in the abstract for an international readership.\\n- There are typographical errors that need correction.\\n\\nReview 4: \\n1. The abstract should include independent and dependent variables, expected results, and actual results.\\n2. The theoretical and empirical significance of the study should be clarified in the introduction.\\n3. The hypothesis/hypotheses should be more clearly defined.\\n4. The argument regarding English proficiency as a confounding variable is weak and needs more justification.\\n5. The recruitment process for participants is unclear, and the sample loss should be addressed.\\n6. The investigation of academic achievements for both first and second year students needs clarification.\\n7. The within-subject experimental design may lead to participant fatigue, affecting results.\\n8. The use of everyday language may threaten ecological validity in the context of academic learning.\\n\\nValidity of the findings:\\n1. Insufficient justification for the subdivision of items related to English proficiency.\\n2. Clarification is needed on the multicollinearity analysis and its results.\\n3. The conclusion regarding the new English Language Skills scale needs empirical support.\\n4. The categorization of background noise lacks literature support and clarity.\\n\\nAdditional comments: The article contains grammatical and stylistic mistakes that need correction.\\n\\nReview 5: nan\\n\\nReview 6: ']"
+}

peerj_json_files/PeerJ_Json_120.json ADDED Viewed

	@@ -0,0 +1,15 @@

+{
+    "v1_col_introduction": "introduction  : The assessment of the clinical competence of a medical student is challenging. A\ncompetency is \"... an observable ability of a health professional related to a specific activity that integrates knowledge, skills, values, and attitudes. Since they are observable, they can be measured and assessed.\" Although seemingly straight forward, competency based education is of limited usefulness in guiding the design and implementation of educational experiences if they are not tied to specific learning objectives.(1) Additionally, learning objectives are of limited usefulness if they are not available to students and faculty when interacting with patients. Finally, observation and assessment help neither students nor patients if they are not captured and documented in a way that facilitates learner specific plans for improvement and excellence. We present a generalizable initiative that makes national curricula functional in local learning environments, and improves and simplifies observation based assessments and performancebased data tracking for faculty and learners.\nMaterials\u00a0&\u00a0Methods Content\u00a0Manager\nWe developed a mobile, cloud-based application called just in time medicine (or JIT) that\nfunctions effectively on smart phones, tablets and laptop computers. The mobile application is supported by a self-service web-based content management system designed with the explicit aim of enabling users with average computing skills to build their own customizable content, including criterion-based checklists that can then be delivered to any internet enabled device such as a smart phone or tablet.\nFor this project, we utilized nineteen core training problems from the nationally validated\nClerkship Directors in Internal Medicine (CDIM) curriculum and combined these training\n14\n15\n16\n17\n18\n19\n20\n21\n22\n23\n24\n25\n26\n27 28\n29\n30\n31\n32\n33\n34\n35\n36\nPeerJ reviewing PDF | (v2013:05:520:1:0:NEW 17 Aug 2013)\nR ev ie w in g M an\nus cr ip t\nproblems with the observable competencies of communication skills, history taking and physical examination to create problem and task specific checklists. For each assessment, the software calculates the students\u2019 performance by determining the percentage of all potential items performed correctly, and an algorithm generated grade of \u201cnot done/unsatisfactory\u201d, \u201cneeds improvement\u201d or \u201cwell done\u201d is calculated depending on the percentage of items performed correctly. In general if a student achieved 80% of the expected items correctly they received a \u201cwell done grade\u201d; performing 30 \u2013 79% of the expected items resulted in a \u201cneeds improvement\u201d grade, and < 30% an unsatisfactory grade. Figures 1 - 2 present screen shots for the process of building checklists using our content manager for the problem altered mental status and the competency history taking. Additionally, Figures 3 - 4 show how the assessment tools are displayed on the user\u2019s device. Figures 5 - 7 show the permanent cloud-based reporting options associated with individual assessments. A fully functional version of JIT can be accessed at: www.justintimemedicine.com/mobile; log in username is testuser@journal.com, and the password is test.\nGoals\u00a0and\u00a0hypotheses \u00a0 In introducing the JIT in our clerkship, we hypothesized that JIT would: 1) facilitate the direct observation and provision of feedback to trainees on their clinical competencies; 2) generally be accepted by faculty; 3) provide a means for recording the observations of trainee performance, and 4) possess adequate reliability and validity. Setting\nThe College of Human Medicine (CHM) at Michigan State University is a community-\nbased medical school with clinical training in 7 communities throughout Michigan. Between July 2010 and October 2012 we implemented JIT as an integral part of the internal medicine clerkship among 367 students. Each student was required to complete ten directly observed\n37\n38\n39\n40\n41\n42\n43\n44\n45\n46\n47\n48\n49\n50\n51 52 53\n54\n55\n56\n57\n58\n59\n60\n61\nPeerJ reviewing PDF | (v2013:05:520:1:0:NEW 17 Aug 2013)\nR ev ie w in g M an\nus cr ip t\nclinical evaluation exercises (i.e. CEX\u2019s) with real patients in authentic clinical settings. A CEX is a short (generally < 20 minutes) directly observed trainee \u2013 patient interaction (e.g. historytaking, examination, counseling etc.); faculty observes rates and provides written comments on the interaction. Students received an orientation to the CEX application and were required to become familiar with the software. Evaluators (attending faculty and residents) received an email on the importance of direct observation and the basic functionality of the CEX application.\nIn general, students chose the patient, problem and competency upon which to be\nassessed. At the time of the assessment, students handed their mobile device, with the checklists displayed for evaluator use during the assessed interaction. A total of 516 evaluators subsequently used JIT to guide their observations and assessments of students\u2019 interacting with patients.\nData\u00a0Collection\nWe collected the following data: the specific training problems and competencies\nobserved and assessed by the evaluators, the grades associated with the observation and descriptive data from faculty on the use of JIT. Descriptive data was collected from the faculty via \u201cpull-down\u201d menus located on the last screen of each assessment. A screen shot of the interface is displayed in figure 4.\nReliability\u00a0and\u00a0validity\u00a0assessments\u00a0\nA group of 17 evaluators, 9 internal medicine residents and 8 general internist faculty\nmembers viewed and rated six scripted videotaped encounters using JIT. Each case was scripted for both satisfactory and unsatisfactory performance. These cases have been previously validated by Holmboe as representing levels of competence which range from unequivocally poor to satisfactory.(2) The sample of raters reflected the number we could reasonably obtain given our small general internal medicine faculty and residency program. We felt it was adequate to provide\n62\n63\n64\n65\n66\n67\n68\n69\n70\n71\n72\n73\n74\n75\n76\n77\n78\n79\n80\n81\n82\n83\n84\n85\nPeerJ reviewing PDF | (v2013:05:520:1:0:NEW 17 Aug 2013)\nR ev ie w in g M an\nus cr ip t\na stable estimate of the inter rater reliability of the assessment process. We calculated the inter rater reliability using a formula developed by Ebel and implemented using software developed by one of the authors.(3, 4) All other statistical analyses were performed with SPSS version 21.",
+    "v2_col_introduction": "introduction  : The assessment of the clinical competence of a medical student is challenging. A\ncompetency is \"... an observable ability of a health professional related to a specific activity that integrates knowledge, skills, values, and attitudes. Since they are observable, they can be measured and assessed.\" Although seemingly straight forward, competency based education is of limited usefulness in guiding the design and implementation of educational experiences if they are not tied to specific learning objectives.(1) Additionally, learning objectives are of limited usefulness if they are not available to students and faculty when interacting with patients. Finally, observation and assessment help neither students nor patients if they are not captured and documented in a way that facilitates learner specific plans for improvement and excellence. We present a generalizable initiative that makes national curricula functional in local learning environments, and improves and simplifies observation based assessments and performance-based data tracking for faculty and learners.\nMaterials\u00a0&\u00a0Methods Content\u00a0Manager\nWe developed a mobile, Cloud-based application called just in time medicine (or JIT)\nthat functions effectively on smart phones, tablets and laptop computers. The mobile application is supported by a self-service web-based content management system designed with the explicit aim of enabling users with average computing skills to build their own customizable content, including criterion-based checklists that can then be delivered to any internet enabled device such as a smart phone or tablet.\n2\n16 17\n18\n19\n20\n21\n22\n23\n24\n25\n26\n27\n28\n29 30 31 32\n33\n34\n35\n36\n37\nPeerJ reviewing PDF | (v2013:05:520:0:1:NEW 26 May 2013)\nR ev ie w in g M an\nus cr ip t\nFor this project, we utilized nineteen core training problems from the nationally\nvalidated Clerkship Directors in Internal Medicine (CDIM) curriculum and combined these training problems with the observable competencies of communication skills, history taking and physical exam to create problem and task specific checklists. For each assessment, the software calculates the students\u2019 performance by determining the percentage of all potential items performed correctly, and an algorithm generated grade of \u201cnot done/unsatisfactory\u201d, \u201cneeds improvement\u201d or \u201cwell done\u201d is calculated depending on the percentage of items performed correctly. Figures 1 - 3 present screen shots for the process of building checklists using our content manager for the problem altered mental status and the competency history taking. Additionally, Figures 4 \u2013 6b show the permanent Cloud-based reports associated with the assessments. Access JIT at www.justintimemedicine.com/mobile; UN: testuser@msu.edu PW: testuser.\nGoals\u00a0and\u00a0hypotheses \u00a0 In introducing the JIT in our clerkship, we hypothesized that JIT would: 1) facilitate the direct observation and provision of feedback to trainees on their clinical competencies; 2) generally be accepted by faculty; 3) provide a means for recording the observations of trainee performance, and 4) possess adequate reliability and validity. Setting\nThe College of Human Medicine (CHM) at Michigan State University is a\ncommunity-based medical school with clinical training in 7 communities throughout Michigan. Between July 2010 and October 2012 we implemented JIT as an integral part of the internal medicine clerkship among 367 students. Each student was required to complete ten directly observed clinical evaluation exercises (i.e. CEX\u2019s) with real patients in authentic clinical\n3\n38\n39\n40\n41\n42\n43\n44\n45\n46\n47\n48\n49\n50 51 52\n53\n54\n55\n56\n57\n58\n59\n60\n61\nPeerJ reviewing PDF | (v2013:05:520:0:1:NEW 26 May 2013)\nR ev ie w in g M an\nus cr ip t\nsettings. Students received an orientation to the CEX app and were required to become familiar with the software. Evaluators (attending faculty and residents) received an email on the importance of direct observation and the basic functionality of the CEX app.\nIn general, students chose the patient, problem and competency upon which to be\nassessed. At the time of the assessment, students handed their mobile device, with the checklists displayed for evaluator use during the assessed interaction. A total of 516 evaluators subsequently used JIT to guide their observations and assessments of students\u2019 interacting with patients.\nData\u00a0Collection\nWe collected the following data: the specific training problems and competencies\nobserved and assessed by the evaluators, the grades associated with the observation and descriptive data from faculty on the use of JIT.\nReliability\u00a0and\u00a0validity\u00a0assessments\u00a0\nA group of 17 evaluators viewed and rated six scripted videotaped encounters using JIT.\nEach case was scripted for both satisfactory and unsatisfactory performance. These cases have been previously validated by Holmboe as representing levels of competence which range from unequivocally poor to satisfactory.(2) To assess predictive validity, we also correlated \u201cgateway\u201d performance assessment examinations taken by 282 students at the end of their third year required clerkships with the CEX assessments obtained by JIT.\nHuman\u00a0Use\nOur medical school has created an \u201cHonest Broker System\u201d for conducting research on\nstudent performance data that are collected as a regular part of the students\u2019 educational\n4\n62\n63\n64\n65\n66\n67\n68\n69\n70 71 72\n73\n74\n75 76 77\n78\n79\n80\n81\n82\n83 84 85\n86\nPeerJ reviewing PDF | (v2013:05:520:0:1:NEW 26 May 2013)\nR ev ie w in g M an\nus cr ip t\nactivities. A designated employee of the medical school, with access to these data, has been designated as the \u201cHonest Broker\u201d. In an honest broker system a person or agency that has access to multiple human subject datasets collected for non-research purposes creates a de-identified dataset that can be used for research purposes without posing risk to the subjects. (3) This approach has been used in various types of clinical research \u2013 though, to the best of our knowledge, it has not been applied in educational research at other institutions.\nAt Michigan State University this designated individual created the dataset used in this\nstudy and made it available to our research team after removing all identifiers. The Social Science/Behavioral/Education Institutional Review Board (SIRB) of Michigan State University has reviewed and determined based on 45 CRF 46(f) these data do not involve human subjects and does not need IRB review.",
+    "v1_text": "results : Number and types of evaluations Five hundred sixteen evaluators used the application to assess 367 students for a total of 3567 separate assessments. The number of CEX\u2019s completed per student was 9.7 (\u00b1 1.8) and the average number of CEX\u2019s completed per faculty was 6.9 (\u00b1 15.8). The average number of training problems a student was assessed on was 6.7; of the three competency domains of communication skills, history taking, and physical examination 68% of the students had at least one evaluation in each of the three categories. In terms of the grades, time variables and satisfaction, ~ 83% of the encounters were associated with a \u201cwell done\u201d grade, and on average students were credited with performing ~ 86% of the items correctly. (Figure 8) Between 43 \u2013 50% of the CEX assessments took < 10 minutes as estimated by the faculty, and in ~ 69% of the encounters feedback was estimated to occur in less than 10 minutes. In 92% of the encounters, faculty rated that they were either satisfied or highly satisfied with the CEX. The estimated inter-rater reliability of a single rater observing the videotaped encounters was 0.69 (slightly higher for faculty at 0.74 vs. residents at .64). In judging the same simulated patient case scripted to be satisfactory and non-satisfactory, the residents and faculty using JIT discriminated between the satisfactory and non-satisfactory performance. The mean number of items checked for the videotapes scripted for unsatisfactory performance was 35% vs. 59% for those scripted for more satisfactory performance. We believe this provides evidence supporting the construct validity of JIT. 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 PeerJ reviewing PDF | (v2013:05:520:1:0:NEW 17 Aug 2013) R ev ie w in g M an us cr ip t To assess predictive validity, we calculated a Pearson product moment correlation between a \u201cgateway\u201d performance assessment examinations taken by 282 students at the end of their third year required clerkships with the CEX assessments obtained by JIT. There was a small (but statistically significant 0.144, p = .008) correlation between students\u2019 CEX scores and communications skills in the gateway performance assessment exam. discussion : Although national learning objectives have been published for all core clerkships, their usefulness for assessing learning outcomes has been limited. As an example, the core competency gathering essential and accurate information seems relatively straight forward. However, when applied to a single condition such as chronic obstructive pulmonary disease, there are at least 28 specified clinical tasks related to history taking and performing a physical examination that a student should demonstrate to meet the expected outcomes as defined in the Clerkship Directors in Internal Medicine (CDIM) curricular objectives for that problem. Of these 28, how many will a faculty evaluator remember when assessing the student? More importantly how many can they remember and what level of consistency will there be among preceptors providing feedback to students? If we take almost any clinical skill and start to dissect it, we find very quickly that existing human memory is insufficient in recalling all of the explicit steps related to potentially hundreds of conditions that help frame the expected outcomes of a trainee\u2019s educational experience and curricula. As the expectations for assessment of discrete competencies increases, the evaluation burden for educators, students and administrators becomes progressively more educationally incomplete and logistically unmanageable. 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 PeerJ reviewing PDF | (v2013:05:520:1:0:NEW 17 Aug 2013) R ev ie w in g M an us cr ip t The inability of faculty to remember and accurately assess for outcomes related to potentially hundreds of discrete educational objectives while evaluating trainees in clinical settings is one of the major reasons faculty have a hard time reliably discriminating unsatisfactory from satisfactory performance, as has been noted by many authors over the past decade using paper-based systems.(2, 5) For example, in a study of mini-CEX evaluations among 300 medical students, Hill noted that problems existed \u201cin trying to ensure that everyone was working to the same or similar standards.\u201d(6) In another study of 400 mini-CEX assessments, Fernando concluded faculty evaluators were unsure of the level of performance expected of the learners.(7) Hasnain noted that poor agreement among faculty evaluating medical students on a Family Medicine clerkship was due to the fact that \u201cStandards for judging clinical competence were not explicit\u201d.(8) In a randomized trial of a faculty development effort, Holmboe studied the accuracy of faculty ratings by having them view videotaped trainee-patient encounters that were scripted to portray three levels of proficiency; unsatisfactory, marginal or satisfactory. Faculty viewing the exact same encounter varied widely in their assessment of trainee competence, with ratings from unequivocally unsatisfactory (CEX scores of scores 1 \u2013 3) to unequivocally superior (CEX scores of 7 \u2013 9), regardless of whether the video was scripted to be unsatisfactory or not. After an intensive 4 day faculty development workshop in which participants were tasked with developing a shared mental model of what specific competencies should look like, problems still existed among faculty in discriminating satisfactory from unsatisfactory performance in these scripted encounters.(2) Kogan noted that in the absence of easily accessible frameworks, faculty evaluators default back to a myriad of highly variable evaluation strategies including such idiosyncratic features as instinct, \u201cgut feelings\u201d, \u201cunsubstantiated assumptions\u201d and the faculty members\u2019 emotional response to providing feedback. What she also noted was that faculty raters commonly 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 PeerJ reviewing PDF | (v2013:05:520:1:0:NEW 17 Aug 2013) R ev ie w in g M an us cr ip t fail to use existing frameworks or external standards in guiding their evaluations of trainees, thus explaining much of the well-recognized problems with poor validity and inter-rater reliability associated with clinical evaluations.(5) Given these realities, it is not surprising that medical trainees commonly do not view the feedback received from faculty as credible nor influential in learning, especially if the feedback was not immediate and tied to the trainees\u2019 clinical work-place performance. (9) Enhancing the effectiveness of clinical assessments, the delivery of feedback related to learning objectives and the creation of better systems for documenting faculty observations are commonly cited needs in medical education.(8, 10-13) Given these and other trends, systems that are capable of disseminating curricular objectives to students and faculty and which also enable criterion-based assessment have become key educational needs. We believe that cloud-based technology, appropriately applied to maximize efficiency, can contribute to optimizing the learning environment by directly aligning learning objectives from national disciplinary curricula with assessment tools for use by students and faculty anywhere and anytime, especially at the bedside. In our first feasibility study, we demonstrated our ability to deliver national educational objectives published by the CDIM to electronic hand-held personal digital assistants (PDAs) such as Palm\u00ae and PocketPC\u00ae devices.(14) In a second feasibility study, we subsequently demonstrated this system could be used to deliver and successfully implement competency-based checklists for student assessment related to the CDIM curricular objectives using PDAs.(15) Data from these studies helped us determine that the distribution and use of curricular objectives and related assessment tools by students and faculty in our geographically dispersed medical school could be facilitated with just in time mobile technology. Importantly, we also determined that 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 PeerJ reviewing PDF | (v2013:05:520:1:0:NEW 17 Aug 2013) R ev ie w in g M an us cr ip t students and preceptors valued the fact that the content and expected competencies were transparent and such transparency facilitated learner assessment.(15) However, technical issues with PDAs -- such as lack of direct internet connection and the requirement to \u201csynchronize\u201d data from PDAs to the web using desktop computers -- limited the practicality of PDA based assessment; a process that is not needed with contemporary internet enabled devices such as iPads, iPhones and other smart phones. These devices have become almost ubiquitous in the past four years and we have leveraged this trend to evolve JIT to a platform-neutral Cloud-based system. The displayed assessment tools function like an \u201capplication\u201d on mobile devices, but are device-agnostic in that they functions on all internet-enabled devices, including smart phones. Out study, like most others, have several inherent limitations. First, this is a single institution study and these results may not be generalizable. Future studies should focus on the use of this technology in other settings. Second, establishing the reliability of all of the customized checklists within the CEX application is needed, as is establishing its reliability in real clinical settings such as the hospital wards. Third, we have not established the validity of the electronic grading algorithm. Fourth, like many tools for direct observation, we have not established the effect of this tool on learning nor the transfer of acquired clinical skills to other areas, or the effect that such direct observation has on the most important outcome of patient care. conclusions : We have established that just in time Cloud-based mobile technology has great potential in competency-based medical education. Although not an objective of this study, we believe such technology holds great promise for use in authentic clinical settings for measuring student achievement related to educational milestones. Additionally, given the time and cost constraints 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 PeerJ reviewing PDF | (v2013:05:520:1:0:NEW 17 Aug 2013) R ev ie w in g M an us cr ip t associated with traditional faculty development efforts, we believe that systems such as JIT have great potential in operationalizing \u201cjust in time\u201d faculty development. 202 203 PeerJ reviewing PDF | (v2013:05:520:1:0:NEW 17 Aug 2013) R ev ie w in g M an us cr ip t References 1. Whitcomb ME. More on competencybased education. Acad Med. 2004;79(6):4934. 2. Holmboe ES, Hawkins RE, Huot SJ. Effects of training in direct observation of medical residents' clinical competence: a randomized trial. Ann Intern Med. 2004;140(11):87481. 3. RL E. Estimation of the reliability of ratings. Psychometrika. 1951;16:40724. 4. Solomon DJ. The rating reliability calculator. BMC Med Res Methodol. 2004;4:11. 5. Kogan JR, Conforti L, Bernabeo E, Iobst W, Holmboe E. Opening the black box of clinical skills assessment via observation: a conceptual model. Med Educ. 2011;45(10):104860. 6. Hill F, Kendall K, Galbraith K, Crossley J. Implementing the undergraduate miniCEX: a tailored applicationroach at Southampton University. Med Educ. 2009;43(4):32634. 7. Fernando N, Cleland J, McKenzie H, Cassar K. Identifying the factors that determine feedback given to undergraduate medical students following formative miniCEX assessments. Med Educ. 2008;42(1):8995. 8. Hasnain M, Connell KJ, Downing SM, Olthoff A, Yudkowsky R. Toward meaningful evaluation of clinical competence: the role of direct observation in clerkship ratings. Acad Med. 2004;79(10 Suppl):S214. 9. Watling C, Driessen E, van der Vleuten CP, Lingard L. Learning from clinical work: the roles of learning cues and credibility judgements. Med Educ. 2012;46(2):192200. 10. Howley LD, Wilson WG. Direct observation of students during clerkship rotations: a multiyear descriptive study. Academic medicine : journal of the Association of American Medical Colleges. 2004;79(3):27680. 11. Torre DM, Simpson DE, Elnicki DM, Sebastian JL, Holmboe ES. Feasibility, reliability and user satisfaction with a PDAbased miniCEX to evaluate the clinical skills of third year medical students. Teaching and learning in medicine. 2007;19(3):2717. 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 PeerJ reviewing PDF | (v2013:05:520:1:0:NEW 17 Aug 2013) R ev ie w in g M an us cr ip t 12. Hauer KE, Kogan JR. Realising the potential value of feedback. Med Educ. 2012;46(2):1402. 13. Whitcomb ME. Competencybased graduate medical education? Of course! But how should competency be assessed? Acad Med. 2002;77(5):35960. 14. Ferenchick G, Fetters M, Carse AM. Just in time: technology to disseminate curriculum and manage educational requirements with mobile technology. Teach Learn Med. 2008;20(1):4452. 15. Ferenchick GS, Foreback J, Towfiq B, Kavanaugh K, Solomon D, Mohmand A. The implementation of a mobile problemspecific electronic CEX for assessing directly observed studentpatient encounters. Med Educ Online. 2010;15. 229 230 231 232 233 234 235 236 237 238 PeerJ reviewing PDF | (v2013:05:520:1:0:NEW 17 Aug 2013) R ev ie w in g M an us cr ip t Figure 1 Step 1. Content Manager for Development of Assessment Tools Using simple interfaces, faculty adds content (e.g. the problem altered mental status) and the specific competency to be assessed (e.g. history taking) PeerJ reviewing PDF | (v2013:05:520:1:0:NEW 17 Aug 2013) R ev ie w in g M an us cr ip t Figure 2 Step 2. Content Manager for Development of Assessment Tools Using the self-service web-based content management system, faculty then adds assessment questions reflecting specific criterion-based outcomes (e.g. The student started the interview with open-ended questions) PeerJ reviewing PDF | (v2013:05:520:1:0:NEW 17 Aug 2013) R ev ie w in g M an us cr ip t Figure 3 Criterion-based assessment for altered mental status and history-taking as displayed on the mobile device for use anytime and anywhere Screen shot A displays how the specific checklist is accessed on the device; screen shot B displays the criterion-based tasks, which are defaulted to No change to Yes (screen shot C) once the task is completed by the learner. Screen shot D displays the alogrithm generated grade PeerJ reviewing PDF | (v2013:05:520:1:0:NEW 17 Aug 2013) R ev ie w in g M an us cr ip t Figure 4 Evaluator information is collected using simple interfaces on the device after the assessment is completed, including open-ended qualitative comments. Faculty enters information concerning their observation (screen shot A), their feedback and action plans (screen shot B). A color coded competency registry is displayed on the learners device (screen shot C). Note in screen shot B, the evaluator has the option to have an email link sent to him/her to complete the qualitative assessment at a later time. All evaluations become part of the learners cloudbased permanent record. PeerJ reviewing PDF | (v2013:05:520:1:0:NEW 17 Aug 2013) R ev ie w in g M an us cr ip t Figure 5 detailed cloud-based reporting options : One of the web-based permanent records of the students\u2019 performance; displaying the item(s) assessed, the percentage of potential items correctly performed, and algorithm generated grade and evaluators written comments on the learners performance (note all of these features are editable, based upon the users\u2019 needs) PeerJ reviewing PDF | (v2013:05:520:1:0:NEW 17 Aug 2013) R ev ie w in g M an us cr ip t Figure 6 jit detailed cloud-based reporting options : With the click of a hyperlink, a detailed list of all the items that were either performed or not by the student PeerJ reviewing PDF | (v2013:05:520:1:0:NEW 17 Aug 2013) R ev ie w in g M an us cr ip t Figure 7 Another option for a cloud-based record or registry of the learners performance. This image represents a milestone based report with the identified milestones (A); the milestone subcompetencies (B); a color-coded table of all of the learners assessments (C). A roll-over option (D) identifies which specific assessment is represented in each cell. This table shows the ACGME competency taxonomy for internal medicine. PeerJ reviewing PDF | (v2013:05:520:1:0:NEW 17 Aug 2013) R ev ie w in g M an us cr ip t Figure 8 Bar chart of grade distribution comparing resident to faculty raters PeerJ reviewing PDF | (v2013:05:520:1:0:NEW 17 Aug 2013) R ev ie w in g M an us cr ip t",
+    "v2_text": "results : Number and types of evaluations Five hundred sixteen evaluators used the app to assess 367 students for a total of 3567 separate assessments. The number of CEX\u2019s completed per student was 9.7 (\u00b1 1.8) and the average number of CEX\u2019s completed per faculty was 6.9 (\u00b1 15.8). The average number of training problems a student was assessed on was 6.7; of the three competency domains of communication skills, history taking, and physical examination 68% of the students had at least one evaluation in each of the three categories. In terms of the grades, time variables and satisfaction, ~ 83% of the encounters were associated with a \u201cwell done\u201d grade, and on average students were credited with performing ~ 86% of the items correctly. Between 43 \u2013 50% of the CEX assessments took < 10 minutes as estimated by the faculty, and in ~ 69% of the encounters feedback was estimated to occur in less 5 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 PeerJ reviewing PDF | (v2013:05:520:0:1:NEW 26 May 2013) R ev ie w in g M an us cr ip t than 10 minutes. In 92% of the encounters, faculty rated that they were either satisfied or highly satisfied with the CEX The inter-rater reliability among faculty observing the videotaped encounters was 0.69 (slightly higher for faculty at 0.74 vs. residents at .64). In judging the exact same clinical performance, these ratings discriminated between satisfactory and non-satisfactory performance, as the mean number of items captured for the performance on the videotapes scripted for unsatisfactory performance was 35% vs. 59% for those scripted for more satisfactory performance. In terms of predictive validity, there was a small (but statistically significant, correlation 0.144, p = .008) correlation between students CEX scores and communications skills in the gateway performance assessment exam. discussion : Although national learning objectives have been published for all core clerkships, their usefulness for assessing learning outcomes has been limited. As an example, the core competency gathering essential and accurate information seems relatively straight forward. However, when applied to a single condition such as chronic obstructive pulmonary disease, there are at least 28 specified clinical tasks related to history taking and performing a physical examination that a student should demonstrate to meet the expected outcomes as defined in the Clerkship Directors in Internal Medicine (CDIM) curricular objectives for that problem. Of these 28, how many will a faculty evaluator remember when assessing the student? More importantly how many can they remember and what level of consistency will there be among preceptors providing feedback to students? 6 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 PeerJ reviewing PDF | (v2013:05:520:0:1:NEW 26 May 2013) R ev ie w in g M an us cr ip t If we take almost any clinical skill and start to dissect it, we find very quickly that existing human memory is insufficient in recalling all of the explicit steps related to potentially hundreds of conditions that help frame the expected outcomes of a trainee\u2019s educational experience and curricula. As the expectations for assessment of discrete competencies increases, the evaluation burden for educators, students and administrators becomes progressively more educationally incomplete and logistically unmanageable. The inability of faculty to remember and accurately assess for outcomes related to potentially hundreds of discrete educational objectives while evaluating trainees in clinical settings is one of the major reasons faculty have a hard time reliably discriminating unsatisfactory from satisfactory performance, as has been noted by many authors over the past decade.(2, 4) For example, in a study of the mini-CEX among 300 medical students, Hill noted that problems existed \u201cin trying to ensure that everyone was working to the same or similar standards.\u201d(5) In another study of 400 mini-CEX assessments, Fernando concluded faculty evaluators were unsure of the level of performance expected of the learners.(6) Hasnain noted that poor agreement among faculty evaluating medical students on a Family Medicine clerkship was due to the fact that \u201cStandards for judging clinical competence were not explicit\u201d.(7) In a randomized trial of a faculty development effort, Holmboe studied the accuracy of faculty ratings by having them view videotaped trainee-patient encounters that were scripted to portray three levels of proficiency; unsatisfactory, marginal or satisfactory. Faculty viewing the exact same encounter varied widely in their assessment of trainee competence, with ratings from unequivocally unsatisfactory (CEX scores of scores 1 \u2013 3) to unequivocally superior (CEX scores of 7 \u2013 9), regardless of whether the video was scripted to be unsatisfactory or not. After an intensive 4 day faculty development workshop in which participants were tasked with 7 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 PeerJ reviewing PDF | (v2013:05:520:0:1:NEW 26 May 2013) R ev ie w in g M an us cr ip t developing a shared mental model of what specific competencies should look like, problems still existed among faculty in discriminating satisfactory from unsatisfactory performance in these scripted encounters.(2) Kogan noted that in the absence of easily accessible frameworks, faculty evaluators default back to a myriad of highly variable evaluation strategies including such idiosyncratic features as instinct, \u201cgut feelings\u201d, \u201cunsubstantiated assumptions\u201d and the faculty members\u2019 emotional response to providing feedback. What she also noted was that faculty raters commonly fail to use existing frameworks or external standards in guiding their evaluations of trainees, thus explaining much of the well-recognized problems with poor validity and inter-rater reliability associated with clinical evaluations.(4) Given these realities, it is not surprising that medical trainees commonly do not view the feedback received from faculty as credible nor influential in learning, especially if the feedback was not immediate and tied to the trainees\u2019 clinical work-place performance. (8) Enhancing the effectiveness of clinical assessments, the delivery of feedback related to learning objectives and the creation of better systems for documenting faculty observations are commonly cited needs in medical education.(7, 9-12) Given these and other trends, systems that are capable of disseminating curricular objectives to students and faculty and which also enable criterion-based assessment have become key educational needs. We believe that Cloud-based technology, appropriately applied to maximize efficiency, can contribute to optimizing the learning environment by directly aligning learning objectives from national disciplinary curricula with assessment tools for use by students and faculty anywhere and anytime, especially at the bedside. 8 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 PeerJ reviewing PDF | (v2013:05:520:0:1:NEW 26 May 2013) R ev ie w in g M an us cr ip t In our first feasibility study, we demonstrated our ability to deliver national educational objectives published by the CDIM to electronic hand-held personal digital assistants (PDAs) such as Palm\u00ae and PocketPC\u00ae devices.(13) In a second feasibility study, we subsequently demonstrated this system could be used to deliver and successfully implement competency-based checklists for student assessment related to the CDIM curricular objectives using PDAs.(14) Data from these studies helped us determine that the distribution and use of curricular objectives and related assessment tools by students and faculty in our geographically dispersed medical school could be facilitated with just in time mobile technology. Importantly, we also determined that students and preceptors valued the fact that the content and expected competencies were transparent and such transparency facilitated learner assessment.(14) However, technical issues with PDAs -- such as lack of direct internet connection and the requirement to \u201csynchronize\u201d data from PDAs to the web using desktop computers -- limited the practicality of PDA based assessment; a process that is not needed with contemporary internet enabled devices such as iPads, iPhones and other smart phones. These devices have become almost ubiquitous in the past four years and we have leveraged this trend to evolve JIT to a platform-neutral Cloud-based system. The displayed assessment tools function like an \u201capp\u201d on mobile devices, but are device-agnostic in that they functions on all internet-enabled devices, including smart phones. Out study, like most others, have several inherent limitations. First, this is a single institution study and these results may not be generalizable. Future studies should focus on the use of this technology in other settings. Second, establishing the reliability of all of the customized checklists within the CEX app is needed, as is establishing its reliability in real clinical settings such as the hospital wards. Third, we have not established the validity of the electronic grading algorithm. Fourth, like many tools for direct observation, we have not 9 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 PeerJ reviewing PDF | (v2013:05:520:0:1:NEW 26 May 2013) R ev ie w in g M an us cr ip t established the effect of this tool on learning nor the transfer of acquired clinical skills to other areas, or the effect that such direct observation has on the most important outcome of patient care. a cloud-based record or registry of the learners performance is created. image below demonstrates : registry reporting organized using the current ACGME competency taxonomy, followed by specific tasks and milestones for an individual learner PeerJ reviewing PDF | (v2013:05:520:0:1:NEW 26 May 2013) R ev ie w in g M an us cr ip t Figure 5 drills down to more detail on the assessed item (next image) PeerJ reviewing PDF | (v2013:05:520:0:1:NEW 26 May 2013) R ev ie w in g M an us cr ip t Figure 6 conclusions : We have established that just in time Cloud-based mobile technology has great potential in competency-based medical education. Although not an objective of this study, we believe such technology holds great promise for use in authentic clinical settings for measuring student achievement related to educational milestones. Additionally, given the time and cost constraints associated with traditional faculty development efforts, we believe that systems such as JIT have great potential in operationalizing \u201cjust in time\u201d faculty development. 10 200 201 202 203 204 205 206 207 208 209 210 PeerJ reviewing PDF | (v2013:05:520:0:1:NEW 26 May 2013) R ev ie w in g M an us cr ip t References 1. Whitcomb ME. More on competencybased education. Acad Med. 2004;79(6):4934. 2. Holmboe ES, Hawkins RE, Huot SJ. Effects of training in direct observation of medical residents' clinical competence: a randomized trial. Ann Intern Med. 2004;140(11):87481. 3. Boyd AD, Hosner C, Hunscher DA, Athey BD, Clauw DJ, Green LA. An 'Honest Broker' mechanism to maintain privacy for patient care and academic medical research. Int J Med Inform. 2007;76(56):40711. 4. Kogan JR, Conforti L, Bernabeo E, Iobst W, Holmboe E. Opening the black box of clinical skills assessment via observation: a conceptual model. Med Educ. 2011;45(10):104860. 5. Hill F, Kendall K, Galbraith K, Crossley J. Implementing the undergraduate miniCEX: a tailored approach at Southampton University. Med Educ. 2009;43(4):32634. 6. Fernando N, Cleland J, McKenzie H, Cassar K. Identifying the factors that determine feedback given to undergraduate medical students following formative miniCEX assessments. Med Educ. 2008;42(1):8995. 7. Hasnain M, Connell KJ, Downing SM, Olthoff A, Yudkowsky R. Toward meaningful evaluation of clinical competence: the role of direct observation in clerkship ratings. Acad Med. 2004;79(10 Suppl):S214. 8. Watling C, Driessen E, van der Vleuten CP, Lingard L. Learning from clinical work: the roles of learning cues and credibility judgements. Med Educ. 2012;46(2):192200. 9. Howley LD, Wilson WG. Direct observation of students during clerkship rotations: a multiyear descriptive study. Academic medicine : journal of the Association of American Medical Colleges. 2004;79(3):27680. 10. Torre DM, Simpson DE, Elnicki DM, Sebastian JL, Holmboe ES. Feasibility, reliability and user satisfaction with a PDAbased miniCEX to evaluate the clinical skills of thirdyear medical students. Teaching and learning in medicine. 2007;19(3):2717. 11 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 PeerJ reviewing PDF | (v2013:05:520:0:1:NEW 26 May 2013) R ev ie w in g M an us cr ip t 11. Hauer KE, Kogan JR. Realising the potential value of feedback. Med Educ. 2012;46(2):1402. 12. Whitcomb ME. Competencybased graduate medical education? Of course! But how should competency be assessed? Acad Med. 2002;77(5):35960. 13. Ferenchick G, Fetters M, Carse AM. Just in time: technology to disseminate curriculum and manage educational requirements with mobile technology. Teach Learn Med. 2008;20(1):4452. 14. Ferenchick GS, Foreback J, Towfiq B, Kavanaugh K, Solomon D, Mohmand A. The implementation of a mobile problemspecific electronic CEX for assessing directly observed studentpatient encounters. Med Educ Online. 2010;15. 12 237 238 239 240 241 242 243 244 245 246 PeerJ reviewing PDF | (v2013:05:520:0:1:NEW 26 May 2013) R ev ie w in g M an us cr ip t Figure 1 jit content manager : Simple web based interfaces allow faculty of average computing skills to enter content for any type of assessment PeerJ reviewing PDF | (v2013:05:520:0:1:NEW 26 May 2013) R ev ie w in g M an us cr ip t Figure 2 Content as displayed on an internet enabled device (e.g. iPhone) for use anwhere anytime PeerJ reviewing PDF | (v2013:05:520:0:1:NEW 26 May 2013) R ev ie w in g M an us cr ip t Figure 3 Faculty enters information concerning their observation, their feedback and action plans A color coded competency registry is displayed on the learners device PeerJ reviewing PDF | (v2013:05:520:0:1:NEW 26 May 2013) R ev ie w in g M an us cr ip t Figure 4 detailed reporting options : One of the web-based permanent records of the students\u2019 performance, with the item assessed, the percentage of potential items correctly performed, and algorithm generated grade and written comments (note all of these features are editable, based upon the users\u2019 needs) PeerJ reviewing PDF | (v2013:05:520:0:1:NEW 26 May 2013) R ev ie w in g M an us cr ip t Figure 7 With the click of a hyperlink, a detailed list of all the items that were either performed or not by the student PeerJ reviewing PDF | (v2013:05:520:0:1:NEW 26 May 2013) R ev ie w in g M an us cr ip t",
+    "url": "https://peerj.com/articles/167/reviews/",
+    "review_1": "Gerard Lazo \u00b7 Aug 31, 2013 \u00b7 Academic Editor\nACCEPT\nThank you for your feedback on the suggested revisions. My read-through went very smoothly and I feel you have addressed all concerns addressed in the initial review. The manuscript was in good shape in the first iteration of review, and it is now even better. Having worked in the area of plant pathology I feel this work can have impact in serving the newest concerns of the science field, and especially to serve well with the latest advances in technology. I applaud your efforts and I expect the feedback to match accordingly when published. Congratulations.",
+    "review_2": "Gerard Lazo \u00b7 Aug 2, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nThe manuscript appears well written and is poised to provide Galaxy work-flows for bench scientists wishing to conduct transcript and peptide analyses for plant pathology related studies. Both reviewers felt the manuscript was appropriate for publication with what I consider to be minor modifications. The Galaxy environment has gained wide acceptance and I feel your application topic may help aid analyses in other plant pathology systems. This would potentially lead toward building common data connections between diverse host-pathogen interactions, and may even extend beyond the limited area of focus. Your presentation is centered around plant pathology based studies; however, it mainly focuses on the tools and not the research findings. Given that the introduction went well into describing the importance of plant pathology studies, perhaps a mention of some plant pathology revelations uncovered from your work may strengthen the impact of this effort. I will forward this to you with a suggestion of minor modifications. I would like you to try to address the points suggested in the reviews; it does not seem to be a major hurdle to accomplish this in a short period of time. Thank you for submitting this manuscript and I expect it to be well received. Congratulations on your efforts.\n\nOther comments which may be useful are to mention software alternatives to Galaxy and to mention to what extent the tools contained within the work-flows can also be used via the command-line. When mentioning third-party software it is best to note their availability; whether a license is required or not. Since the target audience appears the general bench scientist a description of the system requirements in terms of memory and processor requirements would be helpful. A sample data-set may also serve to let the target audience use and test for the expected outcomes.\n\nAdditional edits suggested :\nExample of annotation:\nLINE NO.: / PREVIOUS FORM / SUGGESTED FORM / [ADDITIONAL NOTES]\n\n68: / Apple\u2019s Mac OS X / the Apple OS X / [Mac is semi-redundant; see wikipedia]\n78: / to support / for support / []\n84: / offers is to offer / offers is / []\n114: / server can made / server can be made / []",
+    "review_3": "Mick Watson \u00b7 Jul 29, 2013\nBasic reporting\nThe paper meets the requirements\nExperimental design\nThe paper meets the requirements\nValidity of the findings\nThe paper meets the requirements\nAdditional comments\nThe authors describe a number of tools and tool wrappers that have been integrated into Galaxy, and provide a use-case in molecular plant pathology\n\nThere could be more mention of alternatives to Galaxy, e.g. Taverna and Anvaya\n\nWhilst MIRA has been integrated, no mention is used of the memory requirements - many are reluctant to integrate assemblers into their Galaxy instances for fear that several large memory jobs are launched by users\n\nOn page 5, two workflows are mentioned that are essentially identical, except one uses GetOrfs for gene finding and the second uses Augustus and Glimmer3. Doesn't the second workflow make the first redundant? Why include the first?\n\nOn page 6, technically I feel orthology should be the basis for transferring functional information, not sequence similarity. Similarly on page 7, isn't it more standard to use reciprocal best hit to define orthologues before transferring annotation?\n\nBottom of page 7, GetOrfs is used again - why not use the aforementioned gene predictors?\n\nWas any attempt made to wrap the InterProScan web-service (rather than standalone)?\n\nTop of page 8, I am curious whether the SignalP licence allows for it to be integrated into a public Galaxy?\n\nThe RXLR prediction tools: as I understand it, the authors have implemented several published methods for RXLR motif prediction, and released these into the Galaxy tool shed. Does this paper serve as notice of their publication? Has any testing been done on these implementations to demonstrate their accuracy and efficacy?\n\nOverall the paper is well written and should be published. The above suggestions can be dealt with by adding text to various parts of the manuscript and do not represent a large body of work, therefore I recommend minor revisions\n\nMick Watson\nCite this review as\nWatson M (2013) Peer Review #1 of \"Galaxy tools and workflows for sequence analysis with applications in molecular plant pathology (v0.1)\". PeerJ https://doi.org/10.7287/peerj.167v0.1/reviews/1",
+    "review_4": "Mikel Ega\u00f1a Aranguren \u00b7 Jul 26, 2013\nBasic reporting\nThe paper is very well written and presents the ideas clearly.\n\nSome minor (Discretionary) comments regarding the style:\n\n* The title is too long, how about \"A Galaxy framework for sequence analysis with applications in molecular plant pathology\"?.\n* In the abstract, NCBI BLAST+ is mentioned and then BLAST is mentioned again, but as an example. It id confusing.\n* In the abstract, in the sentence \"The motivating research theme ... \" it is not clear whether the research theme mentioned refers to Galaxy as a whole or the content of this paper. Also, the abstract reads like a presentation of Galaxy, rather than presenting the authors' work (Specific Galaxy tools).\n* The sentence in lines 148-151 is very difficult to understand.\n* The last part of the sentence in lines 244-245 may be clearer written as follows: \"despite being phylogenetically distant\"\n\nPossible mistakes:\n\nLine 112: computING cluster?\nLine 114: can BE made\nLine 115: extra space after \"e.g.\"? Perhaps the authors can use the LaTex command \\newcommand{\\eg}{\\emph{e.g.}\\xspace} (and the xspace package)\nLine 244: sequenceS\nExperimental design\nThe main objection is that the work presented in this paper is not completely reproducible.\n\nThe authors present a set of Galaxy tools and workflows that exploit such tools. However, only the \"backbones\" of the workflows are stored in the Galaxy tool shed. Therefore, if a user wants to reproduce the workflow, she needs to import it into a Galaxy server and run the workflow with datasets of her choice: since the datasets will be different, the workflows are not completely reproducible.\n\nThe authors should publish the workflows with the datasets they used to test them. Since the authors mention in the acknowledgements that they maintain an in-house Galaxy server, they can easily make the workflows mentioned in the paper public, and also publish a history with the datasets used, with clear instructions mapping the datasets to the corresponding workflow steps. This way any reader can run precisely the workflows presented in the paper, with the actual datasets, and judge the results. If the authors are worried about the computational burden for their server, they can set up accounts for the reviewers only, without making their Galaxy server public.\nValidity of the findings\nAs already mentioned, the datasets used to test the workflows have not been made available.\nCite this review as\nEga\u00f1a Aranguren M (2013) Peer Review #2 of \"Galaxy tools and workflows for sequence analysis with applications in molecular plant pathology (v0.1)\". PeerJ https://doi.org/10.7287/peerj.167v0.1/reviews/2",
+    "pdf_1": "https://peerj.com/articles/167v0.2/submission",
+    "pdf_2": "https://peerj.com/articles/167v0.1/submission",
+    "all_reviews": "Review 1: Gerard Lazo \u00b7 Aug 31, 2013 \u00b7 Academic Editor\nACCEPT\nThank you for your feedback on the suggested revisions. My read-through went very smoothly and I feel you have addressed all concerns addressed in the initial review. The manuscript was in good shape in the first iteration of review, and it is now even better. Having worked in the area of plant pathology I feel this work can have impact in serving the newest concerns of the science field, and especially to serve well with the latest advances in technology. I applaud your efforts and I expect the feedback to match accordingly when published. Congratulations.\nReview 2: Gerard Lazo \u00b7 Aug 2, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nThe manuscript appears well written and is poised to provide Galaxy work-flows for bench scientists wishing to conduct transcript and peptide analyses for plant pathology related studies. Both reviewers felt the manuscript was appropriate for publication with what I consider to be minor modifications. The Galaxy environment has gained wide acceptance and I feel your application topic may help aid analyses in other plant pathology systems. This would potentially lead toward building common data connections between diverse host-pathogen interactions, and may even extend beyond the limited area of focus. Your presentation is centered around plant pathology based studies; however, it mainly focuses on the tools and not the research findings. Given that the introduction went well into describing the importance of plant pathology studies, perhaps a mention of some plant pathology revelations uncovered from your work may strengthen the impact of this effort. I will forward this to you with a suggestion of minor modifications. I would like you to try to address the points suggested in the reviews; it does not seem to be a major hurdle to accomplish this in a short period of time. Thank you for submitting this manuscript and I expect it to be well received. Congratulations on your efforts.\n\nOther comments which may be useful are to mention software alternatives to Galaxy and to mention to what extent the tools contained within the work-flows can also be used via the command-line. When mentioning third-party software it is best to note their availability; whether a license is required or not. Since the target audience appears the general bench scientist a description of the system requirements in terms of memory and processor requirements would be helpful. A sample data-set may also serve to let the target audience use and test for the expected outcomes.\n\nAdditional edits suggested :\nExample of annotation:\nLINE NO.: / PREVIOUS FORM / SUGGESTED FORM / [ADDITIONAL NOTES]\n\n68: / Apple\u2019s Mac OS X / the Apple OS X / [Mac is semi-redundant; see wikipedia]\n78: / to support / for support / []\n84: / offers is to offer / offers is / []\n114: / server can made / server can be made / []\nReview 3: Mick Watson \u00b7 Jul 29, 2013\nBasic reporting\nThe paper meets the requirements\nExperimental design\nThe paper meets the requirements\nValidity of the findings\nThe paper meets the requirements\nAdditional comments\nThe authors describe a number of tools and tool wrappers that have been integrated into Galaxy, and provide a use-case in molecular plant pathology\n\nThere could be more mention of alternatives to Galaxy, e.g. Taverna and Anvaya\n\nWhilst MIRA has been integrated, no mention is used of the memory requirements - many are reluctant to integrate assemblers into their Galaxy instances for fear that several large memory jobs are launched by users\n\nOn page 5, two workflows are mentioned that are essentially identical, except one uses GetOrfs for gene finding and the second uses Augustus and Glimmer3. Doesn't the second workflow make the first redundant? Why include the first?\n\nOn page 6, technically I feel orthology should be the basis for transferring functional information, not sequence similarity. Similarly on page 7, isn't it more standard to use reciprocal best hit to define orthologues before transferring annotation?\n\nBottom of page 7, GetOrfs is used again - why not use the aforementioned gene predictors?\n\nWas any attempt made to wrap the InterProScan web-service (rather than standalone)?\n\nTop of page 8, I am curious whether the SignalP licence allows for it to be integrated into a public Galaxy?\n\nThe RXLR prediction tools: as I understand it, the authors have implemented several published methods for RXLR motif prediction, and released these into the Galaxy tool shed. Does this paper serve as notice of their publication? Has any testing been done on these implementations to demonstrate their accuracy and efficacy?\n\nOverall the paper is well written and should be published. The above suggestions can be dealt with by adding text to various parts of the manuscript and do not represent a large body of work, therefore I recommend minor revisions\n\nMick Watson\nCite this review as\nWatson M (2013) Peer Review #1 of \"Galaxy tools and workflows for sequence analysis with applications in molecular plant pathology (v0.1)\". PeerJ https://doi.org/10.7287/peerj.167v0.1/reviews/1\nReview 4: Mikel Ega\u00f1a Aranguren \u00b7 Jul 26, 2013\nBasic reporting\nThe paper is very well written and presents the ideas clearly.\n\nSome minor (Discretionary) comments regarding the style:\n\n* The title is too long, how about \"A Galaxy framework for sequence analysis with applications in molecular plant pathology\"?.\n* In the abstract, NCBI BLAST+ is mentioned and then BLAST is mentioned again, but as an example. It id confusing.\n* In the abstract, in the sentence \"The motivating research theme ... \" it is not clear whether the research theme mentioned refers to Galaxy as a whole or the content of this paper. Also, the abstract reads like a presentation of Galaxy, rather than presenting the authors' work (Specific Galaxy tools).\n* The sentence in lines 148-151 is very difficult to understand.\n* The last part of the sentence in lines 244-245 may be clearer written as follows: \"despite being phylogenetically distant\"\n\nPossible mistakes:\n\nLine 112: computING cluster?\nLine 114: can BE made\nLine 115: extra space after \"e.g.\"? Perhaps the authors can use the LaTex command \\newcommand{\\eg}{\\emph{e.g.}\\xspace} (and the xspace package)\nLine 244: sequenceS\nExperimental design\nThe main objection is that the work presented in this paper is not completely reproducible.\n\nThe authors present a set of Galaxy tools and workflows that exploit such tools. However, only the \"backbones\" of the workflows are stored in the Galaxy tool shed. Therefore, if a user wants to reproduce the workflow, she needs to import it into a Galaxy server and run the workflow with datasets of her choice: since the datasets will be different, the workflows are not completely reproducible.\n\nThe authors should publish the workflows with the datasets they used to test them. Since the authors mention in the acknowledgements that they maintain an in-house Galaxy server, they can easily make the workflows mentioned in the paper public, and also publish a history with the datasets used, with clear instructions mapping the datasets to the corresponding workflow steps. This way any reader can run precisely the workflows presented in the paper, with the actual datasets, and judge the results. If the authors are worried about the computational burden for their server, they can set up accounts for the reviewers only, without making their Galaxy server public.\nValidity of the findings\nAs already mentioned, the datasets used to test the workflows have not been made available.\nCite this review as\nEga\u00f1a Aranguren M (2013) Peer Review #2 of \"Galaxy tools and workflows for sequence analysis with applications in molecular plant pathology (v0.1)\". PeerJ https://doi.org/10.7287/peerj.167v0.1/reviews/2\nReview 5: nan\nReview 6: \n",
+    "LLM_extracted_review": "['Review 1: The manuscript was in good shape in the first iteration of review, and it is now even better. I feel this work can have impact in serving the newest concerns of the science field, especially to serve well with the latest advances in technology.\\n\\nReview 2: The manuscript appears well written and is poised to provide Galaxy work-flows for bench scientists wishing to conduct transcript and peptide analyses for plant pathology related studies. The presentation is centered around plant pathology based studies; however, it mainly focuses on the tools and not the research findings. A mention of some plant pathology revelations uncovered from your work may strengthen the impact of this effort. Other comments which may be useful are to mention software alternatives to Galaxy and to mention to what extent the tools contained within the work-flows can also be used via the command-line. A description of the system requirements in terms of memory and processor requirements would be helpful. A sample data-set may also serve to let the target audience use and test for the expected outcomes.\\n\\nReview 3: The authors describe a number of tools and tool wrappers that have been integrated into Galaxy, and provide a use-case in molecular plant pathology. There could be more mention of alternatives to Galaxy, e.g. Taverna and Anvaya. No mention is used of the memory requirements for MIRA, which may deter users from integrating assemblers into their Galaxy instances. Two workflows mentioned are essentially identical; one uses GetOrfs for gene finding and the second uses Augustus and Glimmer3, making the first redundant. It is more standard to use reciprocal best hit to define orthologues before transferring annotation. The paper should clarify whether the SignalP licence allows for integration into a public Galaxy. The RXLR prediction tools have been implemented, but it is unclear if this paper serves as notice of their publication or if testing has been done to demonstrate their accuracy and efficacy.\\n\\nReview 4: The title is too long; a suggestion is \"A Galaxy framework for sequence analysis with applications in molecular plant pathology.\" The abstract reads like a presentation of Galaxy rather than the authors\\' work. The main objection is that the work presented is not completely reproducible. The authors should publish the workflows with the datasets they used to test them. If the authors are worried about the computational burden for their server, they can set up accounts for the reviewers only, without making their Galaxy server public. The datasets used to test the workflows have not been made available.']"
+}

peerj_json_files/PeerJ_Json_121.json ADDED Viewed

	@@ -0,0 +1,13 @@

+{
+    "v1_Abstract": "The Galaxy Project offers the popular web browser-based platform Galaxy for running bioinformatics tools and constructing simple workflows. Here, we present a broad collection of additional Galaxy tools for large scale analysis of gene and protein sequences. The motivating research theme is the identification of specific genes of interest in a range of nonmodel organisms, and our central example is the identification and prediction of \"effector\" proteins produced by plant pathogens in order to manipulate their host plant. This functional annotation of a pathogen's predicted capacity for virulence is a key step in translating sequence data into potential applications in plant pathology.",
+    "v2_Abstract": "The Galaxy Project offers the popular web browser-based platform Galaxy for running bioinformatics tools and constructing simple workflows. Here, we present a broad collection of additional Galaxy tools for large scale analysis of gene and protein sequences. This collection includes novel tools, wrappers for widely-used third-party tools such as NCBI BLAST+, and workflows that combine several existing tools to enable more complex analyses. Individual bioinformatics software tools (e.g. BLAST) are typically available separately as standalone packages, or in online browser-based form. However, organism scale analyses often require automation, and a corresponding bioinformatic skill set that includes familiarity with command line tools and scripting. The Galaxy framework enables the user to combine these same tools without demanding the same level of computing skill. Workflows created using Galaxy can be saved and are reusable, so may be distributed within and between research groups, facilitating the construction of a set of standardised, reusable bioinformatic protocols. The motivating research theme is the identification of specific genes of interest in a range of non-model organisms, and our central example is the identification and prediction of \"effector\" proteins produced by plant pathogens in order to manipulate their host plant. This functional annotation of a pathogen's predicted capacity for virulence is a key step in translating sequence data into potential applications in plant pathology. The Galaxy tools and workflows described in this manuscript are open source and freely available from the Galaxy Tool Shed (http://usegalaxy.org/toolshed or http://toolshed.g2.bx.psu.edu).",
+    "url": "https://peerj.com/articles/169/reviews/",
+    "review_1": "Michael Somers \u00b7 Aug 31, 2013 \u00b7 Academic Editor\nACCEPT\nThank you for the rvision. The MSreads well.",
+    "review_2": "Michael Somers \u00b7 Aug 23, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nI have read the MS and find it well written and almost ready for acceptance. Please consider the minor issues as given by the referees.",
+    "review_3": "Reviewer 1 \u00b7 Aug 15, 2013\nBasic reporting\nNo comments\nExperimental design\nThis manuscript by Higley and Brosius is well written and presents a well- designed study of a rare tiger beetle that has important implications for its conservation. The introduction presents the relevant literature and an effective background to this study. The data presented and the statistical analysis of these behaviors seems appropriate in providing valid evidence of difference among species.\nValidity of the findings\nThe conclusion that these determined differences in behavior can reduce competition for food seems well supported and justified. The different behaviors of C. n. lincolniana provide a valid insight to its adaptation to the saline habitats where it is found, and can inform management by protecting habitat with seeps and other features it requires\nAdditional comments\nI would offer an alternate or complementary explanation for the foraging microhabitat of togata. It is significantly smaller than the other species, with long legs and consequently can both extend higher above the substrate where it is slightly cooler and would have reduced heat loading and quicker cooling. In tiger beetles (and other insects) smaller body size can result is more effective behavioral thermoregulation that species with larger body size, thus allowing for foraging on warmer substrates or during the warmer parts of the day (Pearson and Vogler).\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Behavioral niche partitioning in a sympatric tiger beetle assemblage and implications for the endangered Salt Creek tiger beetle (v0.1)\". PeerJ https://doi.org/10.7287/peerj.169v0.1/reviews/1",
+    "review_4": "Reviewer 2 \u00b7 Jul 26, 2013\nBasic reporting\nThe authors introduce a very complex set of interactions in an understandable context of ecology and conservation. The English is solid with a minimum of unnecessary jargon. A simple but significant change should be made in lines 59 and 61. The endangered tiger beetle is not an endangered species. It is an endangered population or recognized subspecies. There are a few typos throughout (e.g., line 339 \"on\" should be inserted between reliance and shallow).\nExperimental design\nNo comments\nValidity of the findings\nNo comments\nAdditional comments\nThe authors successfully integrate a complex set of interactions and test them convincingly to accomplish both goals of ecology and conservation.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"Behavioral niche partitioning in a sympatric tiger beetle assemblage and implications for the endangered Salt Creek tiger beetle (v0.1)\". PeerJ https://doi.org/10.7287/peerj.169v0.1/reviews/2",
+    "pdf_1": "https://peerj.com/articles/169v0.2/submission",
+    "pdf_2": "https://peerj.com/articles/169v0.1/submission",
+    "all_reviews": "Review 1: Michael Somers \u00b7 Aug 31, 2013 \u00b7 Academic Editor\nACCEPT\nThank you for the rvision. The MSreads well.\nReview 2: Michael Somers \u00b7 Aug 23, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nI have read the MS and find it well written and almost ready for acceptance. Please consider the minor issues as given by the referees.\nReview 3: Reviewer 1 \u00b7 Aug 15, 2013\nBasic reporting\nNo comments\nExperimental design\nThis manuscript by Higley and Brosius is well written and presents a well- designed study of a rare tiger beetle that has important implications for its conservation. The introduction presents the relevant literature and an effective background to this study. The data presented and the statistical analysis of these behaviors seems appropriate in providing valid evidence of difference among species.\nValidity of the findings\nThe conclusion that these determined differences in behavior can reduce competition for food seems well supported and justified. The different behaviors of C. n. lincolniana provide a valid insight to its adaptation to the saline habitats where it is found, and can inform management by protecting habitat with seeps and other features it requires\nAdditional comments\nI would offer an alternate or complementary explanation for the foraging microhabitat of togata. It is significantly smaller than the other species, with long legs and consequently can both extend higher above the substrate where it is slightly cooler and would have reduced heat loading and quicker cooling. In tiger beetles (and other insects) smaller body size can result is more effective behavioral thermoregulation that species with larger body size, thus allowing for foraging on warmer substrates or during the warmer parts of the day (Pearson and Vogler).\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Behavioral niche partitioning in a sympatric tiger beetle assemblage and implications for the endangered Salt Creek tiger beetle (v0.1)\". PeerJ https://doi.org/10.7287/peerj.169v0.1/reviews/1\nReview 4: Reviewer 2 \u00b7 Jul 26, 2013\nBasic reporting\nThe authors introduce a very complex set of interactions in an understandable context of ecology and conservation. The English is solid with a minimum of unnecessary jargon. A simple but significant change should be made in lines 59 and 61. The endangered tiger beetle is not an endangered species. It is an endangered population or recognized subspecies. There are a few typos throughout (e.g., line 339 \"on\" should be inserted between reliance and shallow).\nExperimental design\nNo comments\nValidity of the findings\nNo comments\nAdditional comments\nThe authors successfully integrate a complex set of interactions and test them convincingly to accomplish both goals of ecology and conservation.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"Behavioral niche partitioning in a sympatric tiger beetle assemblage and implications for the endangered Salt Creek tiger beetle (v0.1)\". PeerJ https://doi.org/10.7287/peerj.169v0.1/reviews/2\nReview 5: nan\nReview 6: \n",
+    "LLM_extracted_review": "['Review 1: The manuscript reads well.\\n\\nReview 2: The manuscript is well written and almost ready for acceptance, but minor issues need to be addressed.\\n\\nReview 3: The study presents a well-designed investigation of a rare tiger beetle with important conservation implications. The introduction effectively covers relevant literature, and the statistical analysis supports the findings. An alternate explanation for the foraging microhabitat of togata is suggested, highlighting the advantages of smaller body size in behavioral thermoregulation.\\n\\nReview 4: The authors present a complex set of ecological interactions in an understandable manner. The English is solid, but a significant change is needed regarding the classification of the endangered tiger beetle. There are a few typos that should be corrected. The authors successfully integrate and test complex interactions for ecological and conservation goals.']"
+}

peerj_json_files/PeerJ_Json_122.json ADDED Viewed

The diff for this file is too large to render. See raw diff

peerj_json_files/PeerJ_Json_123.json ADDED Viewed

The diff for this file is too large to render. See raw diff

peerj_json_files/PeerJ_Json_124.json ADDED Viewed

The diff for this file is too large to render. See raw diff

peerj_json_files/PeerJ_Json_125.json ADDED Viewed

The diff for this file is too large to render. See raw diff

peerj_json_files/PeerJ_Json_126.json ADDED Viewed

The diff for this file is too large to render. See raw diff

peerj_json_files/PeerJ_Json_127.json ADDED Viewed

The diff for this file is too large to render. See raw diff

peerj_json_files/PeerJ_Json_128.json ADDED Viewed

The diff for this file is too large to render. See raw diff

peerj_json_files/PeerJ_Json_129.json ADDED Viewed

The diff for this file is too large to render. See raw diff

peerj_json_files/PeerJ_Json_13.json ADDED Viewed

The diff for this file is too large to render. See raw diff

peerj_json_files/PeerJ_Json_130.json ADDED Viewed

The diff for this file is too large to render. See raw diff

peerj_json_files/PeerJ_Json_131.json ADDED Viewed

	@@ -0,0 +1,18 @@

+{
+    "v1_Abstract": "Small Ubiquitin Like Modifier (SUMO) proteins are involved in many processes in eukaryotes. We here show that Trypanosoma brucei SUMO (Tb927.5.3210) modifies many proteins. The levels of SUMOylation were unaffected by temperature changes but were increased by severe oxidative stress. We obtained evidence that trypanosome homologues of the SUMO conjugating enzyme Ubc9 (Tb927.2.2460) and the SUMO-specific protease SENP (Tb927.9.2220) are involved in SUMOylation and SUMO removal, respectively. Introduction Small Ubiquitin Like Modifier (SUMO) proteins have been found in almost all eukaryotes. Conjugation of SUMO to target proteins alters their functions in multiple ways, and it is therefore central to a multitude of different cellular processes Like ubiquitin, SUMO is attached to its targets via 3 enzymatic steps [1, 2]. First, a SUMOspecific protease (SENP) removes 2-11 amino acids at the SUMO C-terminus, revealing a Cterminal di-glycine motif [3]. Next, SUMO is activated by the SUMO activation complex (E1 complex), which consists of two enzymes, Aos1/SAE1 (budding yeast/human) and Uba2/SAE2 [4, 5]. The C-terminal glycine of SUMO forms a thioester bond with a cysteine residue of Uba2. From there, it is transferred to a cysteine residue of the E2 SUMO conjugating enzyme (Ubc9) [6, 1",
+    "v1_col_introduction": "introduction  : Small Ubiquitin Like Modifier (SUMO) proteins have been found in almost all eukaryotes. Conjugation of SUMO to target proteins alters their functions in multiple ways, and it is therefore central to a multitude of different cellular processes Like ubiquitin, SUMO is attached to its targets via 3 enzymatic steps [1, 2]. First, a SUMOspecific protease (SENP) removes 2-11 amino acids at the SUMO C-terminus, revealing a Cterminal di-glycine motif [3]. Next, SUMO is activated by the SUMO activation complex (E1 complex), which consists of two enzymes, Aos1/SAE1 (budding yeast/human) and Uba2/SAE2 [4, 5]. The C-terminal glycine of SUMO forms a thioester bond with a cysteine residue of Uba2. From there, it is transferred to a cysteine residue of the E2 SUMO conjugating enzyme (Ubc9) [6,\n11\n12\n13\n14\n15\n16\n17\n18\n19\n20\n21\n22\n23\n24\n25\n26\n27\nPeerJ reviewing PDF | (v2012:11:99:1:1:NEW 13 Sep 2013)\n7]. From the E2 conjugation enzyme, SUMO binds to a target lysine residue [1, 2]. This process is assisted by an E3 ligase. SUMO is usually attached as a monomer, although chain formation can occur [8]. SUMO is removed from its targets by a variety of peptidases called SENPs [9]. SENP regulation is critical for homeostasis [3, 10-12] and is also involved in responses to stresses such as heat shock and oxidation [10, 13]. SUMO is essential for growth in S. cerevisiae [4] but not in fission yeast [14] or Aspergillus [15]. Work on Chlamydomonas revealed that the abundance of SUMOylated proteins increases during heat shock and osmotic stress [16]. SUMO was also examined in Toxoplasma gondii [17] and Plasmodium falciparum [18]: in both cases many SUMOylated proteins were observed and identified by mass spectrometry, but details of the roles of SUMO in specific processes are not yet known. In the kinetoplastid Trypanosoma cruzi, the components of the SUMOylation machinery have been identified by BLAST search. Numerous SUMOylated bands were identified by Western blotting using both anti-SUMO antibody and detection of epitope-tagged SUMO. In addition, 236 potentially SUMOylated proteins were identified by tandem affinity purification and mass spectrometry[19], but unfortunately, a recent careful re-examination of the spectra could unambiguously identify only eight SUMOylated peptides on just seven proteins [20]. One of the targets identified by the tandem affinity purification, metacaspase 3, was confirmed by coimmunoprecipitation [19] although the SUMOylated peptide was not found [20]. The paraflagellar rod protein PFR1 (also called PAR3) was suggested as a SUMO target by Western blot analysis and in vitro SUMOylation [21] but again no SUMOylated peptide was found [20]. T. cruzi SUMO itself has a SUMOylation site and is able to polymerize [21]. Together these results suggest that many proteins are SUMOylated in T. cruzi, but purification is very difficult. Possibly, SUMO protease is very active and persists during purification procedures. The amino-acid sequence of Trypanosoma brucei SUMO (TbSUMO, Tb927.5.3210) is 37% identical with that of human SUMO-1 and the 3D structure (solved using NMR) is similar to those of yeast and mammalian SUMO [22]. It was shown by chemical shift analysis that TbSUMO interacts with human Ubc9. RNAi targeting SUMO in procyclic trypanosomes caused growth arrest and cell death, and HA-tagged SUMO was predominantly found in the T. brucei nucleus [23]. RNA intereference targeting SUMO in T. brucei caused growth arrest, followed by death, and antibodies to SUMO in bloodstream-form parasites recognised two prominent bands at around 55 and 60 kDa [24]. In this paper we describe preliminary functional characterisation of further components of the SUMOylation system in T. brucei and investigate the effects of various stresses on protein 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60\nPeerJ reviewing PDF | (v2012:11:99:1:1:NEW 13 Sep 2013)\nR ev ie w in g M an\nus cr ip t\nSUMOylation.",
+    "v2_Abstract": "Small Ubiquitin Like Modifier (SUMO) proteins are involved in many processes in eukaryotes. We here show that Trypanosoma brucei SUMO (Tb927.5.3210) modifies many proteins and is required for bloodstream trypanosome growth. The levels of SUMOylation were unaffected by temperature changes but were increased by severe oxidative stress. We obtained evidence that trypanosome homologues of the SUMO conjugating enzyme Ubc9 (Tb927.2.2460) and the SUMO-specific protease SENP (Tb927.9.2220) are involved in SUMOylation and SUMO removal, respectively.",
+    "v2_col_introduction": "introduction  : Small Ubiquitin Like Modifier (SUMO) proteins have been found in almost all eukaryotes. Conjugation of SUMO to target proteins alters their functions in multiple ways, and it is therefore central to a multitude of different cellular processes. Like ubiquitin, SUMO is attached to its targets via 3 enzymatic steps (Geiss-Friedlander and Melchior, 2007;Ulrich, 2009). First, a SUMO-specific protease (SENP) removes 2-11 amino acids at the SUMO C-terminus, revealing a C-terminal di-glycine motif (Kim and Baek, 2009). Next, SUMO is activated by the SUMO activating complex (E1 complex), which consists of two enzymes, Aos1/SAE1 (budding yeast/human) and Uba2/SAE2 (Desterro et al., 1999;Johnson et al., 1997). The C-terminal glycine of SUMO forms a thioester bond with a cysteine residue of Uba2. From there, it is transferred to a cysteine residue of the E2 SUMO conjugating enzyme (Ubc9) (Desterro et al., 1997;Johnson and Blobel, 1997). From the E2\nPre Prin ts Pre\nconjugating enzyme, SUMO is transferred to a target lysine residue (Geiss-Friedlander and Melchior, 2007;Ulrich, 2009). This process is assisted by an E3 ligase. SUMO is usually attached as a monomer, although chain formation can occur (Ulrich, 2008). SUMO is removed from its targets by a variety of peptidases called SENPs (Mukhopadhyay and Dasso, 2007). SENP regulation is critical for homeostasis (Au et al., 2009;Drag and Salvesen, 2008;Kim and Baek, 2009;Yeh, 2009) and is also involved in responses to stresses such as heat shock and oxidation (Au et al., 2009;Tempe et al., 2008). Most studies of SUMO so far have concentrated on multicellular eukaryotes and yeast. Outside these groups, work on Chlamydomonas revealed that the abundance of SUMOylated proteins increases during heat shock and osmotic stress (Wang et al., 2008). SUMO was also examined in Toxoplasma gondii (Braun et al., 2009) and Plasmodium falciparum (Issar et al., 2008): in both cases many SUMOylated proteins were observed and identified by mass spectrometry, but details of the roles of SUMO in specific processes are not yet known. The amino-acid sequence of Trypanosoma brucei SUMO (TbSUMO, Tb927.5.3210) is 37% identical with that of human SUMO-1 and the 3D structure (solved using NMR) is similar to those of yeast and mammalian SUMO. It was shown by chemical shift analysis that TbSUMO interacts with human Ubc9 (Shang et al., 2009). RNAi targeting SUMO in procyclic trypanosomes caused growth arrest and cell death, and HA-tagged SUMO was predominantly found in the T. brucei nucleus (Liao et al., 2010). Meanwhile, in the kinetoplastid Trypanosoma cruzi, possible SUMOylated proteins have been identified by tandem affinity purification and mass spectrometry (Bayona et al., 2011). SUMOylation of one of the identified targets, metacaspase 3, was confirmed by immunoprecipitation (Bayona et al., 2011). PFR1 (also called PAR3) was also identified as a SUMO target by Western blot analysis and in vitro SUMOylation (Annoura et al., 2012). T. cruzi SUMO itself has a SUMOylation site and is able to polymerize in vitro (Annoura et al., 2012). In this paper we describe preliminary functional characterisation of components of the SUMOylation system in T. brucei and investigate the effects of various stresses on protein SUMOylation.",
+    "v1_text": "results and discussion : Many proteins are SUMOylated in T. brucei To detect SUMOylated proteins, an antibody was raised to His-tagged TbSUMO produced in E. coli. (For details of all plasmid constructs see Table 1.) The anti-T.brucei SUMO antibody was unfortunately insufficiently specific. Although it recognised purified recombinant SUMO, it detected several bands, but not monomeric SUMO, in bloodstream- and procyclic-form cell extracts and the banding pattern was not affected by SUMO RNAi. Next, in bloodstream-form trypanosomes, we integrated a sequence encoding a tandem affinity purification tag (TAP-tag) N-terminally in frame with one allele of SUMO. The TAP-SUMO was seen as a ~40kDa band (Figure 1A); the expected size was 33.5 kDa, comprising 12.5 kDa SUMO + 21 kDa tag. Many SUMOylated proteins were present, with a prominent band at 100kDa, which, without the tag, would correpond to an 80kDa SUMOylated conjugate. In T. cruzi also, using either HA-tagged or untagged SUMO, many SUMOylated bands were seen in addition to monomeric SUMO [31]. The pattern that we observed, however, bears no resemblance to the two prominent bands at 55 and 60 kDa that were previously described for bloodstream-form T. brucei using an anti-SUMO antibody [24]. There are several possible explanations for the discrepancy. Our TAP-SUMO may have impaired function due to the tag, or the 100kDa band could actually be equivalent to the 55KDa band (but with very aberrant migration). Alternatively the previously-published pattern could have been affected by antibody cross-reactions with abundant proteins. Since, in the published results, the antibody staining was abolished after 72h RNAi, the last possibility seems unlikely. It cannot, however, be ruled out since no evidence was presented that the antibody recognised native monomeric SUMO on the Western blots. Also the RNAi had depleted the SUMO RNA within 24h but the signal on the blots was unchanged after 48h RNAi despite extensive morphological degeneration [24]. Our pattern of TAP-SUMO-conjugated proteins was unaffected by heat shock (Figure 1A) or by treatment for 12h or 24h with a sub-lethal level (100 \u00b5M) of phloretin [32] to partially inhibit glucose import (not shown). The bloodstream forms used for these experiments are not able to differentiate into growing procyclic forms, but can undergo some early steps of differentiation after addition of cis-aconitate and transfer to 27\u00b0C. When we did this in the TAP-SUMOexpressing line, the banding pattern remained largely unchanged, but one band migrating at 90 kDa reproducibly disappeared (Figure 1B, marked with a star). In accordance with this result, a changing SUMOylation pattern during differentiation was found in T. cruzi [21]. Our 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 PeerJ reviewing PDF | (v2012:11:99:1:1:NEW 13 Sep 2013) R ev ie w in g M an us cr ip t experiments only detected the most abundant SUMOylated proteins and it is quite possible that less abundant proteins show regulated SUMOylation. In procyclic forms (Figure 1C), we integrated a sequence encoding a V5 epitope tag upstream of the open reading frame [26]. We expected monomeric V5-SUMO to migrate at 13 kDa. This was not reproducibly seen, but we did sometimes see a band or bands running at 20 kDa (Figure 1C). In contrast, slower mobility bands were always present, in particular a prominent double band just below 100 kDa. Comparison of the patterns from bloodstream and procyclic forms (by manipulating the photographs to allow for the sizes of the tags, not shown) suggested that the patterns of abundant SUMOylated proteins were similar in both forms. The SUMOylation pattern in procyclics was unaffected by temperature changes (1h incubations, Figure 1C lanes 7-10). SUMOylation increases after oxidative stress In mammalian cells, peroxide concentrations of 1mM and lower inhibit SUMOylation [33] through formation of a disulfide bond between the catalytic domains of the E2 enzyme Ubc9 and the E1 complex subunit Uba2. In contrast, in trypanosomes, oxidative stress increased the abundance of SUMOylated protein, even at relatively low peroxide levels (32 \u00b5M, 1h incubation; Figure 1C). We do not know the reason for this discrepancy: the trypanosome E1 and E2 enzymes may differ such that the dimerization cannot occur, or the dimerization in mammalian cells may be caused by a specific regulatory process that is absent in trypanosomes. Peroxide concentrations above 1 mM in Saccharomyces cerevisiae [34], or 10mM in mammalian cells [33, 35], increase SUMOylation, probably by inhibiting the SENP proteases [33, 36]. Trypanosomes probably react at lower peroxide concentrations because they are much more susceptible to oxidative stress than mammalian cells and yeast: the EC50 of hydrogen peroxide for bloodstream T. brucei is 223\u00b5M [37], and we found that procyclic trypanosomes were killed by concentrations above 250\u00b5M. Failure to purify SUMOylated proteins from T. brucei extracts We made multiple attempts to purify the SUMOylated proteins from trypanosome extracts, using TAP-, His- and V5 tags and a variety of protocols. As previously reported for T. cruzi, all of these attempts failed [21]. First, we attempted tandem affinity purification. SUMOylation was stable for 2h at 4\u00b0C in the lysis buffer, in which standard protease inhibitor mix and N-ethyl-maleimide (20 mM, NEM) were included in order to inhibit SUMO proteases. Although NEM was removed before the TEV 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 PeerJ reviewing PDF | (v2012:11:99:1:1:NEW 13 Sep 2013) R ev ie w in g M an us cr ip t protease cleavage step, we were unable to elute the tagged preteins from the column. A one-step immunoprecipitation, using V5-tagged SUMO, also yielded no specific protein pattern because only 5% of the V5-tagged SUMO bound to the beads. As SUMO binds covalently to its targets, it might be better to use His-tagged SUMO so that SUMOylated proteins can be purified under denaturating conditions. Role of SUMOylation in T. brucei Reciprocal BLASTp searches using yeast and human sequences, and comparison with T.cruzi [19] revealed several putative homologues of Uba2 and Aos1, the enzymes forming the E1 complex, and also of the E2 enzyme Ubc9 (Table 2). Since SUMO E1 and E2 enzymes resemble those for ubiquitination, the specificities of these proteins is unclear. There were four possible E3 ligases, consistent with the need to regulate SUMOylation of different targets separately. However, only one SENP was found. This is surprising, given that the function of SENPs include both the processing and the removal of SUMO, but the same was reported for T.cruzi [19]. As previously reported [24], RNAi targeting SUMO in bloodstream trypanosomes halted growth 2 days after RNAi induction (Figure 2A). We too observed numerous defects in cell division, which is normal in growth-arrested trypanosomes and does not by itself constitute evidence of a role of SUMO in regulating the cell cycle. RNAi in procyclic forms expressing V5-SUMO gave only a transient decrease in V5-SUMO (on day 2 after induction) although the RNA was clearly decreased; in two independent clones, the doubling time increased from 12.6 h to 14.5 h and 15.1 h (not shown). Liao et al [23] observed stronger growth inhibtion. We first targeted the putative SUMO protease, SENP (Tb09.160.0970/ Tb927.9.2220). RNAi had hardly any effect on cell growth (doubling time increase of only 0.3 h, not shown), but there was a strong increase in the abundance of SUMO modification (Figure 2B, lanes 1-4), confirming that the Tb927.9.2220 protein is important for SUMO removal in trypanosomes. Given this increase in SUMOylation, we speculate that a different enzyme might be involved in the activation of SUMO prior to transfer to the E1 conjugating enzyme. Alternatively, much lower levels of SENP activity may be needed for initial SUMO processing than for SUMO removal. Next, we targeted the possible E2 conjugating enzyme UBC9 (Tb927.2.2460) in procyclic trypanosomes expressing V5-SUMO. A UBC9 RNAi line grew slower than the parent line, even in the absence of tetracycline, and there was only marginal slowing after tetracycline addition (not shown). We did not check the mRNA levels: the mRNA is present at less than one copy per cell so it would be difficult to detect even before RNAi [38]. However, there was a reproducible 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 PeerJ reviewing PDF | (v2012:11:99:1:1:NEW 13 Sep 2013) R ev ie w in g M an us cr ip t decrease in SUMOylation (Figure 2B, lanes 5-9) after RNAi. figure legends : Figure 1 Protein modification by tagged SUMO A. Effect of temperature on the pattern of modification with TAP-SUMO in bloodstream trypanosomes containing the plasmid pHD2020. Lane 1: cells without TAP-SUMO. Lanes 2,3,4: The cells were incubated for 1 h at the indicated temperatures. The antibody used for detection was PAP: peroxidase anti-peroxidase antibody (binds to the IgG-binding domain of the TAP tag). B. Effect of differentiation conditions on the pattern of TAP-SUMOylated proteins. Bloodstream trypanosomes were isolated at 6 x 105 cells/ml (L, lower density, lane 1) or 2.5 x 106 cells/ml (H, 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 PeerJ reviewing PDF | (v2012:11:99:1:1:NEW 13 Sep 2013) R ev ie w in g M an us cr ip t higher density, lane 2). Cis-aconitate was added to the higher-density cells and the culture incubated at 37\u00b0C for 17h (lanes 3-6) [39]. Then, the culture was centrifuged and resuspended in procyclic-form medium at 27\u00b0C (lanes 7 & 8). C. Effect of oxidative stress and temperature stress on the pattern of V5-SUMO modification in procyclic trypanosomes. Parasites were transfected with pHD2021 to V5-in situ tag SUMO at the N-terminus [26]. Lanes 1 and 7 are controls. Lanes 2-6: Cells with V5-in situ-tagged SUMO were incubated for 1 h with 15.6, 31.2, 62.5, 125 or 250 \u00b5M hydrogen peroxide. Lanes 8-10: incubation for 1 h at the indicated temperatures. Proteins were detected with anti-V5; as a control, a monoclonal antibody to tubulin (TUB) (from K. Gull) was used. Figure 2 A. Effect of RNAi targeting SUMO on growth of bloodstream-form trypanosomes. RNAi was induced by addition of tetracycline and growth followed daily, with dilution as required to keep the cell density below 1 x 106/ml. B. The effect of RNAi targeting TbUBC9 and TbSENP on SUMOylation in procyclic trypanosomes. Trypanosomes expressing T7 polymerase and the tet repressor [27] were transfected with pHD 2021 and pHD2038 or pHD2037. RNA interference was induced with tetracycline (100 ng/ml in the absence of other selective drugs) for the times shown and the patterns of SUMOylation assayed by Western blotting. 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 PeerJ reviewing PDF | (v2012:11:99:1:1:NEW 13 Sep 2013) R ev ie w in g M an us cr ip t Figure 1 Figure 1 Protein modification by tagged SUMO A. Effect of temperature on the pattern of modification with TAP-SUMO in bloodstream trypanosomes containing the plasmid pHD2020. Lane 1: cells without TAP-SUMO. Lanes 2,3,4: The cells were incubated for 1 h at the indicated temperatures. The antibody used for detection was PAP: peroxidase anti-peroxidase antibody (binds to the IgG-binding domain of the TAP tag). B. Effect of differentiation conditions on the pattern of TAP-SUMOylated proteins. Bloodstream trypanosomes were isolated at 6 x 105 cells/ml (L, lower density, lane 1) or 2.5 x 106 cells/ml (H, higher density, lane 2). Cis-aconitate was added to the higher-density cells and the culture incubated at 37\u00b0C for 17h (lanes 3-6) [ 39 ] . Then, the culture was centrifuged and resuspended in procyclic- form medium at 27\u00b0C (lanes 7 & 8). C. Effect of oxidative stress and temperature stress on the pattern of V5-SUMO modification in procyclic trypanosomes. Parasites were transfected with pHD2021 to V5-in situ tag SUMO at the N-terminus [ 26 ] . Lanes 1 and 7 are controls. Lanes 2-6: Cells with V5-in situ-tagged SUMO were incubated for 1 h with 15.6, 31.2, 62.5, acknowledgements : We thank Keith Gull (University of Oxford) for the anti-tubulin antibody, J. Haanstra and B. Bakker (Utrecht) for communicating phloretin results and Frauke Melchior (ZMBH) for advice. DD was supported by Sonderforschungsbereich 544 of the Deutsche Forschungsgemeinschaft. methods : conclusions : We confirmed the functions of the trypanosome SENP and UBC9 genes, and could show that SUMO modifies many trypanosome proteins. The pattern of SUMOylation was surprisingly unresponsive to stress and also appeared not to be strongly developmentally regulated. table legends : Table 1 Plasmids used in this work (not all results described in text) Table 2 Trypanosome genes potentially involved in SUMOylation. Genes were identified by reciprocal BLASTp. Only genes giving a yeast SUMO pathway enzyme as the best match are included. The putative PIAS homologues each have the expected RING domain and the single SENP has a cysteine protease domain. plasmids : For the N-terminal in situ TAP tag, a part of the TbSUMO open reading frame (ORF) was amplified using the following primers fw: 5'-gac aag ctt ccg cca ccg acg aac cca ctc ata ac-3' rv: 5'-gtc gat atc tca tgt ctg ctc cac cat cgc-3' and cloned into the p2676 vector [25] using Hind III and EcoR V. For the N-terminal V5 in situ tag, a part of the TbSUMO ORF was amplified (fw: 5' - gac ctc gag gac gaa ccc act cat-3', rv; 5' - gac ggg ccc tca cgc cat gca cca-3'), as well as a part of the 5\u2019 untranslated region (UTR) (fw: 5'- gac ccg cgg tgt cct tgt ggt tac gt-3' rv: 5'-gac tct aga aag agg aag tcg ggg ag-3'). The ORF and UTR fragments were cloned into a vector containing the V5-tag and the Blasticidin resistance as described in [26] using Apa I and Xho I for the ORF and Sac II and Xba I for the UTR fragment. For the RNAi constructs, portions of the open reading frames of the targeted genes were amplified and cloned into p2T7TA blue[27]. The following primers were used: for TbSUMO fw: 5'-ggg ggt acc gac gaa ccc act cat aac-3' rv: 5'-ccc aag ctt tca cgc cat gca cca aag 3'; for Tb09.160.0970 (TbSENP) fw: 5'-cag acg act cac tat cgc ca-3', rv:5'-tgc gct caa atg ttg ttc tc-3' and for Tb927.2.2460 (TbUBC9) fw: 5'-tag ctc agt cac gcc tac ga-3'rv:5'-aca cac gaa atg gct ctt cc-3'. The primers were designed using RNAit [28]. trypanosome culture : Trypanosoma brucei strain Lister 427 expressing the tet repressor, with or without T7 polymerase, were used throughout [27], with culturing and transfecting of trypanosomes as previously described [29]. For growth studies, bloodstream-form cells were diluted to a starting concentration of 5x104 cells/ml, with a maximum density of 1.5-2x106 cells/ml. Procyclics were diluted to 5x105, with a maximum density of 5x106. Tetracycline was added to a final concentration of 0.25\u00b5g/ml to induce expression from tetracycline-regulated promoters. For differentiation, bloodstream-form cells were grown to 1.5-2x106 cells/ml, then cis-aconitate was added to a concentration of 6mM. The cells were grown for 16h at 37\u00b0C then transferred to 27\u00b0C. Inhibition of glucose transport was achieved by adding phloretin to a concentration of 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 PeerJ reviewing PDF | (v2012:11:99:1:1:NEW 13 Sep 2013) R ev ie w in g M an us cr ip t 100\u00b5M. For oxidative stress, H2O2 was added to procyclic trypanosomes to final concentrations of 250\u00b5M, 125\u00b5M, 62,5\u00b5M, 31,25\u00b5M or 15,6\u00b5M; the cells were harvested after an incubation time of 1h. tandem affinity purification : For each tandem affinity purification 4-5x109 cells were harvested at 4\u00b0C and washed twice with ice-cold PBS containing 20mM N-ethyl maleimide (NEM). Bloodstream cells were harvested at a density of 2x106 cells/ml, procyclic cells at a density of 5 x106 cells/ml. The cell pellets were snap-frozen in liquid nitrogen and stored at -80\u00b0C. Cell breakage was performed in 6ml breakage buffer (10mM Tris-Cl, 10mM NaCl, 0.1% NP40, 20mM NEM, one tablet of complete inhibitor (without EDTA, Roche) pH=7.8) by passing the cells 20-25 times through a 21 gauge needle. The lysate was spun at 13,000g for 30 min at 4\u00b0C to remove the cell debris. Then NaCl was added to a final concentration of 150mM. The purification was done according to [30]. 20mM NEM was added to all the buffers, except during the wash and elution step of the IgG beads and during the TEV cleavage, as NEM inhibits TEV protease. v5 immunoprecipitation : 2.5-5x109 procyclic cells were harvested at a density of approximately 5 x106 cells/ml, washed twice with ice-cold PBS containing 10mM NEM and 10mM iodoacetamide (IAA) and snapfrozen in liquid nitrogen and stored at -80\u00b0C. For use, the cell pellet was resuspended in 1ml lysis buffer (10mM Tris-C, 10mM NaCl, 0.1% NP40, 1% SDS, complete protease inhibitor (Roche), 10mM NEM, 10mM IAA, pH=7.5). Cells were passed the cells 20-25 times through a 21 gauge needle. The lysate was spun at 13,000g for 30 min at 4\u00b0C to remove the cell debris and was diluted 1:10 in IP100 (10mM Tris-Cl, 100mM NaCl, 0.1% NP40, complete protease inhibitor (Roche), 10mM NEM, 10mM IAA). Before immunoprecipitation, the lysate was incubated with protein A sepharose for 1 hour on a rotary shaker at 4\u00b0C to absorb non-specifically binding proteins. The supernatant was than added to 50\u00b5l \u03b1-V5 beads (self-made or from Sigma). The lysate was incubated for 3h on a rotary shaker at 4\u00b0C. The beads were washed seven times with IP 100. Elution was done by incubating the beads twice with 125\u00b5l IP buffer mixed with 25\u00b5l V5 peptide for 30min, then boiling in 4x Laemmli buffer. 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 PeerJ reviewing PDF | (v2012:11:99:1:1:NEW 13 Sep 2013) R ev ie w in g M an us cr ip t 125 or 250 \u00b5m hydrogen peroxide. lanes 8-10: incubation for 1 h at the indicated : temperatures. Proteins were detected with anti-V5; as a control, a monoclonal antibody to tubulin (TUB) (from K. Gull) was used. PeerJ reviewing PDF | (v2012:11:99:1:1:NEW 13 Sep 2013) R ev ie w in g M an us cr ip t Figure 2 Figure 2 A. Effect of RNAi targeting SUMO on growth of bloodstream-form trypanosomes. RNAi was induced by addition of tetracycline and growth followed daily, with dilution as required to keep the cell density below 1 x 106/ml. B. The effect of RNAi targeting TbUBC9 and TbSENP on SUMOylation in procyclic trypanosomes. Trypanosomes expressing T7 polymerase and the tet repressor [ 27 ] were transfected with pHD 2021 and pHD2038 or pHD2037. RNA interference was induced with tetracycline (100 ng/ml in the absence of other selective drugs) for the times shown and the patterns of SUMOylation assayed by Western blotting. PeerJ reviewing PDF | (v2012:11:99:1:1:NEW 13 Sep 2013) R ev ie w in g M an us cr ip t Table 1(on next page) Table 1 Plasmids used in this work (not all results described in text). PeerJ reviewing PDF | (v2012:11:99:1:1:NEW 13 Sep 2013) R ev ie w in g M an us cr ip t Table 1 pl as mi d Descri ption cloning strategy pH D2 02 0 SUMO /TAP in-situ tag (Ntermina l) A part of the SUMO ORF was amplified using the following primers fw: 5'-gac aag ctt ccg cca ccg acg aac cca ctc ata ac-3' rv: 5'-gtc gat atc tca tgt ctg ctc cac cat cgc-3' and cloned into the p2676 [25] using Hind III and Eco RV pH D2 02 1 SUMO /V5 insitu tag (N- termina l) The SUMO ORF was amplified using the following primers: fw: 5' - gac ctc gag gac gaa ccc act cat-3' rv; 5' - gac ggg ccc tca cgc cat gca cca-3'. A fragment of the UTR was amplified using the following primers: fw: 5'- gac ccg cgg tgt cct tgt ggt tac gt-3' rv: 5'-gac tct aga aag agg aag tcg ggg ag-3'. The ORF was cloned into the Bla V5 vector [26] using Apa I and Xho I, the 5\u2019 UTR using Sac II and Xba I. pH D2 02 2 SUMO knockout A fragment of the SUMO 3' UTR was amplified using the following primers: fw: 5'- gac tct aga cat aag tgc gcg tag tgg-3' rv: 5'- gtc ccg cgg gca aac gac cgc aga agt-3'. A 5'-UTR fragment was amplified using the following primers: fw: 5'-cac tcg agc cct cat atc cac atc ctc a-3' rv: 5'- gtc aag ctt cgt ggg ctc aga aat gaa-3'. 1 PeerJ reviewing PDF | (v2012:11:99:1:1:NEW 13 Sep 2013) R ev ie w in g M an us cr ip t The 3'-UTR fragment was cloned into pHD1748 (Blasticidin resistance cassette in polylinker) using Xba I and Sac II, the 5' UTR fragment using Xho I and Hind III. pH D2 02 3 SUMO knockout The Blasticidin resistence was removed from pHD2022 using Hind III and Eco RI and replaced by a Puromycin resistence which was acquired by digesting pHD1747 with the same enzymes. pH D2 02 4 SUMO RNAi The SUMO ORF was amplified using the following primers: fw: 5'-ggg ggt acc gac gaa ccc act cat aac-3' rv: 5'-ccc aag ctt tca cgc cat gca cca aag -3' and cloned into p2T7TAblue [27] pH D2 02 5 HisSUMO The SUMO ORF was amplified using the following primers: fw: 5'- gag ggt acc gac gaa ccc act cat aac-3' rv: 5'-ccc aag ctt tca cgc cat gca cca aag-3' and cloned into pQEA38 using Kpn I and Hind III. pQEA38 is an expression vector with ten His tags and a TEV cleavage site, modified from pQTEV (AY243506), from the lab of D. G\u00f6rlich (then at ZMBH). pH D2 02 6 SUMO /His insitu tag (N- termina l) A 5' UTR fragment was amplified using the following primers: fw: 5'- gac ccg cgg tgt cct tgt ggt tac gt-3' rv: 5'- gac tct aga aag agg aag tcg ggg ag-3' and cloned into the Bla V5 vector using Sac II and Xba I. Then the V5 tag was removed using Eco NI and Xho I. The SUMO ORF and the His-tag sequence were cut out of pHD 2025 using Eco RI and Hind III. The vector and the tagged ORF were blunted using the Klenow fragment and ligated. pH D2 03 7 TbSEN P RNAi A fragment of the SENP ORF was amplified using the following primers: fw: 5'-cag acg act cac tat cgc ca-3' rv:5'-tgc gct caa atg ttg ttc tc-3' and cloned into the p2T7TAblue vector pH TbUB A fragment of the UBC9 ORF was amplified using the following PeerJ reviewing PDF | (v2012:11:99:1:1:NEW 13 Sep 2013) R ev ie w in g M an us cr ip t D2 03 8 C9 RNAi primers fw: 5'-tag ctc agt cac gcc tac ga-3' rv:5'-aca cac gaa atg gct ctt cc-3' and cloned into the p2T7TAblue vector PeerJ reviewing PDF | (v2012:11:99:1:1:NEW 13 Sep 2013) R ev ie w in g M an us cr ip t Table 2(on next page) Table 2 Trypanosome genes potentially involved in SUMOylation. Genes were identified by reciprocal BLASTp. Only genes giving a yeast SUMO pathway enzyme as the best match are included. The putative PIAS homologues each have the expected RING domain and the single SENP has a cysteine protease domain. PeerJ reviewing PDF | (v2012:11:99:1:1:NEW 13 Sep 2013) R ev ie w in g M an us cr ip t Table 2 Enzyme Function Tb homologue Name Aos1 E1 complex Tb11.02.5410 AOS1 Uba2 E1 complex Tb927.5.3430 UBA2 Ubc9 E2 complex Tb927.2.2460 UBC9 PIAS4/Siz1 E3 ligase Tb09.211.2400 PIAS4/Siz1 E3 ligase Tb927.2.4420 PIAS4/Nfi1 E3 ligase Tb11.01.8710 PIAS1/Siz1 E3 ligase Tb927.6.4830 SUMO1/Ulp2 SENP Tb09.160.0970 SENP 1 PeerJ reviewing PDF | (v2012:11:99:1:1:NEW 13 Sep 2013) R ev ie w in g M an us cr ip t",
+    "v2_text": "results and discussion : Many proteins are SUMOylated in T. brucei To detect SUMOylated proteins, an antibody was raised to His-tagged TbSUMO produced in E. coli. The anti-T.brucei SUMO antibody was unfortunately insufficiently specific. Although it recognised purified recombinant SUMO, it detected several bands, but not monomeric SUMO, in bloodstream- and procyclic-form cell extracts and the banding pattern was not affected by SUMO RNAi (not shown). Next, in bloodstream-form trypanosomes, we integrated a sequence encoding a tandem affinity purification tag (TAP-tag) N-terminally in frame with one allele of SUMO. The TAP-SUMO was seen as a ~40kDa band (Figure 1A); the expected size was 33.5 kDa, comprising 12.5 kDa SUMO + 21 kDa tag. In addition, many SUMOylated proteins were present, with a prominent band above 100kDa. The pattern was unaffected by heat shock (Figure 1A) or by treatment for 12h or 24h with a sub-lethal level (100 \u00b5M) of phloretin (Haanstra et al., 2011) to partially inhibit Pre Prin ts Pre Prin ts glucose import (not shown). The bloodstream forms used for these experiments are not able to differentiate into growing procyclic forms, but can undergo some early steps of differentiation after addition of cis-aconitate and transfer to 27\u00b0C. When we did this in the TAP-SUMO-expressing line, the banding pattern remained largely unchanged, but one band migrating at 90 kDa reproducibly disappeared (Figure 1B). In accordance with this result, a changing SUMOylation pattern during differentiation was found in T. cruzi (Annoura et al., 2012). Our experiments only detected the most abundant SUMOylated proteins and it is quite possible that many more, less abundant, proteins show regulated SUMOylation. In procyclic forms (Figure 1C), we integrated a sequence encoding a V5 epitope tag upstream of the open reading frame (Shen et al., 2001). We expected monomeric V5-SUMO to migrate at 13 kDa. This was not reproducibly seen, but we did sometimes see a band or bands running at 20 kDa (Figure 1C). In contrast, slower mobility bands were always present, in particular a prominent double band just below 100 kDa. Comparison of the patterns from bloodstream and procyclic forms (by manipulating the photographs to allow for the sizes of the tags, not shown) suggested that the patterns of abundant SUMOylated proteins were similar in both forms. The SUMOylation pattern in procyclics was unaffected by temperature changes (1h incubations, Figure 1C lanes 7-10). SUMOylation increases after oxidative stress In mammalian cells, peroxide concentrations of 1mM and lower inhibit SUMOylation (Bossis and Melchior, 2006) through formation of a disulfide bond between the catalytic domains of the E2 enzyme Ubc9 and the E1 complex subunit Uba2. In contrast, in trypanosomes, oxidative stress increased the abundance of SUMOylated protein, even at relatively low peroxide levels (15,6\u00b5M or 31,25\u00b5M, 1h incubation; Figure 1C, lane 2 and 3). We do not know the reason for this discrepancy: the trypanosome E1 and E2 enzymes may differ such that the dimerization cannot occur, or the dimerization in mammalian cells may be caused by a specific regulatory process that is absent in trypanosomes. Peroxide concentrations above 1 mM in Saccharomyces cerevisiae (Zhou et al., 2004), or 10mM in mammalian cells (Bossis and Melchior, 2006;Saitoh and Hinchey, 2000), increase SUMOylation, probably by inhibiting the SENP proteases (Bossis and Melchior, 2006;Xu et al., 2008). Trypanosomes probably react at lower peroxide concentrations because they are much more susceptible to oxidative stress than mammalian cells and yeast: the EC50 of hydrogen peroxide for bloodstream T. brucei is 223\u00b5M (Krieger et al., 2000), and Pre Prin ts Pre Prin ts we found that procyclic trypanosomes were killed by concentrations above 250\u00b5M. Failure to purify SUMOylated proteins from T. brucei extracts We made multiple attempts to purify the SUMOylated proteins from trypanosome extracts, using TAP- and V5-tags and a variety of protocols. As previously reported for T. cruzi, all of these attempts failed (Annoura et al., 2012). First, we attempted tandem affinity purification, as previously successfully applied to the exosome (Estevez et al., 2001). SUMOylation was stable for 2h at 4\u00b0C in TAP lysis buffer, as observed by Western blot (not shown). Both protease inhibitor and N-ethyl-maleimide (NEM, 20 mM) were included in the lysis buffer in order to inhibit SUMO proteases. However, NEM was removed by washing before the TEV cleavage step, as it inhibits the TEV protease. Eluates from cells expressing TAP-tagged SUMO gave exactly the same pattern on SDS-PAGE as eluates from cells expressing TAP alone and SUMO was not visible. Further analysis indicated that the majority of TAP-tagged SUMO had bound to the IgG beads, but it was never recovered. Additional bands were however not seen after analysis of the boiled beads. A one-step immunoprecipitation, using V5 tagged SUMO, also yielded no specific protein pattern. Western blot analysis, however, showed that only about 5% of the V5-tagged SUMO bound to the beads, suggesting that the problem lay with poor binding of the tagged protein to the beads. One reason for the failure of the purification of SUMO might be that we failed to inhibit SUMO proteases. As SUMO binds covalently to its targets, it should be possible to purify SUMOylated proteins under denaturating conditions, with all proteases completely inactive during the purification.Therefore, expression of His-tagged SUMO in trypanosomes, would allow purification on nickel agarose beads, in the presence of urea. Bayona et al. (Bayona et al., 2011) have, in contrast to our experience and that of Annoura et al., reported purification of SUMOylated T. cruzi proteins. They started with ten times as many parasites, included TLCK, iodoacetamide, and PMSF as protease inhibitors, and used a tandem affinity approach with His and HA tags, without TEV protease cleavage and with a control extract lacking tagged protein. The list of putatively SUMOylated proteins included most highly abundant proteins, so - although the authors attempted quantitation by comparing spectral counts, and did the experiments in triplicate - it is unclear how many of the identified proteins really are SUMOylated. Nevertheless, they could confirm that a very small proportion of metacaspase 3 is indeed SUMOylated. Pre Prin ts Pre Prin ts Role of SUMOylation in T. brucei To identify enzymes involved in SUMOylation, we performed reciprocal BLASTp searches using yeast and human sequences, as well as comparing the results to the homologues found in T.cruzi (Bayona et al., 2011). We found several putative homologues of Uba2 and Aos1, the enzymes forming the E1 complex, and also for the E2 enzyme Ubc9 (Table 2). Since SUMO E1 and E2 enzymes resemble those for ubiquitination, the specificities of these proteins is unclear. There were four possible E3 ligases, consistent with the need to regulate SUMOylation of different targets separately. However, only one SENP was found. This is surprising, given that the function of SENPs include both the processing and the removal of SUMO. The same was reported for T.cruzi (Bayona et al., 2011). RNAi targeting SUMO in bloodstream trypanosomes halted growth 2 days after RNAi induction (Figure 2A). The RNAi was confirmed by Northern blot (not shown). We observed numerous defects in cell division, which is normal in growth-arrested trypanosomes and does not constitute evidence of a role of SUMO in regulating the cell cycle. RNAi in procyclic forms expressing V5-SUMO gave only a transient decrease in V5-SUMO (on day 2 after induction); in two independent clones, the doubling time increased from 12.6 h to 14.5 h and 15.1 h (not shown). Liao et al (Liao et al., 2010), in contrast, observed stronger growth inhibition. Thus in trypanosomes, as in S. cerevisiae (Johnson et al., 1997), SUMO is essential for growth. In contrast, SUMO is not essential in fission yeast (Tanaka et al., 1999) or Aspergillus (Szewczyk et al., 2008). Next, we targeted the possible E2 conjugating enzyme UBC9 (Tb927.2.2460) in procyclic trypanosomes expressing V5-SUMO A UBC9 RNAi line grew slower than the parent line, even in the absence of tetracycline (not shown). There was only marginal slowing after tetracycline addition and the decrease in SUMOylation was reproducible but moderate (Figure 2B, lanes 5-9), suggesting that substantial levels of UBC9 activity remained. Finally, we targeted the putative SUMO protease, SENP (Tb09.160.0970/ Tb927.9.2220). RNAi had hardly any effect on cell growth (doubling time increase of only 0.3 h, not shown), but there was reproducibly a clear increase in the abundance of SUMO modification (Figure 2B, lanes 1-4), confirming that the identified gene indeed is important for SUMO removal in trypanosomes. It is paradoxical that only one SUMO protease was identified, since the activity is needed for both activation and removal of SUMO. Since RNAi targeting SENP resulted in a general increase in SUMOylation, we speculate that a different enzyme might be involved in the activation of SUMO prior to transfer to the E1 conjugating enzyme. Alternatively, Pre Prin ts Pre Prin ts much lower levels of SENP activity may be needed for the initial SUMO processing than for SUMO removal. figure legends : Figure 1 Protein modification by tagged SUMO A. Effect of temperature on the pattern of modification with TAP-SUMO in bloodstream trypanosomes containing the plasmid pHD2020. Lane 1: cells without TAP-SUMO. Lanes 2,3,4: The cells were incubated for 1 h at the indicated temperatures. The antibody used for detection was PAP: peroxidase anti-peroxidase antibody (binds to the IgG-binding domain of the TAP tag). B. Effect of differentiation conditions on the pattern of TAP-SUMOylated proteins. Bloodstream trypanosomes were isolated at 6 x 105 cells/ml (L, lower density, lane 1) or 2.5 x 106 cells/ml (H, higher density, lane 2). Cis-aconitate was added to the higher-density cells and the culture incubated at 37\u00b0C for 17h (lanes 3-6) (Queiroz et al., 2009). Then, the culture was centrifuged and resuspended in procyclic-form Pre Prin ts Pre Prin ts medium at 27\u00b0C (lanes 7 & 8). The asterisk marks the lane disappearing during the treatment. The arrow marks the lane corresponding to the unblound SUMO protein. C. Effect of oxidative stress and temperature stress on the pattern of V5-SUMO modification in procyclic trypanosomes. Parasites were transfected with pHD2021 to V5-in situ tag SUMO at the N-terminus (Shen et al., 2001). Lanes 1 and 7 are controls. Lanes 2-6: Cells with V5-in situ-tagged SUMO were incubated for 1 h with 15.6, 31.2, 62.5, 125 or 250 \u00b5M hydrogen peroxide. Lanes 8-10: incubation for 1 h at the indicated temperatures. Proteins were detected with anti-V5 (AbD Serotech); as a control, a monoclonal antibody to tubulin (KMX1) (from K. Gull) was used. Figure 2 Roles of SUMO, UBC9 and SENP A. Effect of RNAi targeting SUMO on growth of bloodstream-form trypanosomes. RNAi was induced by addition of tetracycline and growth followed daily, with dilution as required to keep the cell density below 1 x 106/ml. B. The effect of RNAi targeting TbUBC9 and TbSENP on SUMOylation in procyclic trypanosomes. Trypanosomes expressing T7 polymerase and the tet repressor (Alibu et al., 2004) were transfected with pHD 2021 and pHD2038 or pHD2037. RNA interference was induced with tetracycline (100 ng/ml in the absence of other selective drugs) for the times shown and the patterns of SUMOylation assayed by Western blotting. Pre Prin ts Pre Prin ts Figure 1 Protein modification by tagged SUMO Pre Prin ts Pre Prin ts Figure 2 Roles of SUMO, UBC9 and SENP Pre Prin ts Pre Prin ts acknowledgements : We thank Keith Gull (University of Oxford) for the anti-tubulin antibody J. Haanstra and B. Bakker (Utrecht) for communicating phloretin results and Frauke Melchior (ZMBH) for advice. DD was supported by Sonderforschungsbereich 544 of the Deutsche Forschungsgemeinschaft. methods : conclusions : We confirmed the functions of the trypanosome SENP and UBC9 genes, and could show that SUMO modifies many trypanosome proteins. The pattern of SUMOylation was surprisingly unresponsive to stress and also appeared not to be strongly developmentally regulated. tables : Table 1 Trypanosome genes potentially involved in SUMOylation. Genes were identified by reciprocal BLASTp. Only genes giving a yeast SUMO pathway enzyme as the best match are included. The putative PIAS homologues each have the expected RING domain and the single SENP has a cysteine protease domain. Enzyme Function Tb homologue Name Aos1 E1 complex Tb11.02.5410 AOS1 Uba2 E1 complex Tb927.5.3430 UBA2 Ubc9 E2 complex Tb927.2.2460 UBC9 PIAS4/Siz1 E3 ligase Tb09.211.2400 PIAS4/Siz1 E3 ligase Tb927.2.4420 PIAS4/Nfi1 E3 ligase Tb11.01.8710 PIAS1/Siz1 E3 ligase Tb927.6.4830 SUMO1/Ulp2 SENP Tb09.160.0970 SENP plasmids : For the N-terminal in situ TAP tag, a part of the TbSUMO open reading frame (ORF) was amplified using the following primers fw: 5'-gac aag ctt ccg cca ccg acg aac cca Pre Prin ts Pre Prin ts ctc ata ac-3' rv: 5'-gtc gat atc tca tgt ctg ctc cac cat cgc-3' and cloned into the p2676 vector (Kelly et al., 2007) using Hind III and EcoR V. For the N-terminal V5 in situ tag, a part of the TbSUMO ORF was amplified (fw: 5' - gac ctc gag gac gaa ccc act cat-3', rv: 5' - gac ggg ccc tca cgc cat gca cca-3'), as well as a part of the 5\u2019 untranslated region (UTR) (fw: 5'- gac ccg cgg tgt cct tgt ggt tac gt-3' rv: 5'-gac tct aga aag agg aag tcg ggg ag-3'). The ORF and UTR fragments were cloned into a vector containing the V5-tag and the Blasticidin resistance as described in (Shen et al., 2001) using Apa I and Xho I for the ORF and Sac II and Xba I for the UTR fragment. For the expression of recombinant TbSUMO in E.coli, the ORF was amplified with the following primers: fw: 5'- gag ggt acc gac gaa ccc act cat aac-3' rv: 5'-ccc aag ctt tca cgc cat gca cca aag-3' and cloned into pQEA38 using Kpn I and Hind III. pQEA38 is an expression vector with ten His tags and a TEV cleavage site, modified from pQTEV (AY243506), from the lab of D. G\u00f6rlich (then at ZMBH). Protein expression was done using BL21 cells. Protein purification was done under native conditions, using Ni-NTA agarose beads (Qiagen) according to the Qiagen manual. For the RNAi constructs, portions of the open reading frames of the targeted genes were amplified and cloned into p2T7TA blue(Alibu et al., 2004). The following primers were used: for TbSUMO fw: 5'-ggg ggt acc gac gaa ccc act cat aac-3' rv: 5'-ccc aag ctt tca cgc cat gca cca aag 3'; for Tb09.160.0970 (TbSENP) fw: 5'-cag acg act cac tat cgc ca-3', rv:5'-tgc gct caa atg ttg ttc tc-3' and for Tb927.2.2460 (TbUBC9) fw: 5'-tag ctc agt cac gcc tac ga-3'rv:5'-aca cac gaa atg gct ctt cc-3'. The primers were designed using RNAit (Redmond et al., 2003). trypanosome culture : Trypanosoma brucei strain Lister 427 expressing the tet repressor, with or without T7 polymerase, were used throughout (Alibu et al., 2004), with culturing and transfecting of trypanosomes as previously described (van Deursen et al., 2001). For growth studies, bloodstream-form cells were diluted to a starting concentration of 5x104 cells/ml, with a maximum density of 1.5-2x106 cells/ml. Procyclics were diluted to 5x105, with a maximum density of 5x106. Tetracycline was added to a final concentration of 0.25\u00b5g/ml to induce expression from tetracycline-regulated promoters. For differentiation, bloodstream-form cells were grown to 1.5-2x106 cells/ml, then cis-aconitate was added to a concentration of 6mM. The cells were grown for 16h at 37\u00b0C then transferred to 27\u00b0C. Inhibition of glucose transport was achieved by adding phloretin to a concentration of 100\u00b5M. For oxidative stress, H2O2 was added Pre Prin ts Pre Prin ts to procyclic trypanosomes to final concentrations of 250\u00b5M, 125\u00b5M, 62,5\u00b5M, 31,25\u00b5M or 15,6\u00b5M; the cells were harvested after an incubation time of 1h. tandem affinity purification : For each tandem affinity purification approximately 5x109 cells were harvested at 4\u00b0C and washed twice with ice-cold PBS containing 20mM N-ethyl maleimide (NEM). Bloodstream cells were harvested at a density of 2x106 cells/ml, procyclic cells at a density of 5 x106 cells/ml. The cell pellets were snap-frozen in liquid nitrogen and stored at -80\u00b0C. Cell breakage was performed in 6ml breakage buffer (10mM Tris-Cl, 10mM NaCl, 0,1% IGEPAL, 20mM NEM, one tablet of complete protease inhibitor (without EDTA, Roche) pH=7.8) by passing the cells 20-25 times through a 21 gauge needle. The lysate was spun at 16,200g for 30 min at 4\u00b0C to remove the cell debris. Then NaCl was added to a final concentration of 150mM. The purification was done according to (Puig et al., 2001). 20mM NEM was added to all the buffers, except during the wash and elution step of the IgG beads and during the TEV cleavage, as NEM inhibits TEV protease. v5 immunoprecipitation : Anti-V5 antibodies (AbD serotech) were diluted in 1ml 1xPBS and incubated with protein A sepharose (1mg antibody per mg beads) for 1h at room temperature. Then the bound antibody was coupled to the column according to (Harlow and Lane, 1999). The beads were washed five times with PBS before use. 2.5-5x109 procyclic cells a density of approximately 5 x106cells/ml were harvested, washed twice with ice-cold PBS containing 10mM NEM and 10mM iodoacetamide (IAA) and snap-frozen in liquid nitrogen and stored at -80\u00b0C. For use, the cell pellet was resuspended in 1ml lysis buffer (10mM Tris-C, 10mM NaCl, 0,1% IGEPAL, 1% SDS, complete protease inhibitor (Roche), 10mM NEM, 10mM IAA, pH=7.5). Cells were passed the cells 20-25 times through a 21 gauge needle. The lysate was spun at 16,200g for 30 min at 4\u00b0C to remove the cell debris and was diluted 1:10 in IP100 (10mM Tris-Cl, 100mM NaCl, 0,1% IGEPAL, complete protease inhibitor (Roche), 10mM NEM, 10mM IAA). Before immunoprecipitation, the lysate was incubated with protein A sepharose for 1 hour on a rotary shaker at 4\u00b0C to absorb non-specifically binding proteins. The supernatant was than added to 50\u00b5l \u03b1-V5 beads (self-made or from Sigma). The lysate was incubated for 3h on a rotary shaker at 4\u00b0C. The beads were washed seven times with IP100. Elution was done by incubating the beads twice with 125\u00b5l Pre Prin ts Pre Prin ts IP buffer mixed with 25\u00b5l V5 peptide (2mg/ml) for 30min, then were boiled in 4x Laemmli buffer. Eluates were TCA-precipitated and loaded on a 15% SDS gel, which was stained with silver stain. Purification of metabolically labelled proteins A small scale purification of metabolic labelled proteins was done to test the efficiency of the V5 pull-down. About 6x107 procyclic cells were spun down at 600g, transferred to a 1.5ml Eppendorf tube and washed once with PBS. The cells were resuspended in 400ml labelling medium, as described in (Clayton, 1987). Then 12\u00b5l 35S methionine solution (120\u00b5Ci) was added. The cells were incubated for 1h at 30\u00b0C. Then the cells were pelleted and washed twice with PBS containing 10mM NEM and 10mM IAA. The purification was done according to the already described protocols, with slight modifications. 5\u00b5l \u03b1-V5-tagged beads were used. Cell lysis was done by keeping the cells on ice for 45min, with occasional vortexing. The samples were incubated on a rotary shaker rotated at 4\u00b0C for 3,5h. The beads were spun down and washed five times with 1ml IP100. The supernatant of the V5-tagged cells was given on 5\u00b5l new \u03b1-V5-tagged beads and rotated at 4\u00b0C for additional 1.5h. The proteins were eluted with 5\u00b5l V5 peptide (2mg/ml). To yield additional proteins, the beads were boiled in 2x Laemmli buffer.",
+    "url": "https://peerj.com/articles/181/reviews/",
+    "review_1": "Alla Kostyukova \u00b7 Sep 26, 2013 \u00b7 Academic Editor\nACCEPT\nFont size in some parts of the manuscript is smaller than it should be. If it is still a result of transferring the doc file to the pdf file, please, inform the production staff about this problem.",
+    "review_2": "Alla Kostyukova \u00b7 Sep 9, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nYou don't need to do any additional experiments but you should revise your discussion and address other reviewer's comments.",
+    "review_3": "Reviewer 1 \u00b7 Sep 8, 2013\nBasic reporting\nThe article provides sufficient introduction and background to support the experimental design of the study. Formatting of Figures appears appropriate and the article seems to conform with template information.\n\nMinor comments:\nIn the abstract, it should be clarified that the donor (AEDANS) is bound to the tropomyosin isoforms.\nIn the first paragraph of the introduction (line 18), S1 should be introduced as being the S1 subfragment of myosin which may not be clear to an audience not familiar with the field of research.\nTypo in line 75: Change \u2018Than cDNA from the first\u2026, to \u2018Then, cDNA from the first\u2026\u2019\nTypo in line 186: Change \u2018the donor-acceptor distance the absence\u2019 to \u2018the donor-acceptor distance in the absence\u2019\nReference \u2018Barua et al., 2013 (lines 281-283) needs volume and page information\nExperimental design\nThe research question is clearly defined and directly links this study with the findings of the authors from a previous study. The approach and experimental design has been validated by the previous study. The research question is a logical consequence from their previous findings. The method are described with sufficient detail.\nValidity of the findings\nIn the present study, Slivinska and Moraczewska determined isoform-specific localisation of tropomyosin C-termini along actin filaments and muscle myosin S1 subfragment-dependent changes in the localisation of the C-termini. This work is a direct follow-up study where the authors have tested the properties of the N-termini of the same tropomyosin isoforms. The current study reveals interesting information on the displacement of the C-termini of low- and high molecular weight tropomyosins in the presence of myosin S1 binding. These findings provide significant experimental evidence for an isoform specific mechanism by which different tropomyosins regulate the access of actin-associated proteins to the filaments.\nThe recognition of tropomyosins as \u2018gate keepers\u2019 of actin filaments is an emerging concept in the field of studying the regulation of cellular architecture. A large diversity in the regulation of actin-associated proteins by tropomyosins has particularly been explored previously for non-muscle tropomyosins. In this context it would be interesting to test differences in the S1 subfragment-dependent C-terminus displacement of tropomyosins for which isoform-specific regulation of the interaction between actin-associated proteins and the filament has been shown in different cell types. This includes for example the different properties of the TPM1 gene product TmBr3 (containing exons 1b and 9c) and the TPM3 gene product Tm5NM1 (containing exon 1b and 9d) (Bryce et al., 2003). Similarly, it would be interesting to test the effect of non-muscle myosin S1 (e.g. non-muscle myosin IIb S1) on displacement of the tropomyosin C-Termini due to the different binding properties of the muscle and non-muscle S1.\nAdditional comments\nno comments\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Structural differences between C-terminal regions of tropomyosin isoforms (v0.1)\". PeerJ https://doi.org/10.7287/peerj.181v0.1/reviews/1",
+    "review_4": "Dmitrii Levitsky \u00b7 Sep 5, 2013\nBasic reporting\nThe manuscript 647 by \u015aliwi\u0144ska and Moraczewska describes the influence of the C-terminal sequence of tropomyosin (C-terminal exons 9a or 9d) on the distances between this part of Tm bound to F-actin and Cys374 of actin. For this purpose, four Tm isoforms with different C-terminal exons, 9a or 9d, were used, and the distances between fluorescence donor attached to Cys residue introduced into this Tm exon (Cys 269 or Cys 232) and acceptor attached to Cys374 of actin were measured by FRET in two different states of Tm on the surface of actin filament (closed and open).\n\nThis paper continues the previous paper of the authors (Sliwinska et al. (2011) Cytoskeleton 68, 300-312 ). Both papers are very similar (and sometimes even identical) in respect to Tm isoforms and the methods used, with the only difference that in the previous paper the fluorescence donor was attached to Cys residue in the N-terminal part of Tm, whereas in the present work it was introduced into the Tm C-terminal part. Taken together, the results of both studies show that the distances between Tm end-to-end overlap region and actin are different for Tm isoforms, and their changes in response to myosin S1 binding are individual for each isoform. Generally, I agree with authors\u2019 conclusion that the Tm isoforms are differently oriented on actin filament and the degree of Tm shift in response to the filament activation by myosin S1 binding (i.e. transition from closed to open state) is individual for each type of the studied isoform.\nExperimental design\nThe FRET experiments are carefully performed and clearly described, and the results are well explained.\nValidity of the findings\nThe work is mostly novel, although the main aspects of the research have been previously published by the authors (see above).\nAdditional comments\nn my opinion, the paper could be quite suitable for publication in PeerJ after some corrections according to the following comments.\n\n1). First line in the abstract: \u201cTropomyosins are actin regulatory proteins...\u201d\nIt seems to me, \u201cTropomyosins are actin-binding regulatory proteins..\u201d is more correct.\n\n2). Results, line 102: \u201c...illustration of the overlap regions are shown in Fig. 1.\u201d\nIt should be either \u201d...illustration of the overlap regions is shown in Fig. 1\u201d or \u201c...illustrations of the overlap regions are shown in Fig. 1.\u201d\n\n3). Table 1, upper line (exon 9a, red): it should be A/C, but not A/.\n\n4). The English needs to be edited carefully, and the formatting of the text should be improved (e.g., on lines 87-89, 185, 193-197, and 258-263 the size of the letters is too small).\nCite this review as\nLevitsky D (2013) Peer Review #2 of \"Structural differences between C-terminal regions of tropomyosin isoforms (v0.1)\". PeerJ https://doi.org/10.7287/peerj.181v0.1/reviews/2",
+    "pdf_1": "https://peerj.com/articles/181v0.2/submission",
+    "pdf_2": "https://peerj.com/articles/181v0.1/submission",
+    "review_5": "Reviewer 3 \u00b7 Sep 1, 2013\nBasic reporting\nIn this paper the authors have used FRET to study the relationship of the C terminus of tropomyosin to Cys 374 on actin in Tm isoforms that would have different overlap complexes, and the cooperativity of myosin binding. The rationale is sound: there are structures of the overlap regions of tropomyosins, and models for actin-Tm based on Tm models and structures that do not contain the overlap region. Since the ends are known to be so important for Tm function and a source of isoform diversity (work from Dr. Morazcewsaka\u2019s lab and others), this is a reasonable approach and significant goal. Cysteine mutations were introduced just outside of the overlap region in four different Tms and shown to have little effect on the function of the Tm. The relative affinities are consistent with her previous work (Moraczewska et al., 1999) but not cited, and the fluorescence quenching is consistent with the affinities. The appropriate controls were carried out.\nThe FRET distances were determined for four Tms with sequences encoded by four different combinations of N-terminal and C-terminal sequences encoded by alternatively-expressed exons. Three occur naturally, the fourth is a synthetic construct (1b9a). While Tm2, Tm5a and 1b9a are all alpha Tm (TPM1) sequences, that used for the skTm is not clear: unacetylated alpha TM binds poorly; the citation is in a paper that studied gamma skTm (TPM3) that is 285 residues long due to an extra Met at the N terminus. This needs to be clarified, and also any differences from the common exons in the alpha-Tm encoded forms. The calculated FRET distances for the long forms from the donor on Cys374 on actin were smaller than for the short Tm isoforms. The cooperativity of the response from S1 binding was cooperative in all forms, but the probe in long forms moved further from the donor, while that on the short forms moved closer.\nExperimental design\nsee above\nValidity of the findings\nWith structures and models now available, a value of FRET studies is the opportunity to interpret the observed FRET distances in terms of these models. This is where the manuscripts falls short, and it is quite confusing at best. The data used to make actin-Tm based on 3D reconstructions do not contain information about the axial position of Tm on actin. Previously published work from the Lehrer and Miki laboratories include computational studies to try to use the FRET data in positioning Tm on actin, considering the Tm probe could have FRET with the donor on more than one actin in the filament. The models in the present paper integrate the present FRET results into a model, rather than using the data as a way to test the model. Undoubtedly the FRET data are consistent with more than one model, given the size of the FRET distances.\nCartoons in Figure 4, that derive from published models (but it is not clear which one or how), are especially confusing, it that the authors have inferred that the complex between the N and C termini of 1b9d dissociates on actin, leading to different positions of the ends on actin (i.e. no continuous cable along the actin filament), and failure to interact with the expected sites on actin, D25, K326 and K238. This is especially surprising given the published results that the 1b9a ends form a tighter complex than 1b9d (based on model peptides). The interpretation of the ends coming apart is on the basis of trying to explain the long FRET value for the C terminus, compared to the others, and the shorter value of the N-terminal receptors (previously-published work). Is there another possible interpretation? Some of the reported difficulties working with 1b9a raise concern. Also, why is there no cartoon of stTm for comparison?\nThe observation that the N terminus determines the direction of movement in relation to the actin donor when myosin binds is particularly interesting. With the available of a molecular model of Tm on actin in the presence of myosin (Behrmann et al., 2012), some further discussion of this result would be worthwhile.\nAdditional comments\nSome other comments. The discussions of the structures in the PDB is a bit misleading. The Greenfield and Frye structures are very similar, despite being different genes and different methods (NMR and X-ray). The dynamics showing that the overlap complex is flexible come from the Greenfield structure and were confirmed by Frye in comparing their two structures. Also, it is the Greenfield work that showed the C-terminus opens up when forming a complex with the N terminus. Frye et al. did not determine the structure of the ends alone, and the other published structures of the C terminus from the Cohen and Maeda labs are of complexes of the C terminus with itself, in an antiparallel manner. The Murekami structure is completely unrelated to and inconsistent with those determined by Greenfield and Frye. In the overlap complexes, the C-terminal helices are not parallel; they are in the solution structure of the C terminus alone, but the chains splay apart when they form a complex with the N terminus in all the structures.\nThere is discussion about the myosin cooperativity, but not that myosin increases the affinity of Tm for actin (and visa versa). In 162 ff, this should be mentioned.\nThe paper needs to be reviewed carefully for English usage, particularly the use of articles. Also, the use of the term \u201corientation\u201d is misleading. In a polar structure such as the actin filament, the first reaction is that the orientation of the ends relative to each other is variable; better to use \u201cposition\u201d, I think that is what is suggested.\nIn summary, the data seem valid, but the interpretation is incomplete. Without incorporating the data into existing models in a relevant way, the value of the data significantly diminishes.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #3 of \"Structural differences between C-terminal regions of tropomyosin isoforms (v0.1)\". PeerJ https://doi.org/10.7287/peerj.181v0.1/reviews/3",
+    "all_reviews": "Review 1: Alla Kostyukova \u00b7 Sep 26, 2013 \u00b7 Academic Editor\nACCEPT\nFont size in some parts of the manuscript is smaller than it should be. If it is still a result of transferring the doc file to the pdf file, please, inform the production staff about this problem.\nReview 2: Alla Kostyukova \u00b7 Sep 9, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nYou don't need to do any additional experiments but you should revise your discussion and address other reviewer's comments.\nReview 3: Reviewer 1 \u00b7 Sep 8, 2013\nBasic reporting\nThe article provides sufficient introduction and background to support the experimental design of the study. Formatting of Figures appears appropriate and the article seems to conform with template information.\n\nMinor comments:\nIn the abstract, it should be clarified that the donor (AEDANS) is bound to the tropomyosin isoforms.\nIn the first paragraph of the introduction (line 18), S1 should be introduced as being the S1 subfragment of myosin which may not be clear to an audience not familiar with the field of research.\nTypo in line 75: Change \u2018Than cDNA from the first\u2026, to \u2018Then, cDNA from the first\u2026\u2019\nTypo in line 186: Change \u2018the donor-acceptor distance the absence\u2019 to \u2018the donor-acceptor distance in the absence\u2019\nReference \u2018Barua et al., 2013 (lines 281-283) needs volume and page information\nExperimental design\nThe research question is clearly defined and directly links this study with the findings of the authors from a previous study. The approach and experimental design has been validated by the previous study. The research question is a logical consequence from their previous findings. The method are described with sufficient detail.\nValidity of the findings\nIn the present study, Slivinska and Moraczewska determined isoform-specific localisation of tropomyosin C-termini along actin filaments and muscle myosin S1 subfragment-dependent changes in the localisation of the C-termini. This work is a direct follow-up study where the authors have tested the properties of the N-termini of the same tropomyosin isoforms. The current study reveals interesting information on the displacement of the C-termini of low- and high molecular weight tropomyosins in the presence of myosin S1 binding. These findings provide significant experimental evidence for an isoform specific mechanism by which different tropomyosins regulate the access of actin-associated proteins to the filaments.\nThe recognition of tropomyosins as \u2018gate keepers\u2019 of actin filaments is an emerging concept in the field of studying the regulation of cellular architecture. A large diversity in the regulation of actin-associated proteins by tropomyosins has particularly been explored previously for non-muscle tropomyosins. In this context it would be interesting to test differences in the S1 subfragment-dependent C-terminus displacement of tropomyosins for which isoform-specific regulation of the interaction between actin-associated proteins and the filament has been shown in different cell types. This includes for example the different properties of the TPM1 gene product TmBr3 (containing exons 1b and 9c) and the TPM3 gene product Tm5NM1 (containing exon 1b and 9d) (Bryce et al., 2003). Similarly, it would be interesting to test the effect of non-muscle myosin S1 (e.g. non-muscle myosin IIb S1) on displacement of the tropomyosin C-Termini due to the different binding properties of the muscle and non-muscle S1.\nAdditional comments\nno comments\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Structural differences between C-terminal regions of tropomyosin isoforms (v0.1)\". PeerJ https://doi.org/10.7287/peerj.181v0.1/reviews/1\nReview 4: Dmitrii Levitsky \u00b7 Sep 5, 2013\nBasic reporting\nThe manuscript 647 by \u015aliwi\u0144ska and Moraczewska describes the influence of the C-terminal sequence of tropomyosin (C-terminal exons 9a or 9d) on the distances between this part of Tm bound to F-actin and Cys374 of actin. For this purpose, four Tm isoforms with different C-terminal exons, 9a or 9d, were used, and the distances between fluorescence donor attached to Cys residue introduced into this Tm exon (Cys 269 or Cys 232) and acceptor attached to Cys374 of actin were measured by FRET in two different states of Tm on the surface of actin filament (closed and open).\n\nThis paper continues the previous paper of the authors (Sliwinska et al. (2011) Cytoskeleton 68, 300-312 ). Both papers are very similar (and sometimes even identical) in respect to Tm isoforms and the methods used, with the only difference that in the previous paper the fluorescence donor was attached to Cys residue in the N-terminal part of Tm, whereas in the present work it was introduced into the Tm C-terminal part. Taken together, the results of both studies show that the distances between Tm end-to-end overlap region and actin are different for Tm isoforms, and their changes in response to myosin S1 binding are individual for each isoform. Generally, I agree with authors\u2019 conclusion that the Tm isoforms are differently oriented on actin filament and the degree of Tm shift in response to the filament activation by myosin S1 binding (i.e. transition from closed to open state) is individual for each type of the studied isoform.\nExperimental design\nThe FRET experiments are carefully performed and clearly described, and the results are well explained.\nValidity of the findings\nThe work is mostly novel, although the main aspects of the research have been previously published by the authors (see above).\nAdditional comments\nn my opinion, the paper could be quite suitable for publication in PeerJ after some corrections according to the following comments.\n\n1). First line in the abstract: \u201cTropomyosins are actin regulatory proteins...\u201d\nIt seems to me, \u201cTropomyosins are actin-binding regulatory proteins..\u201d is more correct.\n\n2). Results, line 102: \u201c...illustration of the overlap regions are shown in Fig. 1.\u201d\nIt should be either \u201d...illustration of the overlap regions is shown in Fig. 1\u201d or \u201c...illustrations of the overlap regions are shown in Fig. 1.\u201d\n\n3). Table 1, upper line (exon 9a, red): it should be A/C, but not A/.\n\n4). The English needs to be edited carefully, and the formatting of the text should be improved (e.g., on lines 87-89, 185, 193-197, and 258-263 the size of the letters is too small).\nCite this review as\nLevitsky D (2013) Peer Review #2 of \"Structural differences between C-terminal regions of tropomyosin isoforms (v0.1)\". PeerJ https://doi.org/10.7287/peerj.181v0.1/reviews/2\nReview 5: Reviewer 3 \u00b7 Sep 1, 2013\nBasic reporting\nIn this paper the authors have used FRET to study the relationship of the C terminus of tropomyosin to Cys 374 on actin in Tm isoforms that would have different overlap complexes, and the cooperativity of myosin binding. The rationale is sound: there are structures of the overlap regions of tropomyosins, and models for actin-Tm based on Tm models and structures that do not contain the overlap region. Since the ends are known to be so important for Tm function and a source of isoform diversity (work from Dr. Morazcewsaka\u2019s lab and others), this is a reasonable approach and significant goal. Cysteine mutations were introduced just outside of the overlap region in four different Tms and shown to have little effect on the function of the Tm. The relative affinities are consistent with her previous work (Moraczewska et al., 1999) but not cited, and the fluorescence quenching is consistent with the affinities. The appropriate controls were carried out.\nThe FRET distances were determined for four Tms with sequences encoded by four different combinations of N-terminal and C-terminal sequences encoded by alternatively-expressed exons. Three occur naturally, the fourth is a synthetic construct (1b9a). While Tm2, Tm5a and 1b9a are all alpha Tm (TPM1) sequences, that used for the skTm is not clear: unacetylated alpha TM binds poorly; the citation is in a paper that studied gamma skTm (TPM3) that is 285 residues long due to an extra Met at the N terminus. This needs to be clarified, and also any differences from the common exons in the alpha-Tm encoded forms. The calculated FRET distances for the long forms from the donor on Cys374 on actin were smaller than for the short Tm isoforms. The cooperativity of the response from S1 binding was cooperative in all forms, but the probe in long forms moved further from the donor, while that on the short forms moved closer.\nExperimental design\nsee above\nValidity of the findings\nWith structures and models now available, a value of FRET studies is the opportunity to interpret the observed FRET distances in terms of these models. This is where the manuscripts falls short, and it is quite confusing at best. The data used to make actin-Tm based on 3D reconstructions do not contain information about the axial position of Tm on actin. Previously published work from the Lehrer and Miki laboratories include computational studies to try to use the FRET data in positioning Tm on actin, considering the Tm probe could have FRET with the donor on more than one actin in the filament. The models in the present paper integrate the present FRET results into a model, rather than using the data as a way to test the model. Undoubtedly the FRET data are consistent with more than one model, given the size of the FRET distances.\nCartoons in Figure 4, that derive from published models (but it is not clear which one or how), are especially confusing, it that the authors have inferred that the complex between the N and C termini of 1b9d dissociates on actin, leading to different positions of the ends on actin (i.e. no continuous cable along the actin filament), and failure to interact with the expected sites on actin, D25, K326 and K238. This is especially surprising given the published results that the 1b9a ends form a tighter complex than 1b9d (based on model peptides). The interpretation of the ends coming apart is on the basis of trying to explain the long FRET value for the C terminus, compared to the others, and the shorter value of the N-terminal receptors (previously-published work). Is there another possible interpretation? Some of the reported difficulties working with 1b9a raise concern. Also, why is there no cartoon of stTm for comparison?\nThe observation that the N terminus determines the direction of movement in relation to the actin donor when myosin binds is particularly interesting. With the available of a molecular model of Tm on actin in the presence of myosin (Behrmann et al., 2012), some further discussion of this result would be worthwhile.\nAdditional comments\nSome other comments. The discussions of the structures in the PDB is a bit misleading. The Greenfield and Frye structures are very similar, despite being different genes and different methods (NMR and X-ray). The dynamics showing that the overlap complex is flexible come from the Greenfield structure and were confirmed by Frye in comparing their two structures. Also, it is the Greenfield work that showed the C-terminus opens up when forming a complex with the N terminus. Frye et al. did not determine the structure of the ends alone, and the other published structures of the C terminus from the Cohen and Maeda labs are of complexes of the C terminus with itself, in an antiparallel manner. The Murekami structure is completely unrelated to and inconsistent with those determined by Greenfield and Frye. In the overlap complexes, the C-terminal helices are not parallel; they are in the solution structure of the C terminus alone, but the chains splay apart when they form a complex with the N terminus in all the structures.\nThere is discussion about the myosin cooperativity, but not that myosin increases the affinity of Tm for actin (and visa versa). In 162 ff, this should be mentioned.\nThe paper needs to be reviewed carefully for English usage, particularly the use of articles. Also, the use of the term \u201corientation\u201d is misleading. In a polar structure such as the actin filament, the first reaction is that the orientation of the ends relative to each other is variable; better to use \u201cposition\u201d, I think that is what is suggested.\nIn summary, the data seem valid, but the interpretation is incomplete. Without incorporating the data into existing models in a relevant way, the value of the data significantly diminishes.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #3 of \"Structural differences between C-terminal regions of tropomyosin isoforms (v0.1)\". PeerJ https://doi.org/10.7287/peerj.181v0.1/reviews/3\nReview 6: \n",
+    "LLM_extracted_review": "[\"Review 1: Font size in some parts of the manuscript is smaller than it should be. If it is still a result of transferring the doc file to the pdf file, please, inform the production staff about this problem.\\n\\nReview 2: You don't need to do any additional experiments but you should revise your discussion and address other reviewer's comments.\\n\\nReview 3: In the abstract, it should be clarified that the donor (AEDANS) is bound to the tropomyosin isoforms. In the first paragraph of the introduction (line 18), S1 should be introduced as being the S1 subfragment of myosin which may not be clear to an audience not familiar with the field of research. Typo in line 75: Change \u2018Than cDNA from the first\u2026, to \u2018Then, cDNA from the first\u2026\u2019 Typo in line 186: Change \u2018the donor-acceptor distance the absence\u2019 to \u2018the donor-acceptor distance in the absence\u2019. Reference \u2018Barua et al., 2013 (lines 281-283) needs volume and page information.\\n\\nReview 4: The English needs to be edited carefully, and the formatting of the text should be improved (e.g., on lines 87-89, 185, 193-197, and 258-263 the size of the letters is too small). First line in the abstract: \u201cTropomyosins are actin regulatory proteins...\u201d It seems to me, \u201cTropomyosins are actin-binding regulatory proteins..\u201d is more correct. Results, line 102: \u201c...illustration of the overlap regions are shown in Fig. 1.\u201d It should be either \u201d...illustration of the overlap regions is shown in Fig. 1\u201d or \u201c...illustrations of the overlap regions are shown in Fig. 1.\u201d Table 1, upper line (exon 9a, red): it should be A/C, but not A/.\\n\\nReview 5: The calculated FRET distances for the long forms from the donor on Cys374 on actin were smaller than for the short Tm isoforms. The cooperativity of the response from S1 binding was cooperative in all forms, but the probe in long forms moved further from the donor, while that on the short forms moved closer. The data used to make actin-Tm based on 3D reconstructions do not contain information about the axial position of Tm on actin. The interpretation of the ends coming apart is on the basis of trying to explain the long FRET value for the C terminus, compared to the others, and the shorter value of the N-terminal receptors (previously-published work). The discussions of the structures in the PDB is a bit misleading. The paper needs to be reviewed carefully for English usage, particularly the use of articles. Also, the use of the term \u201corientation\u201d is misleading. In a polar structure such as the actin filament, the first reaction is that the orientation of the ends relative to each other is variable; better to use \u201cposition\u201d, I think that is what is suggested.\"]"
+}

peerj_json_files/PeerJ_Json_132.json ADDED Viewed

The diff for this file is too large to render. See raw diff

peerj_json_files/PeerJ_Json_133.json ADDED Viewed

The diff for this file is too large to render. See raw diff

peerj_json_files/PeerJ_Json_134.json ADDED Viewed

	@@ -0,0 +1,17 @@

+{
+    "v1_Abstract": "Citation metrics and h indices differ using different bibliometric databases. We compiled the number of publications, number of citations, h index and year since the first publication from 340 soil researchers from all over the world. On average, Google Scholar has the highest h index, number of publications and citations per researcher, and the Web of Science the lowest. The number of papers in Google Scholar is on average 2.3 times higher and the number of citations is 1.9 times higher compared to the data in the Web of Science. Scopus metrics are slightly higher than that of the Web of Science. The h index in Google Scholar is on average 1.4 times larger than Web of Science, and the h index in Scopus is on average 1.1 times larger than Web of Science. Over time, the metrics increase in all three databases but fastest in Google Scholar. The h index of an individual soil scientist is about 0.7 times the number of years since the first publication. There is a large difference between the number of citations, number of publications and the h index using the three databases. From this analysis it can be concluded that the choice of the database affects widely used citation and evaluation metrics but that bibliometric transfer functions exist to relate the metrics from these three databases. We also investigated the relationship between journal\u2019s impact factor and Google Scholar\u2019s h5-index. The h5-index is a better measure of a journal\u2019s citation than the 2 or 5 year window impact factor.",
+    "v1_col_introduction": "introduction  : Scientific impact measures are increasingly being used for academic promotions, grant evaluations and evaluation of job vacancy candidates. They are also being used for the evaluations of university departments and research centres. Traditionally, the impact factor of a journal has been used \u2013 a metric developed by Garfield (1955) whereby the citations and number of papers published over a given period are divided. For most journals it shows considerable inter-annual fluctuation and it provides no information on individual papers nor individual authors. Since 2005, the h index has been used as an index for quantifying the scientific productivity of scientists based on their publication record (Hirsch, 2005). It is a personal index and provides information on the number of publications of an author and the number of citations: A scholar with an index of h has published h papers with at least h citations each. The h index can also be calculated for journals, departments, universities or countries.\nThe three widely used bibliometric databases for analysis and evaluations of\ncitations and the h index are Web of Science (Thomson Reuters), Scopus (Elsevier), and Google Scholar. Some papers have compared citations between these three databases. Although Google Scholar and Scopus seem to provide higher numbers of citations (Falagas et al., 2008), there is mixed information on the h index. For example, Bar-Ilan (2008) compared the h index for 47 highly-cited Israeli researchers across the three databases and concluded that the results from Google Scholar are considerably different from Web of Science and Scopus. Mingers and Lipitakis (2010) looked at 4,600 publications from three UK Business Schools, and found that Web of Science poorly covers the management discipline compared to Google Scholar. De Groote and Raszewski (2012) examined 31 faculty members from nursing faculty in the Midwestern USA, and concluded that more than one database should be used to calculate the h index. They further recommended that since the h index rankings differ among databases, comparisons between researchers should be done only within a specified database.\nThe difference between the three databases has been fairly well established and\nthe three databases will calculate different citations and h indices. As far as we know, the relationships between the three databases have not been investigated and derived. The aims of this paper are therefore: (i) to compare citations and h index across the three databases, (ii) to derive transfer functions to convert metrics from one database to\n11\n12\n13\n14\n15\n16\n17\n18\n19\n20\n21\n22\n23\n24\n25\n26\n27\n28\n29\n30\n31\n32\n33\n34\n35\n36\n37\n38\n39\n40\n41\n42\nPeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013)\nR ev ie w in g M an\nus cr ip t\nthe others, and (iii) to compare impact factors for journals and the h index. Hereto we have compared the data from 340 researchers and 31 journals. Since we are all three soil scientists, we have used only soil researchers and journals in this study.\nSoil science is a study of soil as a natural phenomenon and resource (Brevik and\nHartemink, 2010). It is a relatively small discipline, in terms of number of researchers, number of papers per annum, and citations. The IUSS (International Union of Soil Sciences) database lists about 50,000 soil scientists worldwide, but only a fraction of these are in research and actively publish, with a guestimate of 5,000 to 10,000 publishing researchers. The \u201csoil\u201d topic has lower number of papers and citations when compared to other subjects of natural resources such as \u201cair\u201d and \u201cwater\u201d (Minasny et al., 2007). The number of published papers in 2011 according to Scopus with \u201csoil\u201d in the abstract and keywords is 39,504, with a rate of increase of about 2,000 papers per year. In comparison the number of papers in 2011 on \u201cair\u201d is 1.4 times larger and the number of papers on \u201cwater\u201d is 3.5 times larger. The h index ratios for water, air, and soil (for the papers published in 2011) are 1.7, 1.3, and 1.0. Nevertheless soil is becoming more important with strong links to global issues of food security, biodiversity, land use change, and climate change (McBratney et al., 2014). While this study only used soil researchers the bibliometric results are illustrative to other agricultural, environmental, earth science and biology disciplines, and to small scientific disciplines in general.",
+    "v2_Abstract": "Citation metrics and h indices differ using different bibliometric databases. We compiled the number of publications, number of citations, h index and year since first publication for 340 soil researchers from all over the world. On average, Google Scholar has the highest h index, number of publications and citations per researcher, and the Web of Science the lowest. The number of papers in Google Scholar is on average 2.3 times larger and the number of citations is 1.9 times larger compared with data in the Web of Science. Scopus metrics are slightly larger than that of the Web of Science. Over time, the metrics increase in all three databases but fastest in Google Scholar. The h index of an individual soil scientist is about 0.7 times the number of years since the first publication. About 10% of the h index is caused by self-citation but that may be higher for younger authors. There is a large difference between the number of citations, number of publications and the h index using the three different databases. We also compared journal impact factor and the h5-index from Google Scholar in 31 soil science journals. The h5-index is a better measure of a journal\u2019s citation than the 2or 5-year window impact factor. From this analysis it can be concluded that the choice of the database affects widely used citation and evaluation metrics but that pedobibliometric transfer functions exist to relate the metrics from these three databases.",
+    "v2_col_introduction": "introduction  : Scientific impact measures are increasingly being used for academic promotions, grant evaluations and evaluation of job vacancy candidates. They are also being used for the evaluations of university departments and research centres. Traditionally, the impact factor of a journal was being used \u2013 a metric developed by Garfield (1955) whereby the citations and number of papers published over a given period (usually 2 years) are divided. For most journals it shows considerable inter-annual fluctuation and it provides no information on individual papers nor individual authors. Since 2005, the h index has been used as an index for quantifying the scientific productivity of scientists based on their publication record (Hirsch, 2005). It is a personal index and provides information on the number of publications of an author and the number of citations: A scholar with an index of h has published h papers with at least h citations each. The h index can also be calculated for journals, departments, universities or countries.\nThe three widely used bibliometric databases for analysis and evaluations of\ncitations and the h index are Web of Science (Thomson Reuters), Scopus (Elsevier), and Google Scholar. Some papers have compared citations between these three databases. Although Google Scholar and Scopus seem to provided higher number of citations (Falagas et al., 2008), there is mixed information on the h index. For example, Bar-Ilan (2008) compared the h index for 47 highly-cited Israeli researchers across the three databases and concluded that the results from Google Scholar are considerably different from Web of Science and Scopus. Mingers and Lipitakis (2010) looked at 4,600 publications from three UK Business Schools, and found that Web of Science poorly covers the management discipline compared to Google Scholar. De Groote and Raszewski (2012) examined 31 faculty members from nursing faculty in the mid-west USA, and concluded that more than one databases should be used to calculate the h index. They further recommended that since the h index rankings differ among databases, comparisons between researchers should be done only within a specified database.\nThe difference between the three databases has been fairly well established and\nthe three databases will calculate different citations and h indices. As far as we know, the relationships between the three databases have not been investigated and derived. The aims of this paper are therefore: (i) to compare citations and h index across the\n10\n11\n12\n13\n14\n15\n16\n17\n18\n19\n20\n21\n22\n23\n24\n25\n26\n27\n28\n29\n30\n31\n32\n33\n34\n35\n36\n37\n38\n39\n40\n41\nPeerJ reviewing PDF | (v2013:09:794:0:1:NEW 9 Sep 2013)\nR ev ie w in g M an\nus cr ip t\nthree databases, (ii) to derive transfer functions to convert metrics from one database to the others, and (iii) to compare impact factors for journals and the h index. Hereto we have compared the data from 340 researchers and 31 journals. Since we are all three soil scientists, we have used only soil researchers and journals in this study.\nData and Methods\nGoogle Scholar (GS) is a bibliographic database freely available from Google. It was introduced in 2004 and contains scholarly works across many disciplines and sources, including theses, books, reports, abstracts, peer-reviewed and non-reviewed articles, and web pages that are deemed scholarly. Google Scholar lists these automatically from its search engine activities (Harzing and van der Wal, 2009; Vine, 2006). An individual Google Scholar page was featured in 2012, where a researcher can create a webpage, with fields of interest. Google Scholar automatically searches and populates the individual\u2019s publications, calculates and displays the individual's total number of citations, h index, and i10 index. Scopus, or SciVerse Scopus, is a bibliographic database from Elsevier which contains abstracts and citations for academic journal articles, conference papers, and book chapters. Inclusion in the database is through the Scopus Content Selection and Advisory Board. Although its record goes back as early as 1823, its citations is reliable after 1995. The Web of Science is a bibliographic database from Thompson Reuters which only contained abstracts and citations for articles listed in the Web of Science indexed journals since 1900 (Harzing and van der Wal, 2009).\nData from researchers with the following areas of interest in: \u201csoil science\u201d, \u201csoil\u201d,\n\u201cpedology\u201d, \u201csoil physics\u201d, \u201csoil biology\u201d, \u201csoil chemistry\u201d, \u201csoil fertility\u201d, \u201csoil erosion\u201d, \u201csoil ecology\u201d, and \u201csoil carbon\u201d were retrieved from the Google Scholar\u2019s author page. The same researcher was located in Scopus and the Web of Science. In Scopus, the \u2018Author Identifier\u2019 tool was used to locate the researcher. In the Web of Science, the author\u2019s surname and first name\u2019s initial was used, together with \u201csoil\u201d in the search subject. When the name and publication were inconsistent across all three databases, the researcher was not included in our analysis. At the end, we collected data from 340 researchers and this included: number of total citations, h index, number of papers, and year of the first publication. These data were obtained for each researcher and from each of the three databases. The publications and citations are until June 2013.\n42\n43\n44\n45\n46\n47\n48\n49\n50\n51\n52\n53\n54\n55\n56\n57\n58\n59\n60\n61\n62\n63\n64\n65\n66\n67\n68\n69\n70\n71\n72\n73\nPeerJ reviewing PDF | (v2013:09:794:0:1:NEW 9 Sep 2013)\nR ev ie w in g M an\nus cr ip t",
+    "v1_text": "results and discussion : Number of papers, citations and h index Table 1 shows the statistics of h index, number of publications, number of citations, and year of the first paper for 340 soil researchers in the three databases. Our data encompass a wide range of researchers from early-career to well-established and highly-cited researchers. The database is much larger and more diverse than previous studies where a small and focussed group of researchers was used to compare citation metrics between the databases (e.g. Franceschet, 2010; Meho and Rogers 2008; Patel et al., 2013). The median number of papers for the 340 soil researchers ranged from 23 (Web of Science) to 79 (Google Scholar) with Scopus having intermediate values. The number of citations is also highest in Google scholar, with a median of 866 citations per author whereas it is 291 in the Web of Science. The h index and its annual increase are lowest 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t in the Web of Science. This pattern holds for all of the metrics presented here: Google Scholar has the highest numbers and the Web of Science the lowest whereas the Scopus numbers are in between. Part of this may be the different types of publications included and also the periods of time covered by the 3 databases are slightly different. A simple linear regression without intercept was performed between the citation indices of the three databases (Table 2). Google Scholar has on average 2.3 times more articles and 1.9 times more citations than the Web of Science. The Scopus database (all years) has 1.1 times more papers than the Web of Science but a similar number of citations compared to the Web of Science. Since the citations are more correct and complete after 1995, a revision was made to the relationship for post 1995 authors; it shows that Scopus has about 1.2 times more citations than the Web of Science. The 20% higher citations are consistent with the findings by Falagas et al. (2008) in the field of medicine. Similarly, for articles in medical journals, Kulkarni et al. (2009) found that Google Scholar and Scopus retrieved more citations compared to Web of Science (1.22 and 1.20 times respectively). <Table 1. Somewhere here> <Table 2. Somewhere here> The relationship between number of papers and citations is scattered, especially for the number of papers (Fig. 1), however the relationships between h index values across the 3 databases appear to be quite linear. The h index in Google Scholar is on average 1.4 times larger than Web of Science, and the h index in Scopus (post 1995 authors) is on average 1.1 times larger than Web of Science. However for pre-1995 authors, their Scopus h index is similar and sometimes can be smaller when compared to Web of Science. While Google Scholar contains more grey literature (informally published written material) and its citations may contain errors (Harzing and van der Wal, 2009), the h index appears to be quite robust and comparable with Web of Science and Scopus. This is due to the fact that h index does not vary greatly if the number of articles increases (e.g. book chapters or unrefereed articles). In addition, extra citations do not have a large effect on the h index, as once a paper has reached h citations additional citations to that paper do not affect its value (Franceschini et al., 2013; Courtault and Hayek, 2008). 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t <Fig. 1. Somewhere here> Our results are different from the study by Franceschet (2010) who evaluated 13 computer scientists from his university\u2019s department and found that on average Google Scholar had five times more papers, eight times more citations and a three-fold larger h index. Our results from 340 soil researchers are more in line with De Groote and Raszewski (2012) who looked at 30 researchers from nursing and found that the h index from Google Scholar is 1.3 times larger than the Web of Science, and Scopus is 1.1 times larger than the h index in the Web of Science. Similar results were obtained by Meho and Rogers (2008) who evaluated 22 human-computer interaction researchers from the UK and found that the h index in Google Scholar is on average 1.6 times higher than Web of Science. Patel et al. (2013) compared publications and citations for 195 Nobel laureates in Physiology and Medicine using the three databases. They found no concordance between the three databases when considering the number of publications and citations count per laureate. However, the h index was the most reliably calculated bibliometric index across the three databases. We calculated the Spearman\u2019s rank correlation (\u03c1) of the h index of the 340 researchers from the three databases. The three databases show excellent correlation for the h index, with WoS and GS as having the largest concordance. Unexpectedly, the rank correlation in terms of no. citations and no. papers (Table 3) also indicates that the three databases are comparable. This implies that the ranking of individuals within a database is comparable with the other databases. Our correlation is also much higher compared to the 13 computer scientists studied by Franceschet (2010) who only obtained \u03c1 = 0.65. We used a much larger dataset, and the GS data came from the page that was created by the researcher, thus the listed papers and citations are assumed to be more complete. <Table 3. Somewhere here> a comparison between 5 year impact factor (if) and google scholar h5-index for 31 soil : science journals in 2012. PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t Figure 6 (a) Relationship between cites and 5 year Impact factor (IF), and (b) relationship between cites and Google Scholar h5-index for soil science journals in 2012. Cites is the number of citations in 2012 for papers that were published in 2007 \u2013 2011. PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t Table 1(on next page) Descriptive statistics of publication indices from Google Scholar, Scopus and Web of Science database for 340 soil researchers PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t Table 1. Descriptive statistics of publication indices from Google Scholar, Scopus and Web of Science database for 340 soil researchers Number of citations Number of papers First year of publication h index m (rate of increase per year) conclusions : From this analysis the following can be concluded: - There is a large difference between the number of citations, number of publications and the h index using the three different databases. - On average, Google Scholar gives the largest number of publications, largest number of citations and the highest h index. The Web of Science gives the lowest averages. - There are solid relationships between the h indices in these three databases. - The h5-index has a correlation with the five year impact factor, however it is more robust and less affected by citation manipulation. It should be considered as an alternative to the journal\u2019s impact factor. This analysis has shown that the choice of the database affects the assessment of scientific impact for academic promotions, grant evaluations, job vacancy candidates or the evaluations of university departments and research centres. It is recommended that we should quote these bibliometric indices for all three databases as they reflect different types of publications. Web of Science uses mostly refereed journal articles, Scopus includes conference proceedings and book chapters, whereas Google Scholar includes other publications (including software). The established relationships between the databases (Table 2) can be used as bibliometric transfer functions by anyone interested in relating databases. We are not aware of whether these functions have been established for other scientific disciplines but assume they will be similar. As a test, we applied our function relating the h index of WoS and GS to the 30 nursing faculty data of DeGroote and Raszewski (2012) and the function gives a good prediction with a Spearman rank correlation of 0.852. We envisage that these functions would work better for science than socio-economical disciplines. However, this needs to be investigated. Although we focussed on the relatively small discipline of soil science, the reported researchers has a lot of cross-over with other disciplines, in particular earth science, agricultural science, biogeochemistry and ecology. Many researchers in ecology and microbiology work with soil as a medium, while they do not necessary study soil as a natural body in classical pedology. Their contributions also elevated the citations as 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t compared to pure soil research. The trend of h index for soil researchers appears to be in between the water and biochemistry disciplines (McCarty and Jawitz, 2013). However, soil science publication rate (average 2.5 papers per year per researcher) is lower compared to water and biochemistry (average of 3.1 and 3.8 papers per year, respectively). 335 336 337 338 339 PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t the h index of soil researchers : In an earlier paper (Minasny et al., 2007) we investigated the relationship between the h index of 228 soil researchers and found that the index was 0.7 times the number of years since the first publication (which we called scientific age, or t). That means if a 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t researcher has been publishing for 10 years his/her h index should be about 7. We calculated this index using the Web of Science database, and now we repeated this using analysis to the 340 soil researchers in this study (which are different from the previous list) (Fig. 2). Although the data are scattered the relationship holds: h index = 0.73 x scientific age (R2 = 0.72). The Web of Science database shows that the average rate of h index increase over time (m) is 0.7, with the lowest value of 0.06 and highest value of 2.9 (Table 1). The average m value for Scopus is 0.7 and for Google Scholar it is 0.8 (Table 1). McCarty and Jawitz (2013) evaluated the linear relationship between scientific age and h index for 4 disciplines and found the following mean m values of 0.83, 0.47, 0.43, and 0.36 for biochemistry, water, economics, and anthropology, respectively. Thus the trend of soil science is in between water and biochemistry. For selected researchers, we tried to calculate the distribution of m (h index divided by the number of years since first publication) as a function of sub-disciplines in soil science. Table 5 shows the distribution of m for WoS and GS according to 6 subdisciplines. It shows that h index varies between sub-diciplines, for WoS, soil biology, biogeochemistry and ecology have the highest m values (median of 0.8). This is followed by soil physics, soil fertility and management, soil geography and pedometrics, chemistry and lastly pedology (average m = 0.5). The order in Google Scholar is slightly different, but it is consistent in that soil biology has the highest m value and pedology is the lowest. Therefore within soil science, the sub-disciplines also vary in terms of h index. The citation ratios are:- Soil biology, ecology and biogeochemistry, Soil management and fertility, Soil geography & pedometrics, Soil physics, Soil chemistry and mineralogy, Pedology 1: 0.9: 0.8: 0.8: 0.8: 0.6; respectively. Although the number of citations for researchers across the three databases can be quite different, the relationship between the number of citations and the h index is quite consistent across the three databases (Fig. 3): h index = \u00bd n\u00bd (R2 = 0.95) 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t This relationship follows the function postulated by Hirsch (2005) where number of citations is about 3 to 5 times h2, and it appears the h index follows an absorption-type relationship (Warrick, 2003), increasing rapidly at low numbers of citations with the rate decreasing with increasing number of citations. <Fig. 2. Somewhere here> <Fig. 3. Somewhere here> Table 4 shows the relationship of the average number of papers and citations per year for the 340 soil researchers. This can be interpreted as: on average, a soil researcher produces 5 articles per year, 2 articles in international refereed journals, 1 in a conference proceedings, and 2 other unrefereed publications. The researcher receives 65 citations per year from journal articles, an additional 13 citations from conference proceedings and another 44 citations from other publications. <Table 4. Somewhere here> <Table 5. Somewhere here> 1 times larger than web of science. over time, the metrics increase in all three databases : but fastest in Google Scholar. The h index of an individual soil scientist is about 0.7 times the number of years since the first publication. There is a large difference between the number of citations, number of publications and the h index using the three databases. From this analysis it can be concluded that the choice of the database affects widely used citation and evaluation metrics but that bibliometric transfer functions exist to relate the metrics from these three databases. We also investigated the relationship between journal\u2019s impact factor and Google Scholar\u2019s h5-index. The h5-index is a better measure of a journal\u2019s citation than the 2 or 5 year window impact factor. PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t Citations and the h index of soil researchers and journals in the Web of Science, data and methods : Google Scholar (GS) is a bibliographic database freely available from Google. It was introduced in 2004 and contains scholarly works across many disciplines and sources, including theses, books, reports, abstracts, peer-reviewed and non-reviewed articles, and web pages that are deemed scholarly. Google Scholar lists these automatically from its search engine activities (Harzing and van der Wal, 2009; Vine, 2006). An individual Google Scholar page was featured in 2012, where a researcher can create a webpage, with fields of interest. Google Scholar automatically searches and populates the individual\u2019s publications, calculates and displays the individual's total number of citations, h index, and i10 index. Scopus, or SciVerse Scopus, is a bibliographic database from Elsevier which contains abstracts and citations for academic journal 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t articles, conference papers, and book chapters. Inclusion in the database is through the Scopus Content Selection and Advisory Board. Although its record goes back as early as 1823, its citations are reliable after 1995. The Web of Science is a bibliographic database from Thompson Reuters which only contained abstracts and citations for articles listed in the Web of Science indexed journals since 1900 (Harzing and van der Wal, 2009). Data from researchers who listed their areas of interest as: \u201csoil science\u201d, \u201csoil\u201d, \u201cpedology\u201d, \u201csoil physics\u201d, \u201csoil biology\u201d, \u201csoil chemistry\u201d, \u201csoil fertility\u201d, \u201csoil erosion\u201d, \u201csoil ecology\u201d, and \u201csoil carbon\u201d were retrieved from the Google Scholar author pages. The same researchers were located in Scopus and the Web of Science. In Scopus, the \u2018Author Identifier\u2019 tool was used to locate the researcher. In the Web of Science, the author\u2019s surname and first name\u2019s initial was used, together with \u201csoil\u201d in the search subject. When the name and publication record were inconsistent across all three databases, the researcher was not included in our analysis. At the end, we collected data from 340 researchers and this included: number of total citations, h index, number of papers, and year of the first publication. These data were obtained for each researcher and from each of the three databases. The publications and citations are until June 2013. 340 soil researchers from all over the world. on average, google scholar has the highest h : index, number of publications and citations per researcher, and the Web of Science the lowest. The number of papers in Google Scholar is on average 2.3 times higher and the number of citations is 1.9 times higher compared to the data in the Web of Science. Scopus metrics are slightly higher than that of the Web of Science. The h index in Google Scholar is on average 1.4 times larger than Web of Science, and the h index in Scopus is on average PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t Figure 2 Relationship between the scientific age (t) of 340 soil researchers and the h index (Web of Science data). PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t Figure 3 Relationship between the number of citations and the h index of 340 soil researchers from 3 databases Black dots are data from Web of Science, green squares are from Scopus, and blue triangles are from Google Scholar. PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t Figure 4 (a) Relationship between the h index with and without self-citation, (b) relationship between the scientific age of 340 soil researchers and percentage self-citation based on Scopus data. PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t Figure 5 scopus, and google scholar : Budiman Minasny 1 *, Alfred E. Hartemink 2 , Alex. McBratney 1, Ho-Jun Jang1 1 The University of Sydney, Department of Environmental Sciences, Faculty of Agriculture & Environment, NSW 2006, Australia. 2 University of Wisconsin \u2013 Madison, Department of Soil Science, FD Hole Soils Lab, 1525 Observatory Drive. Madison, WI, 53706, USA. ( * corresponding author) E-mail budiman.minasny@sydney.edu.au 1 2 3 4 5 6 7 8 9 10 PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t self-citations : A way to boost the h index is by self-citation. Hyland (2003) found that self-citation is 12% of all references in biology, engineering and physics, compared to 4% in sociology, philosophy, linguistics, or marketing. For soil science journals, we found a mean of 12% self-citations but it differs between the sub-disciplines (Minasny et al., 2010). High rates of self-citation were accompanied by high journal impact factor ranking; China and the USA had the highest rates of self-citation whereas Egypt, Algeria, Ukraine, and Indonesia have low levels of self-citations in soil science (Minasny et al., 2010). So high rates of self-citation may influence the h index and the Scopus database allows calculation of the h index with and without self-citation. Self-citation here is the so-called diachronous kind (Lawani, 1982), which is self-citation from the citations received by the author. The other type is called synchronous which is more difficult to 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t calculate, i.e. author\u2019s self-referencing relative to the total number of references cited in a paper. The relationship for the 340 soil researchers is consistent and the average h index without self-citation is about 12% lower (Fig. 4): h index without self-citation = 0.88 \u00d7 h index (R2 = 0.97). We found a weak relationship between percentage of self-citation and scientific age (t) (Fig. 4). It suggests that some younger authors appear to have high rates of self citation as their works were not known widely and their citations mainly come from themselves, as the researchers mature their papers are more widely known and more external citations were gained thus a lower percentage of self-citations: Percent self-citation = 42 - 5 t0.5 (R2 = 0.18). <Fig. 4. Somewhere here> journal citations : We retrieved 31 Soil Science journal impact factors (IF) and other metrics from the 2012 Thompson Reuters Journal Citation Reports (JCR, released in June 2013). Google Scholar also has measures of journal\u2019s metric, the h5-index, which is the h index for articles published in that journal for the last five years. The list of journals for the soil science discipline in Google Scholar is slightly different from the Thompson Reuters Journal Citation Reports (JCR), and therefore we used the journals listed in JCR as the basis for comparison. We searched for the h5-index for the journals in Google Scholar metrics for 2012 (released July 2013). Table 6 shows that Google Scholar h5-index has a better correlation with the five year IF (impact factor) than the two year IF, and Figure 5 shows the comparison between GS h5-index and the five year IF. While the h5-index and five year IF have a high rank correlation (\u03c1 = 0.90), the ranking is different for different journals. The journals \u2018Soil Biology and Biochemistry\u2019 and \u2018Plant and Soil\u2019 both consistently ranked no. 1 and 2 in JCR and GS while other journals appear to be slightly different in their ranking (1 to 3 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t places difference). The top two journals are able to maintain a large number of citations relative to the number of papers they publish. <Table 6. Somewhere here> <Fig. 5. Somewhere here> There are 4 journals that are ranked much higher (>= 4 difference in rank) in Google Scholar compared to the IF: \u2018Soil Science Society of America Journal\u2019, \u2018Journal of Plant Nutrition and Soil Science\u2019, \u2018Pedosphere\u2019, and \u2018Revista Brasileira de Ci\u00eancia do Solo\u2019. All these journals are published by national soil science societies (USA, Germany, China and Brazil). In the case of \u2018Revista Brasileira de Ci\u00eancia do Solo\u2019 which ranked 12 in GS and 25 in JCR, Google scholar includes more citations from non-English articles. Contrarily, there are four journals that are ranked much lower (<= 4 difference in rank) in Google Scholar: \u2018European Journal of Soil Science\u2019, \u2018Soil Use and Management\u2019, \u2018Journal of Soil and Water Conservation\u2019, and \u2018Soil Science\u2019. The Thompson Reuters Journal Citation Reports suffers from a miscalculation, for example \u2018Australian Journal of Soil Research\u2019 was reported to have a 2 year IF of 3.443. This is a miscalculation, as the journal changed its name to \u2018Soil Research\u2019 in 2011, and the IF calculation for Australian Journal of Soil Research only accounts for papers published until 2010. \u2018Soil Research\u2019 was again listed as a separate journal in JCR. We have recalculated the actual impact factor for this journal in our analysis. While there is a positive correlation between cites (citations in 2012 to papers published from the previous 5 years) and IF, we can see that there are 2 trends (Fig. 6a). For journals that published <700 papers between 2007 and 2011 (or on average less than 140 papers per year) IF tends to increase rapidly with increasing citations (1.2 increase in IF per 1000 citations). For the other 7 journals that published more than 700 papers, the slope is half as much (0.6 IF increase per 1000 citations). So there is a drawback for journals that publish more papers. Meanwhile the h5-index is mostly controlled by number of citations following an absorption relationship (Fig. 6b). Although the citations come from WoS, the h5-index still holds the square-root relationship supporting its robustness. Table 6 also shows that the GS h5-index is more correlated to the Eigenfactor metric compared to IF. The Eigenfactor metric is based on the Google PageRank 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t algorithm which calculated the \u201cinfluence\u201d of the journal based on the citations weighted by the \u201cquality\u201d of the journal (Bergstrom, 2007). A citation from a highly cited journal will have a higher weighting than a lower cited journal. We showed that this metric is less vulnerable to self-citation than the impact factor (Minasny et al., 2010). Interestingly, Google Scholar does not apply its PageRank for citations. Vanclay (2006) and Cortault and Hayek (2008) established that the h index is robust and is relatively unaffected by grey literature and errors in citations such as in Google Scholar. Most of the errors (and distortions) in citation databases are found in the \u2018long tails\u2019 of the citation distribution and they tend not to affect the h index much. For the journals considered here, we calculated the ratio between h5-index and number of papers, and it shows that only 1- 9% (median 5%) of the total papers that contributed to h5-index. In other words, less than 10% of the cited papers are influencing the h5-index. We also demonstrated that the h5-index keeps its relationship even when using WoS citations. Harzing and van der Wal (2009) recommended the use of the GS h index for Management and International Business journals. We also concur that the h5-index is a better measure of a journal\u2019s citation performance than the impact factor as it is more robust and less affected by citation manipulation. It is now acknowledged that there are ways of manipulating impact factor, which include self-citation, and editorials that listed references to previously published articles (Falagas et al., 2008). The h index is less sensitive to the increase in number of citations, while individual highly cited papers can artificially increase the impact factor. In addition, it only considers the top influential h papers in the journal, thus it does not penalise a journal for publishing a larger number of papers. Although h index can also be manipulated by self-citation, in order to increase the h index considerably, a journal has to be more tactical by increasing a significant number of citations to certain papers. <Fig. 6. Somewhere here> 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t google scholar : Minimum 16 3 1953 1 0.09 25th Quantile 266 32 1985 26 0.56 Median 866 79 1993 15 0.85 75th Quantile 2596 146 2001 26 1.18 Maximum 49447 1159 2011 115 3.67 PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t Variable By Variable Spearman \u03c1 Prob > |\u03c1| WoS h index GS h index 0.939 <.0001 Scopus h index GS h index 0.931 <.0001 WoS h index Scopus h index 0.922 <.0001 WoS no. citations GS no. citations 0.939 <.0001 Scopus no. citations GS no. citations 0.955 <.0001 WoS no. citations Scopus no. citations 0.945 <.0001 WoS no. papers GS no. papers 0.840 <.0001 Scopus no. papers GS no. papers 0.896 <.0001 WoS no. papers Scopus no. papers 0.905 <.0001 2 3 PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t Table 4(on next page) Comparison of the h index over time using the data from 340 soil researchers in Google scopus : Minimum 1 1 1955 1 0.03 25th Quantile 116 14 1989 5 0.45 Median 469 34 1996 11 0.71 75th Quantile 1361 65 2004 19 1.00 Maximum 28693 423 2011 70 2.87 Authors who started to publish after 1995 Scopus h index = 1.11 WoS h index 0.02 0.954 2 3 PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t Table 3(on next page) Spearman\u2019s rank correlation coefficient ( ) of the \u03c1 h index of the 340 researchers using web of science : Minimum 1 1 1957 1 0.06 25th Quantile 76 10 1991 5 0.41 Median 291 23 1998 10 0.67 75th Quantile 945 48 2004 17 1.00 Maximum 32837 424 2011 96 2.87 1 2 PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t Table 2(on next page) Comparison of publication indices from Google Scholar (GS), Scopus and Web of science (wos) : PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t Standard error of estimates R2 gs no. papers = 2.33 wos no. papers 0.06 0.797 : scholar (gs), scopus and web of science (wos) : PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t Table 4. Comparison of the h index over time using the data from 340 soil researchers in Google Scholar (GS), Scopus and Web of Science (WoS) Standard error of estimates R2 GS h index = 0.84 \u00d7 year 0.02 0.745 WoS h index = 0.73 \u00d7 year 0.02 0.717 Scopus h index = 0.73 \u00d7 year 0.02 0.759 GS no. papers = 5.5 \u00d7 year 0.23 0.620 WoS no. papers = 2.5 \u00d7 year 0.12 0.567 Scopus no. papers = 3.0 \u00d7 year 0.12 0.656 GS no. citations = 122 \u00d7 year 9.1 0.344 WoS no. citations = 65 \u00d7 year 5.8 0.269 Scopus no.citations = 78 \u00d7 year 5.8 0.346 1 2 PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t Table 5(on next page) The distribution of m (h index divided by the number of years since first publication) according to sub-disciplines in soil science using the data from Google Scholar, Scopus and Web of Science, n is the number of samples, Q25 and Q75 refers t PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t n Min Q25 Median Q75 Max Google Scholar Pedology 25 0.09 0.37 0.56 0.83 1.67 Soil chemistry 42 0.09 0.45 0.78 1.16 2.00 Soil physics 67 0.21 0.56 0.79 1.00 2.55 Soil geography & pedometrics 28 0.16 0.58 0.87 1.02 1.93 Soil management & fertility 42 0.28 0.69 0.98 1.30 2.12 Soil biology, ecology, biogeochemistry 88 0.23 0.78 1.01 1.65 3.67 Web of Science Pedology 25 0.08 0.32 0.46 0.78 1.67 Soil chemistry 42 0.06 0.32 0.63 1.00 1.67 Soil geography & pedometrics 67 0.11 0.50 0.64 0.86 1.50 Soil management & fertility 28 0.15 0.47 0.67 1.00 1.63 Soil physics 42 0.17 0.46 0.72 1.00 2.14 Soil biology, ecology, biogeochemistry 88 0.14 0.63 0.83 1.33 2.87 2 3 4 5 6 PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t Table 6(on next page) Spearman\u2019s rank correlation coefficient ( ) of the Google Scholar h5-index and impact \u03c1 factor (IF), no. papers, citations, and Eigenfactor metrics from Journal Citation Reports for 31 soil science journals. Cites is the number of citations in 2012 for pape PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t Table 6. Spearman\u2019s rank correlation coefficient (\u03c1) of the Google Scholar h5-index and impact factor (IF), no. papers, citations, and Eigenfactor metrics from Journal Citation Reports for 31 soil science journals. Cites is the number of citations in 2012 for papers that were published in 2007 \u2013 2011. variable by variable spearman \u03c1 prob > |\u03c1| : h5-index Cites (5 years) 0.972 <.0001 h5-index Eigenfactor 0.970 <.0001 h5-index 5 year IF 0.903 <.0001 h5-index 2 year IF 0.870 <.0001 h5-index No. papers (5 years) 0.721 <.0001 2 3 4 5 6 PeerJ reviewing PDF | (v2013:09:794:1:1:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t",
+    "v2_text": "results and discussion : Number of papers, citations and h index Table 1 shows the statistics of h index, number of publications, number of citations, and year of the first paper from 340 soil researchers in the three databases. Our data encompass a wide range of researchers from early-career to well-established and highly-cited researchers. The database is much larger and more diverse than previous studies where a small and focussed group of researchers was used to compare citation metrics between the databases (e.g. Franceschet, 2010; Meho and Rogers 2008; Patel et al., 2013). The median number of papers for the 340 soil researchers ranged from 23 (Web of Science) to 79 (Google Scholar) with Scopus having intermediate values. The number of citation is also highest in Google scholar and the median is 866 citations per author whereas it is 291 in the Web of Science. The h index and its annual increase are lowest in the Web of Science. This pattern holds for all of the metrics presented here: Google Scholar has the highest numbers and the Web of Science the lowest whereas the Scopus numbers are in between. Part of this may be the effect of different types of publications being included and also the periods are slightly different. A simple linear regression without intercept was performed between the citation indices of the three databases (Table 2). Google Scholar has on average 2.3 times more articles and 1.9 times more citations than the Web of Science. The Scopus database (all years) has 1.1 times more papers than the Web of Science but a similar number of citations compared to the Web of Science. Since the citations are complete and more correct after 1995 and revising the relationship for post 1995 authors, it shows that Scopus has about 1.2 times more citations than the Web of Science. The 20% extra citations are consistent with the finding by Falagas et al. (2008) in the field of medicine. Similarly, for articles in medical journals, Kulkarni et al. (2009) found that Google Scholar and Scopus retrieved more citations compared to Web of Science (1.22 and 1.20 times respectively). 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 PeerJ reviewing PDF | (v2013:09:794:0:1:NEW 9 Sep 2013) R ev ie w in g M an us cr ip t <Table 1. Somewhere here> <Table 2. Somewhere here> The relationship of number of papers and citations is scattered, especially for the number of papers (Fig. 1), however the relationships between h index values across the 3 databases appear to be quite linear. The h index in Google Scholar is on average 1.4 times larger than Web of Science, and the h index in Scopus (post 1995 authors) is on average 1.1 times higher than Web of Science. However for pre-1995 authors, their Scopus h index is similar and can be smaller when compared to Web of Science. While Google Scholar contains more grey literature (informally published written material) and its citations may contain errors (Harzing and van der Wal, 2009), the h index appears to be quite robust and comparable with Web of Science and Scopus. This is due to the fact that h index does not vary greatly if the number of articles increases (e.g. book chapters or unrefereed articles). In addition, extra citations will not affect much of the h index, as once a paper has reached h citations the additional citations to that paper does not affect its value (Franceschini et al., 2013; Courtault and Hayek, 2008). <Fig. 1. Somewhere here> Our results are different from the study from Franceschet (2010) who evaluated 13 computer scientists from his university\u2019s department. Franceschet (2010) found a much higher number when comparing Google Scholar to Web of Science indicators: on average Google Scholar has five times more papers, eight times more citations and a three times higher h index. Our results of 340 soil reearchers are more in line with De Groote and Raszewski (2012) who looked at 30 researchers from nursing and found that for the h index Google Scholar is 1.3 times higher than the Web of Science, and Scopus is 1.1 higher than the h index in the Web of Science. Similar results were obtained by Meho and Rogers (2008) who evaluated 22 human-computer interaction researchers from UK and found that the h index in Google Scholar is on average 1.6 times higher than Web of Science. Patel et al. (2013) compared publications and citations for 195 Nobel laureates in Physiology and Medicine using the three databases. They found no concordance between the three databases when considering the number of publications 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 PeerJ reviewing PDF | (v2013:09:794:0:1:NEW 9 Sep 2013) R ev ie w in g M an us cr ip t and citations count per laureate. However, the h index was the most reliably calculated bibliometric across the three databases. We calculated the Spearman\u2019s rank correlation (\u03c1) of the h index of the 340 researchers from the three databases. The three databases show excellent correlation for the h index, with WoS and GS as having the largest concordance. Unexpectedly, the rank correlation in terms of no. citations and no. papers (Table 3) also indicates that the three databases are comparable. This implies that the ranking of individuals within a database is comparable with the other other databases. Our correlation is also much higher compared to the 13 computer scientists studied by Franceschet (2010) who only obtained \u03c1 = 0.65. We used a much larger dataset, and the GS data came from the page that was created by the researcher, thus the listed papers and citations are assumed to be more complete. <Table 3. Somewhere here> a comparison between 5 year impact factor (if) and google scholar h5-index for 31 soil : science journals in 2012. PeerJ reviewing PDF | (v2013:09:794:0:1:NEW 9 Sep 2013) R ev ie w in g M an us cr ip t Figure 6 The relationship between cites and 5 year Impact factor (IF) and Google Scholar h5index for soil science journals in 2012. Cites is the number of citations in 2012 for papers that were published in 2007 \u2013 2011. PeerJ reviewing PDF | (v2013:09:794:0:1:NEW 9 Sep 2013) R ev ie w in g M an us cr ip t Table 1(on next page) Tables 1-6 PeerJ reviewing PDF | (v2013:09:794:0:1:NEW 9 Sep 2013) R ev ie w in g M an us cr ip t Table 1. Descriptive statistics of publication indices from Google Scholar, Scopus and Web of Science database for 340 soil researchers. Number of citations Number of papers First year of publication h index m (rate of increase per year) conclusions : From this analysis the following can be concluded: - There is a large difference between the number of citations, number of publications and the h index using the three different databases. - On average, Google Scholar gives the largest number of publications, largest number of citation and the highest h index. The Web of Science gives the lowest averages. - There are solid relationships of the h index between these three databases. - The h5-index has a high correlation with the impact factor, however it is more robust and less affected by citation manipulation. It should be considered as an alternative to the journal\u2019s impact factor. This analysis has shown that the choice of the database affects for assessing scientific impact for academic promotions, grant evaluations, job vacancy candidates or the evaluations of university departments and research centres. It is recommended that we should quote these bibliometric indices for all three databases as they reflect different types of publications. Web of Science uses refereed journal articles, Scopus includes conference proceedings and book chapters, whereas Google Scholar include other 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 PeerJ reviewing PDF | (v2013:09:794:0:1:NEW 9 Sep 2013) R ev ie w in g M an us cr ip t publications (including software). The established relationships between the databases (Table 2) can be used as pedobibliometric transfer functions by anyone interested relating databases. We are not aware whether these functions have been established for other scientific disciplines but assume they will be similar. As a test, we applied our function relating h index of WoS and GS to the 30 nursing faculty data of DeGroote and Raszewski (2012) and the function gives a good prediction with a Spearman rank correlation of 0.852. the h index of soil researchers : In an earlier paper (Minasny et al., 2007) we investigated the relationship between the h index of 228 soil researchers and found that the index was about 0.7 times the number of years since the first publication (which we called scientific age, or t). That means if a researcher has been publishing since 10 years his/her h index should be about 7. We had calculated this index using the Web of Science database, and now we have repeated this using the 340 soil researchers (which are different from the previous list) (Fig. 2). Although the data are scattered the relationship holds: h index = 0.7 x scientific age (i.e. first year of publication) (R2 = 0.72). The Web of Science database shows that the average rate of h index increase over time (m) is 0.7, with the lowest value of 0.06 and highest value of 2.9 (Table 4). The average m value for Scopus is 0.7 and for Google Scholar it is 0.8 (Table 1). 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 PeerJ reviewing PDF | (v2013:09:794:0:1:NEW 9 Sep 2013) R ev ie w in g M an us cr ip t For selected researchers, we tried to calculate the distribution of m (h index divided by the number of years since first publication) as a function of sub-disciplines in soil science. Table 5 shows the distribution of m for WoS and GS according to 6 subdisciplines. It shows that h index varies between sub-diciplines, for WoS, soil biology, biogeochemistry and ecology having the highest m values (median of 0.8). This is followed by soil physics, soil fertility and management, soil geography and pedometrics, chemistry and lastly pedology (average m = 0.5). The order in Google Scholar is slightly different, but it is consistent that soil biology has the highest m value and pedology is the lowest. Therefore within soil science, the sub-discipline also varies in terms of h index. The citation ratios are:- Soil biology, ecology and biogeochemistry, Soil management and fertility, Soil geography & pedometrics, Soil physics, Soil chemistry and mineralogy, Pedology 1: 0.9: 0.8: 0.8: 0.8: 0.6; respectively. Although the number of citations across the three databases can be quite different, the relationship between the number of citations and the h index is quite consistent across the three databases (Fig. 3): h index = \u00bd n\u00bd (R2 = 0.95) This relationship follows the function postulated by Hirsch (2005) where number of citations is about 3 to 5 times h2, and it appears the h index follows an absorption-type relationship (Warrick, 2003), increasing rapidly at low number of citations and the rate decreases with increasing number of citations. <Fig. 2. Somewhere here> <Fig. 3. Somewhere here> Table 4 shows the relationship of the average number of papers and citations per year for the 340 soil researchers. This can be interpreted as: on average, a soil researcher produces 5 articles per year, 2 articles in a journal, 1 in a conference proceeding, and 2 other unrefereed publications. The researcher receives 65 citations from journal articles, additional 13 citations from conference proceedings and another 44 citations from other publications. 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 PeerJ reviewing PDF | (v2013:09:794:0:1:NEW 9 Sep 2013) R ev ie w in g M an us cr ip t <Table 4. Somewhere here> <Table 5. Somewhere here> journal citations : We retrieved 31 Soil Science journal impact factors (IF) and other metrics from the 2012 Thompson Reuters Journal Citation Reports (JCR, released in June 2013). Google Scholar also has measures of journal\u2019s metric, h5-index, which is the h index for articles published in that journal for the last five years. The list of journals for soil science discipline in Google Scholar is slightly different from the Journal Citation Reports, and therefore we used the journals listed in Thompson Reuters Journal Citation Reports as the basis for comparison. We searched for the h5-index for the journals in Google Scholar metric for 2012 (released July 2013). Table 5 shows that Google Scholar h5-index has a better correlation with the five year IF (impact factor) than the two year IF, whereas Figure 5 shows the comparison between GS h5-index and the five year IF. While the h5-index and five year IF has a high rank correlation (\u03c1 = 0.90), the ranking is different for different journals. The journals \u2018Soil Biology and Biochemistry\u2019 and \u2018Plant and Soil\u2019 both consistently ranked no. 1 and 2 in JCR and GS while other journals appear to be slightly different in the ranking (1 to 3 places difference). The 2 journals are able to maintain high number of citations relative to the number of papers they published. <Table 6. Somewhere here> <Fig. 5. Somewhere here> There are 4 journals that are ranked much higher (>= 4 difference in rank) in Google Scholar compared to the IF: \u2018Soil Science Society of America Journal\u2019, \u2018Journal of Plant Nutrition and Soil Science\u2019, \u2018Pedosphere\u2019, and \u2018Revista Brasileira de Ci\u00eancia do Solo\u2019. All these are journals are published by national soil science societies (USA, Germany, China and Brazil). In the case of \u2018Revista Brasileira de Ci\u00eancia do Solo\u2019 which ranked 12 in GS and ranked 25 in JCR, Google scholar includes more citations from non-English articles. Contrarily, there are four journals that are ranked much lower (<= 4 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 PeerJ reviewing PDF | (v2013:09:794:0:1:NEW 9 Sep 2013) R ev ie w in g M an us cr ip t difference in rank) in Google Scholar: \u2018European Journal of Soil Science\u2019, \u2018Soil Use and Management\u2019, \u2018Journal of Soil and Water Conservation\u2019, and \u2018Soil Science\u2019. The Thompson Reuters Journal Citation Reports suffers from a miscalculation, for example \u2018Australian Journal of Soil Research\u2019 was reported to have a 2 year IF of 3.443. This is a miscalculation, as the journal changed its name to \u2018Soil Research\u2019 in 2010, and the IF calculation only accounts for papers published in 2010. \u2018Soil Research\u2019 was again listed as a separate journal in JCR. We have recalculated the actual impact factor for this journal in our analysis. While there is a positive correlation between citations (citations in 2012 to papers published from the previous 5 years) and IF, we can see that there are 2 trends (Fig. 6a). For journals that published <700 papers between 2007 and 2011 (or on average less than 140 papers per year) IF tends to increase much rapidly with increasing citations (1.2 increase in IF per 1000 citation). For the other 7 journals that published more than 700 papers, the slope is half (0.6 IF increase per 1000 citation). So there is a drawback for journals to publish more papers. Meanwhile h5-index is mostly controlled by no. citations following an absorption relationship (Fig. 6b). Although the citations come from WoS, h5-index still holds the square-root relationship supporting its robustness. Table 5 also shows that GS h5-index is more correlated to the Eigenfactor metric compared to IF. The Eigenfactor metric is based on the Google PageRank algorithm which calculated the \u201cinfluence\u201d of the journal based on the citations weighted by the \u201cquality\u201d of the journal (Bergstrom, 2007). Citation from a highly cited journal will have a higher weighting than a lower cited journal. We showed that this metric is less vulnerable to self-citation as compared to impact factor (Minasny et al., 2010). Interestingly, Google Scholar does not apply its PageRank for citations. Vanclay (2006) and Cortault and Hayek (2008) established that the h index is robust and is relatively unaffected by grey literature and errors in citations such as in Google Scholar. Most of the errors (and distortions) in citation databases are found in the \u2018long tails\u2019 of the citation distribution and they tend not to affect the h index much. For the journals considered here, we calculated the ratio between h5-index and number of papers, and it shows that only 1- 9% (median 5%) of the total papers that contributed to h5-index. And we also demonstrated that h5-index keeps its relationship even when using WoS citations. 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 PeerJ reviewing PDF | (v2013:09:794:0:1:NEW 9 Sep 2013) R ev ie w in g M an us cr ip t Harzing and van der Wal (2009) recommended the use of GS h index for Management and International Business journals. We also concur that the h5-index is a better measure of a journal\u2019s citation performance when compared to impact factor as it is more robust and less affected by citation manipulation. It is now acknowledged that there are ways of manipulating impact factor, which include self-citation, and editorials that listed references to previously published articles (Falagas and Alexiou, 2008). The h index is less sensitive to the increase in number of citations, while individual highly cited papers can artificially increases the impact factor. Although h index can also be manipulated by self-citation, in order to increase the h index considerably, a journal has to be more tactical by increasing a significant number of citations to certain papers. In addition, it only considers the top influential h papers in the journal, thus it does not penalise a journal for publishing a larger number of papers. <Fig. 6. Somewhere here> science, scopus, and google scholar : Budiman Minasny 1 *, Alfred E. Hartemink 2 , Alex. McBratney 1, Ho-Jun Jang1 1 The University of Sydney, Department of Environmental Sciences, Faculty of Agriculture & Environment, NSW 2006, Australia. 2 University of Wisconsin \u2013 Madison, Department of Soil Science, FD Hole Soils Lab, 1525 Observatory Drive. Madison, 53706 WI, USA. ( * corresponding author) E-mail budiman.minasny@sydney.edu.au 1 2 3 4 5 6 7 8 9 PeerJ reviewing PDF | (v2013:09:794:0:1:NEW 9 Sep 2013) R ev ie w in g M an us cr ip t self-citations : A way to boost the h index is by self citation. Hyland (2003) found that self-citation is 12% of all references in biology, engineering and physics, compared to 4% in sociology, philosophy, linguistics, or marketing. For all soil science journals, we found a mean of 12% self-citations but it differs between the subdisciplines (Minasny et al., 2010). High rates of self-citation were accompanied by high journal impact factor ranking; China and the USA had the highest rates of self-citation whereas Egypt, Algeria, Ukraine, and Indonesia have low levels of self-citations in soil science (Minasny et al., 2010). So high rates of self-citation may influence the h index and the Scopus database allows calculation of the h index with and without self-citation. Self-citation here is the so-called diachronous (Lawani, 1982) kind, which is self-citation from the citations received by the author. The other type is called synchronous which is more difficult to calculate, i.e. author\u2019s self-referencing relative to the total number of references quoted in a paper. The relationship for the 340 soil researchers is consistent and the average h index without self-citation is about 12% lower (Fig. 4): h index without self-citation = 0.88 \u00d7 h index (R2 = 0.97) We found a weak relationship between percentage self-citation and scientific age (t) which suggests that older and more established authors cite less of their own work (Fig. 4). It also suggests that some younger authors have very high rates of self citation as their works were not known widely, as the researcher matures the papers are more widely known and more citations were gained. Percent self-citation = 42 - 4.8 \u00d7 \u221at (R2 = 0.18) 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 PeerJ reviewing PDF | (v2013:09:794:0:1:NEW 9 Sep 2013) R ev ie w in g M an us cr ip t <Fig. 4. Somewhere here> google scholar : Minimum 16 3 1953 1 0.09 25th Quantile 266 32 1985 26 0.56 Median 866 79 1993 15 0.85 75th Quantile 2596 146 2001 26 1.18 Maximum 49447 1159 2011 115 3.67 scopus : Minimum 1 1 1955 1 0.03 25th Quantile 116 14 1989 5 0.45 Median 469 34 1996 11 0.71 75th Quantile 1361 65 2004 19 1.00 Maximum 28693 423 2011 70 2.87 web of science : Minimum 1 1 1957 1 0.06 25th Quantile 76 10 1991 5 0.41 Median 291 23 1998 10 0.67 75th Quantile 945 48 2004 17 1.00 Maximum 32837 424 2011 96 2.87 1 2 PeerJ reviewing PDF | (v2013:09:794:0:1:NEW 9 Sep 2013) R ev ie w in g M an us cr ip t Standard error of estimates R2 GS no. papers = 2.33 WoS no. papers 0.06 0.797 GS no. citations = 1.87 WoS no. citations 0.05 0.809 GS h index = 1.44 WoS h index 0.02 0.956 Scopus no. papers = 1.09 WoS no. papers 0.02 0.902 Scopus no. citations = 1.03 WoS no. citations 0.02 0.867 Scopus h index = 0.99 WoS h index 0.01 0.936 Authors who started to publish after 1995 Scopus no. papers = 1.11 WoS no. papers 0.03 0.900 Scopus no. citations = 1.17 WoS no. citations 0.02 0.949 Scopus h index = 1.11 WoS h index 0.02 0.954 3 4 PeerJ reviewing PDF | (v2013:09:794:0:1:NEW 9 Sep 2013) R ev ie w in g M an us cr ip t Variable By Variable Spearman \u03c1 Prob > |\u03c1| WoS h index GS h index 0.939 <.0001 Scopus h index GS h index 0.931 <.0001 WoS h index Scopus h index 0.922 <.0001 WoS no. citations GS no. citations 0.939 <.0001 Scopus no. citations GS no. citations 0.955 <.0001 WoS no. citations Scopus no. citations 0.945 <.0001 WoS no. papers GS no. papers 0.840 <.0001 Scopus no. papers GS no. papers 0.896 <.0001 WoS no. papers Scopus no. papers 0.905 <.0001 5 6 PeerJ reviewing PDF | (v2013:09:794:0:1:NEW 9 Sep 2013) R ev ie w in g M an us cr ip t Table 4. Comparison of the h index over time using the data from 340 soil researchers in Google Scholar (GS), Scopus and Web of Science (WoS) Standard error of estimates R2 GS h index = 0.84 \u00d7 year 0.02 0.745 WoS h index = 0.73 \u00d7 year 0.02 0.717 Scopus h index = 0.73 \u00d7 year 0.02 0.759 GS no. papers = 5.5 \u00d7 year 0.23 0.620 WoS no. papers = 2.5 \u00d7 year 0.12 0.567 Scopus no. papers = 3.0 \u00d7 year 0.12 0.656 GS no. citations = 122 \u00d7 year 9.1 0.344 WoS no. citations = 65 \u00d7 year 5.8 0.269 Scopus no.citations = 78 \u00d7 year 5.8 0.346 7 8 PeerJ reviewing PDF | (v2013:09:794:0:1:NEW 9 Sep 2013) R ev ie w in g M an us cr ip t n Min Q25 Median Q75 Max Google Scholar Pedology 25 0.09 0.37 0.56 0.83 1.67 Soil chemistry 42 0.09 0.45 0.78 1.16 2.00 Soil physics 67 0.21 0.56 0.79 1.00 2.55 Soil geography & pedometrics 28 0.16 0.58 0.87 1.02 1.93 Soil management & fertility 42 0.28 0.69 0.98 1.30 2.12 Soil biology, ecology, biogeochemistry 88 0.23 0.78 1.01 1.65 3.67 Web of Science Pedology 25 0.08 0.32 0.46 0.78 1.67 Soil chemistry 42 0.06 0.32 0.63 1.00 1.67 Soil geography & pedometrics 67 0.11 0.50 0.64 0.86 1.50 Soil management & fertility 28 0.15 0.47 0.67 1.00 1.63 Soil physics 42 0.17 0.46 0.72 1.00 2.14 Soil biology, ecology, biogeochemistry 88 0.14 0.63 0.83 1.33 2.87 9 10 11 12 13 PeerJ reviewing PDF | (v2013:09:794:0:1:NEW 9 Sep 2013) R ev ie w in g M an us cr ip t Table 6. Spearman\u2019s rank correlation coefficient (\u03c1) of the Google Scholar h5-index and impact factor (IF), no. papers, citations, and Eigenfactor metrics from Journal Citation Reports for 31 soil science journals. Cites is the number of citations in 2012 for papers that were published in 2007 \u2013 2011. Variable By Variable Spearman \u03c1 Prob > |\u03c1| h5-index Cites (5 years) 0.972 <.0001 h5-index Eigenfactor 0.970 <.0001 h5-index 5 year IF 0.903 <.0001 h5-index 2 year IF 0.870 <.0001 h5-index No. papers (5 years) 0.721 <.0001 14 15 16 17 PeerJ reviewing PDF | (v2013:09:794:0:1:NEW 9 Sep 2013) R ev ie w in g M an us cr ip t",
+    "url": "https://peerj.com/articles/184/reviews/",
+    "review_1": "Pei-Yuan Qian \u00b7 Oct 1, 2013 \u00b7 Academic Editor\nACCEPT\nI am very happy to know that you have addressed most issues raised by the reviewers and I am happy to accept this ms for publication. Thank you for choosing PeerJ as outlet of your research findings.",
+    "review_2": "Pei-Yuan Qian \u00b7 Aug 18, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nThe issues raised by both reviewers need to be carefully addressed before this ms can be accepted for publication. Please revise your ms carefully and submit your revised version, together with a detailed response to reviewers' comments in a format of point-to-point.",
+    "review_3": "Matthew Cole \u00b7 Aug 15, 2013\nBasic reporting\nOverall\n\nThis paper considers whether microplastics are ingested by barnacles that reside of floating debris in the North Pacific gyre. There are few studies of in-situ organisms ingesting microplastics so the data presented is a particularly important addition to the literature. The introduction is clear and well-informed, the methods and results clear and concise, and the discussion comprehensive. Overall I consider this a well-informed and highly interesting paper that I would be happy to recommend for publication.\nExperimental design\nMethodology\n\nLine 85-86 \u2013 Dissection procedure could do with a little more detail (e.g. how was intestine removed? Was it cut open longitudinally? Was observation for microplastic debris systematic? Was the intestine washed through or plastics picked off? What was the magnification used?). Further, were any protocols included to mitigate contamination of samples (i.e. airborne or introduced via scalpel or clothing for example)? If not, perhaps a sentence to justify why this was not required in this study?\n\nLine 92 \u2013 Does (N=30,518) refer to number of tows (as positioning of text suggests), or number of plastic particles collected?\nValidity of the findings\nResults\n\nFigure 1 \u2013 did you consider comparing the sites used with number of microplastic particles found in the seawater trawls?\n\nLine 117-119 \u2013 Looking at Figure 2b this correlation is not particularly evident; instead of looking at a correlation, I could see a histogram working much better here (similar to used in Fig2a). If you have data from the seawater trawls, would you be able to compare number of particles ingested Vs number of particles found in trawls around location the barnacle was sampled from?\n\nLine125-127 \u2013 What types of plastic dominated the seawater trawls? (i.e. were there lots of fibres or nurdles present in the seawater that the barnacles were not eating?)\n\n\n\nDiscussion\n\nIn method (Line 103) Raman laser melted some plastics. There is no follow-up on alternate method that might have been used instead?\nAdditional comments\nGeneral Comments\n\nWhile most readers will be familiar with what a barnacle looks like, \u201ccapitulum\u201d is a very specific term that will not be well known to the majority of readers. Perhaps a small figure with an image of one of your barnacle specimens with some basic annotations would be appropriate? Would also be interesting to see a photo of some of the plastics recovered.\nCite this review as\nCole M (2013) Peer Review #1 of \"Gooseneck barnacles (Lepas spp.) ingest microplastic debris in the North Pacific Subtropical Gyre (v0.1)\". PeerJ https://doi.org/10.7287/peerj.184v0.1/reviews/1",
+    "review_4": "Daniel Rittschof \u00b7 Aug 1, 2013\nBasic reporting\nfine\nExperimental design\nwell explained\nValidity of the findings\nvalid for over half the particles.\nAdditional comments\nThis is a well done offering. There are several things to consider with this paper:\nHow long do particles stay in a barnacle gut? Are you really measuring just a snapshot of the percentabge of barnacles that ate particles in the last x hours? Is ingested the best word to use, or would a phrase like found in their guts be more appropriate. It is clear that fish eat Lepas, it is very likely that there is considerable predation on Lepas in the open ocean. Do you believe for example that oceanic Lepas die of old age? Ms could be improved by addressing these issues and at least commenting on alternative methods that might be used to determine what kinds of plastic the colored particles that melt are made out of. Have the macroplastics been characterized? How might they compare.\nCite this review as\nRittschof D (2013) Peer Review #2 of \"Gooseneck barnacles (Lepas spp.) ingest microplastic debris in the North Pacific Subtropical Gyre (v0.1)\". PeerJ https://doi.org/10.7287/peerj.184v0.1/reviews/2",
+    "pdf_1": "https://peerj.com/articles/184v0.2/submission",
+    "pdf_2": "https://peerj.com/articles/184v0.1/submission",
+    "all_reviews": "Review 1: Pei-Yuan Qian \u00b7 Oct 1, 2013 \u00b7 Academic Editor\nACCEPT\nI am very happy to know that you have addressed most issues raised by the reviewers and I am happy to accept this ms for publication. Thank you for choosing PeerJ as outlet of your research findings.\nReview 2: Pei-Yuan Qian \u00b7 Aug 18, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nThe issues raised by both reviewers need to be carefully addressed before this ms can be accepted for publication. Please revise your ms carefully and submit your revised version, together with a detailed response to reviewers' comments in a format of point-to-point.\nReview 3: Matthew Cole \u00b7 Aug 15, 2013\nBasic reporting\nOverall\n\nThis paper considers whether microplastics are ingested by barnacles that reside of floating debris in the North Pacific gyre. There are few studies of in-situ organisms ingesting microplastics so the data presented is a particularly important addition to the literature. The introduction is clear and well-informed, the methods and results clear and concise, and the discussion comprehensive. Overall I consider this a well-informed and highly interesting paper that I would be happy to recommend for publication.\nExperimental design\nMethodology\n\nLine 85-86 \u2013 Dissection procedure could do with a little more detail (e.g. how was intestine removed? Was it cut open longitudinally? Was observation for microplastic debris systematic? Was the intestine washed through or plastics picked off? What was the magnification used?). Further, were any protocols included to mitigate contamination of samples (i.e. airborne or introduced via scalpel or clothing for example)? If not, perhaps a sentence to justify why this was not required in this study?\n\nLine 92 \u2013 Does (N=30,518) refer to number of tows (as positioning of text suggests), or number of plastic particles collected?\nValidity of the findings\nResults\n\nFigure 1 \u2013 did you consider comparing the sites used with number of microplastic particles found in the seawater trawls?\n\nLine 117-119 \u2013 Looking at Figure 2b this correlation is not particularly evident; instead of looking at a correlation, I could see a histogram working much better here (similar to used in Fig2a). If you have data from the seawater trawls, would you be able to compare number of particles ingested Vs number of particles found in trawls around location the barnacle was sampled from?\n\nLine125-127 \u2013 What types of plastic dominated the seawater trawls? (i.e. were there lots of fibres or nurdles present in the seawater that the barnacles were not eating?)\n\n\n\nDiscussion\n\nIn method (Line 103) Raman laser melted some plastics. There is no follow-up on alternate method that might have been used instead?\nAdditional comments\nGeneral Comments\n\nWhile most readers will be familiar with what a barnacle looks like, \u201ccapitulum\u201d is a very specific term that will not be well known to the majority of readers. Perhaps a small figure with an image of one of your barnacle specimens with some basic annotations would be appropriate? Would also be interesting to see a photo of some of the plastics recovered.\nCite this review as\nCole M (2013) Peer Review #1 of \"Gooseneck barnacles (Lepas spp.) ingest microplastic debris in the North Pacific Subtropical Gyre (v0.1)\". PeerJ https://doi.org/10.7287/peerj.184v0.1/reviews/1\nReview 4: Daniel Rittschof \u00b7 Aug 1, 2013\nBasic reporting\nfine\nExperimental design\nwell explained\nValidity of the findings\nvalid for over half the particles.\nAdditional comments\nThis is a well done offering. There are several things to consider with this paper:\nHow long do particles stay in a barnacle gut? Are you really measuring just a snapshot of the percentabge of barnacles that ate particles in the last x hours? Is ingested the best word to use, or would a phrase like found in their guts be more appropriate. It is clear that fish eat Lepas, it is very likely that there is considerable predation on Lepas in the open ocean. Do you believe for example that oceanic Lepas die of old age? Ms could be improved by addressing these issues and at least commenting on alternative methods that might be used to determine what kinds of plastic the colored particles that melt are made out of. Have the macroplastics been characterized? How might they compare.\nCite this review as\nRittschof D (2013) Peer Review #2 of \"Gooseneck barnacles (Lepas spp.) ingest microplastic debris in the North Pacific Subtropical Gyre (v0.1)\". PeerJ https://doi.org/10.7287/peerj.184v0.1/reviews/2\nReview 5: nan\nReview 6: \n",
+    "LLM_extracted_review": "['Review 1: The reviewer is happy with the revisions made and accepts the manuscript for publication.\\n\\nReview 2: The manuscript requires careful revision to address issues raised by both reviewers before it can be accepted for publication.\\n\\nReview 3: \\n- Basic reporting: The paper is a significant addition to the literature on microplastics and barnacles.\\n- Experimental design: More detail is needed on the dissection procedure and contamination protocols. Clarification is needed on the number referenced (N=30,518).\\n- Validity of findings: Suggestions for improving data presentation and comparisons with seawater trawls are provided. Questions about the types of plastics found in seawater are raised.\\n- Discussion: There is a lack of follow-up on alternative methods for analyzing plastics that melted during Raman laser analysis. A suggestion is made to include a figure for better understanding of barnacle anatomy.\\n\\nReview 4: \\n- Basic reporting: The reporting is fine.\\n- Experimental design: The methodology is well explained.\\n- Validity of findings: Validity is confirmed for over half the particles.\\n- Additional comments: Questions are raised about the duration particles remain in barnacle guts and the appropriateness of terminology. Suggestions for addressing predation and characterizing macroplastics are made.']"
+}

peerj_json_files/PeerJ_Json_135.json ADDED Viewed

	@@ -0,0 +1,16 @@

+{
+    "v1_Abstract": "Substantial quantities of small plastic particles, termed \u201cmicroplastic,\u201d have been found in many areas of the world ocean, and have accumulated in particularly high densities on the surface of the subtropical gyres. While plastic debris has been documented on the surface of the North Pacific Subtropical Gyre (NPSG) since the early 1970s, the ecological implications remain poorly understood. Organisms associated with floating objects, termed the \u201crafting assemblage,\u201d are an important component of the NPSG ecosystem. These objects are often dominated by abundant and fast-growing gooseneck barnacles (Lepas spp.), which predate on plankton and larval fishes at the sea surface. To assess the potential effects of microplastic on the rafting community, we examined the gastrointestinal tracts of 385 barnacles collected from the NPSG for evidence of plastic ingestion. We found that 33.5% of the barnacles had plastic particles present in their gastrointestinal tract, ranging from one plastic particle to a maximum of 30 particles. Particle ingestion was positively correlated to capitulum length, and no blockage of the stomach or intestines was observed. The majority of ingested plastic was polyethylene, with polypropylene and polystyrene also present. Our results suggest that barnacle ingestion of microplastic is relatively common, with unknown trophic impacts on the rafting community and the NPSG ecosystem.",
+    "v1_col_introduction": "introduction  : Oceanic litter, termed \u201cmarine debris\u201d or \u201cplastic pollution,\u201d is a matter of increasing\nscientific and public concern (STAP-GEF 2011, U.S. Environmental Protection Agency 2011,\nConvention on Biological Diversity and STAP-GEF 2012). The durability and longevity that\nmake plastic a useful substance also leads to its persistence in the marine environment, with\nconsequences that include entanglement, damage to habitats, species transport, and ingestion\n(National Research Council 2009). One study estimated that more than 267 species have been\ndocumented to ingest plastic (Allsopp et al. 2006), including mammals (Eriksson and Burton\n2003, Jacobsen et al. 2010), seabirds (Moser and Lee 1992, Ryan 2008, van Franeker et al. 2011),\nturtles (Schuyler et al. 2013), and a wide variety of fishes (Possatto et al. 2011, Lusher et al.\n2013, Anastasopoulou et al. 2013). Negative effects of plastic ingestion may include intestinal\nblockage, diminished feeding stimulus, lowered steroid hormone levels, delayed ovulation and\nreproductive failure (Azzarello and Van Vleet 1987, Derraik 2002). Because oceanic plastic\ndebris can contain high levels of hydrophobic toxins (Endo et al. 2005, Frias et al. 2010, Rios et\nal. 2010, Rochman et al. 2013), ingestion of plastic debris may also increase toxic exposure\n(Teuten et al. 2009, Gassel et al. 2013).\nMost plastic ingestion has been documented in vertebrates (Convention on Biological\nDiversity and STAP-GEF 2012), but the extent of plastic ingestion in marine invertebrates\nremains poorly known. Laboratory experiments suggest that many invertebrate species ingest plastic (reviewed in Wright et al. 2013). Suspended plastic particles (2-60 \u03bcm in diameter) were\n1\n2\n3\n4\n5\n6\n7\n8\n9\n10\n11\n12\n13\n14\n15\n16\n17\n18\n19\n20\n21\n22\n23\n24\nPeerJ reviewing PDF | (v2013:07:631:1:0:NEW 25 Sep 2013)\nR ev ie w in g M an\nus cr ip t\nsuccessfully fed to calanoid copepods, cladocerans, and salps in the context of studying particle\nsize selectivity (Burns 1968, Wilson 1973, Frost 1977, Kremer and Madin 1992). In laboratory\nstudies focused specifically on the incidence of plastic ingestion, plastic particles were readily\nconsumed by an assortment of zooplankton (Cole et al. 2013) and benthic invertebrates\n(Thompson et al. 2004, Browne et al. 2008, Graham and Thompson 2009, Wegner et al. 2012,\nvon Moos et al. 2012, Besseling et al. 2013). However, the evidence from natural ecosystems is\nfar sparser. To date, we are aware of only three studies that have found in situ plastic ingestion in\ninvertebrates: sandhopper amphipods (Talitrus saltator; Ugolini et al. 2013), Norway lobster\n(Nephrops norvegicus; Murray and Cowie 2011), and flying squid (Ommastrephes bartrami; Day\n1988 cited in Laist 1997).\nThough plastic pollution has been documented in the North Atlantic and North Pacific\nsubtropical gyres since the early 1970s (Carpenter and Smith 1972, Wong et al. 1974, Day and\nShaw 1987, Moore et al. 2001, Law et al. 2010, Goldstein et al. 2012), the ecological\nimplications have been relatively little studied. In this open ocean ecosystem, the majority of\nmarine debris are small particles (termed \u201cmicroplastic,\u201d less than 5 mm in diameter) that float at\nthe sea surface (Hidalgo-Ruz et al. 2012), though wind mixing moves some particles deeper\n(Kukulka et al. 2012). Floating plastics in these areas are primarily comprised of polyethylene,\nwith polypropylene and polystyrene also present (Rios et al. 2007, Goldstein 2012). Ingestion has\nbeen found in surface-feeding seabirds (Fry et al. 1987, Avery-Gomm et al. 2012) and epipelagic\nand mesopelagic fishes (Boerger et al. 2010, Davison and Asch 2011, Jantz et al. 2013, Choy and Drazen 2013), but the biota most likely to be impacted by microplastic pollution is the neuston, a specialized community associated with the air-sea interface which includes both zooplankton and substrate-associated rafting organisms (Cheng 1975).\nRafting organisms in the open ocean are increasingly associated with floating plastic\ndebris, which has supplemented natural substrates such as pumice and macroalgae (Thiel and\n25\n26\n27\n28\n29\n30\n31\n32\n33\n34\n35\n36\n37\n38\n39\n40\n41\n42\n43\n44\n45\n46\n47\n48\n49\nPeerJ reviewing PDF | (v2013:07:631:1:0:NEW 25 Sep 2013)\nR ev ie w in g M an\nus cr ip t\nGutow 2005a). Two species of lepadomorph barnacles (Lepas anatifera and Lepas pacifica) are widespread throughout the North Pacific Subtropical Gyre (NPSG) and frequently dominate the rafting assemblage (Tsikhon-Lukanina et al. 2001). (A third species, Lepas (Dosima) fascularis, forms its own float at the end of the juvenile stage and drifts independently, and is therefore not a major component of the rafting assemblage; Newman and Abbott 1980.) These barnacles are omnivorous, feeding opportunistically on the neustonic zooplankton, and are said to \u201chold a singular position in having more sources of food to draw upon than any other organisms in the neuston (Bieri 1966).\u201d The barnacles are themselves preyed upon by omnivorous epipelagic crabs and the rafting nudibranch Fiona pinnata (Bieri 1966, Davenport 1992).\nIn this study, we hypothesized that Lepas\u2019 indiscriminate feeding strategy and position at\nthe sea surface could cause this species to ingest microplastic, with unknown implications for NPSG ecology. To this end, we examined the gastrointestinal tracts of 385 Lepas from the NPSG for evidence of microplastic ingestion.",
+    "v2_Abstract": "Substantial quantities of small plastic particles, termed \u201cmicroplastic,\u201d have been found in many areas of the world ocean, and have accumulated in particularly high densities on the surface of the subtropical gyres. While plastic debris has been documented on the surface of the North Pacific Subtropical Gyre (NPSG) since the early 1970s, the ecological implications remain poorly understood. Organisms associated with floating objects, termed the \u201crafting assemblage,\u201d are an important component of the NPSG ecosystem. These objects are often dominated by abundant and fast-growing gooseneck barnacles (Lepas spp.), which predate on plankton and larval fishes at the sea surface. To assess the potential effects of microplastic on the rafting community, we examined the gastrointestinal tracts of 385 barnacles collected from the NPSG for evidence of plastic ingestion. We found that 33.5% of the barnacles had plastic particles present in their gastrointestinal tract, ranging from one plastic particle to a maximum of 30 particles. Particle ingestion was positively correlated to capitulum length, and no blockage of the stomach or intestines was observed. The majority of ingested plastic was polyethylene, with polypropylene and polystyrene also present. Our results suggest that barnacle ingestion of microplastic is relatively common, with unknown trophic impacts on the rafting community and the NPSG ecosystem.",
+    "v2_col_introduction": "introduction  : Oceanic litter, termed \u201cmarine debris\u201d or \u201cplastic pollution,\u201d is a matter of increasing\nscientific and public concern (STAP-GEF 2011, U.S. Environmental Protection Agency 2011,\nConvention on Biological Diversity and STAP-GEF 2012). The durability and longevity that\nmake plastic a useful substance also leads to its persistence in the marine environment, with\nconsequences that include entanglement, damage to habitats, species transport, and ingestion\n(National Research Council 2009). One study estimated that more than 267 species have been\ndocumented to ingest plastic (Allsopp et al. 2006), including mammals (Eriksson and Burton\n2003, Jacobsen et al. 2010), seabirds (Moser and Lee 1992, Ryan 2008, van Franeker et al. 2011),\nturtles (Bjorndal et al. 1994, Bugoni et al. 2001), and a wide variety of fishes (Possatto et al.\n2011, Lusher et al. 2013, Anastasopoulou et al. 2013). Negative effects of plastic ingestion may\ninclude intestinal blockage, diminished feeding stimulus, lowered steroid hormone levels,\ndelayed ovulation and reproductive failure (Azzarello and Van Vleet 1987, Derraik 2002).\nBecause oceanic plastic debris can contain high levels of hydrophobic toxins (Endo et al. 2005,\nFrias et al. 2010, Rios et al. 2010, Rochman et al. 2013), ingestion of plastic debris may also\nincrease toxic exposure (Teuten et al. 2009, Gassel et al. 2013).\nMost plastic ingestion has been documented in vertebrates (Convention on Biological\nDiversity and STAP-GEF 2012), but the extent of plastic ingestion in marine invertebrates\nremains poorly known. Laboratory experiments suggest that many invertebrate species ingest\n1\n2\n3\n4\n5\n6\n7\n8\n9\n10\n11\n12\n13\n14\n15\n16\n17\n18\n19\n20\n21\n22\n23\nPeerJ reviewing PDF | (v2013:07:631:0:1:NEW 30 Jul 2013)\nR ev ie w in g M an\nus cr ip t\nplastic (reviewed in Wright et al. 2013). Suspended plastic particles (2-60 \u03bcm in diameter) were\nsuccessfully fed to calanoid copepods, cladocerans, and salps in the context of studying particle\nsize selectivity (Burns 1968, Wilson 1973, Frost 1977, Kremer and Madin 1992). In laboratory\nstudies focused specifically on the incidence of plastic ingestion, plastic particles were readily\nconsumed by an assortment of zooplankton (Cole et al. 2013) and benthic invertebrates\n(Thompson et al. 2004, Browne et al. 2008, Graham and Thompson 2009, Wegner et al. 2012,\nvon Moos et al. 2012, Besseling et al. 2013). However, the evidence from natural ecosystems is\nfar sparser. To date, we are aware of only three studies that have found in situ plastic ingestion in\ninvertebrates: sandhopper amphipods (Talitrus saltator; Ugolini et al. 2013), Norway lobster\n(Nephrops norvegicus; Murray and Cowie 2011), and flying squid (Ommastrephes bartrami; Day\n1988 cited in Laist 1997).\nThough plastic pollution has been documented in the North Atlantic and North Pacific\nsubtropical gyres since the early 1970s (Carpenter and Smith 1972, Wong et al. 1974, Day and\nShaw 1987, Moore et al. 2001, Law et al. 2010, Goldstein et al. 2012a), the ecological\nimplications have been relatively little studied. In this open ocean ecosystem, the majority of\nmarine debris are small particles (termed \u201cmicroplastic,\u201d less than 5 mm in diameter) that float at\nthe sea surface (Goldstein et al. 2012b), though wind mixing moves some particles deeper\n(Kukulka et al. 2012). Floating plastics in these areas are primarily comprised of polyethylene,\nwith polypropylene and polystyrene also present (Rios et al. 2007, Goldstein 2012). Ingestion has\nbeen found in surface-feeding seabirds (Fry et al. 1987, Avery-Gomm et al. 2012) and epipelagic\nand mesopelagic fishes (Boerger et al. 2010, Davison and Asch 2011, Jantz et al. 2013, Choy and\nDrazen 2013), but the biota most likely to be impacted by microplastic pollution is the neuston, a\nspecialized community associated with the air-sea interface which includes both zooplankton and substrate-associated rafting organisms (Cheng 1975).\n24\n25\n26\n27\n28\n29\n30\n31\n32\n33\n34\n35\n36\n37\n38\n39\n40\n41\n42\n43\n44\n45\n46\n47\nPeerJ reviewing PDF | (v2013:07:631:0:1:NEW 30 Jul 2013)\nR ev ie w in g M an\nus cr ip t\nRafting organisms in the open ocean are increasingly associated with floating plastic\ndebris, which has supplemented natural substrates such as pumice and macroalgae (Thiel and\nGutow 2005a). Two species of lepadomorph barnacles (Lepas anatifera and Lepas pacifica) are\nwidespread throughout the North Pacific Subtropical Gyre (NPSG) and frequently dominate the rafting assemblage (Tsikhon-Lukanina et al. 2001). (A third species, Lepas (Dosima) fascularis, forms its own float at the end of the juvenile stage and drifts independently, and is therefore not a major component of the rafting assemblage; Newman and Abbott 1980.) These barnacles are omnivorous, feeding opportunistically on the neustonic zooplankton, and are said to \u201chold a singular position in having more sources of food to draw upon than any other organisms in the neuston (Bieri 1966).\u201d The barnacles are themselves preyed upon by omnivorous epipelagic crabs (Planes spp.) and the rafting nudibranch Fiona pinnata (Bieri 1966, Davenport 1992).\nIn this study, we hypothesized that Lepas\u2019 indiscriminate feeding strategy and position at\nthe sea surface could cause this species to ingest microplastic, with unknown implications for NPSG ecology. To this end, we examined the gastrointestinal tracts of 385 Lepas from the NPSG for evidence of microplastic ingestion.",
+    "v1_text": "results : Of the 385 barnacles examined, 129 individuals (33.5%) had ingested plastic (Fig 2, Table 1). These included 243 Lepas anatifera and 85 Lepas pacifica (57 barnacles could not be 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 PeerJ reviewing PDF | (v2013:07:631:1:0:NEW 25 Sep 2013) R ev ie w in g M an us cr ip t identified to species), of which 90 L. anatifera, 34 L. pacifica, and 5 Lepas spp. contained plastic. Forty-one of the barnacles that ingested plastic had one plastic particle in their stomach or intestines, 26 individuals had two particles, and 57 individuals contained three or more particles, to a maximum of 30 particles (Fig 3A). Overall, the number of ingested particles was positively correlated to capitulum length (Kendall\u2019s tau=0.099, p=0.015). However, when we considered only barnacles that had ingested plastic, the correlation between plastic ingestion and capitulum length was not significant (Kendall\u2019s tau=-0.080, p=0.229). Individuals with a capitulum length between 2 and 3 cm consumed the greatest number plastic particles (Fig 3B). With the exception of one individual, all the barnacles that consumed plastic had a capitulum length of 1.7 cm or greater. In total, 518 plastic particles were recovered from barnacle digestive tracts. Of these, 99% were degraded fragments and 1% were monofilament line. None of the pre-production pellets known as \u201cnurdles\u201d were found. The median diameter of ingested particles was 1.41 mm, and the median surface area 1.00 mm2, smaller than the median diameter of 1.78 mm and median surface area of 1.27 mm2 for all particles collected in nets during 2009 (Fig 4, KolmogorovSmirnov test p<0.001). The smallest particle ingested by barnacles had a maximum diameter of 0.609 mm and the largest (a long thin fragment) a maximum diameter of 6.770 mm. No blockage of the stomach or intestine was observed, and particles did not accumulate in any area of the digestive tract. All particles were of a plausible size to pass through the anus. Of the randomly selected subset of 219 ingested plastic particles that were analyzed for plastic type, 58.4% were polyethylene, 5.0 % were polypropylene, and 1.4% were polystyrene. As noted in the Methods section, we were unable to identify 35% of the subset due to darker pigmentation in these particles, which caused melting under the Raman spectrophotometer. Of the 29 barnacles that had ingested more than one piece of plastic, 66% contained more than one type of plastic. The plastic types of 12 floating debris items to which barnacles were attached 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 PeerJ reviewing PDF | (v2013:07:631:1:0:NEW 25 Sep 2013) R ev ie w in g M an us cr ip t were more diverse than those of ingested particles. Four substrates were polystyrene, 3 were polyethylene, 2 were polypropylene, 2 were polyethylene terephthalate, and one was tire rubber. acknowledgements : We thank the captain, crews, and students of the SEAPLEX cruise on the R/V New Horizon and Sea Education Association cruises S-242 and S-243 on the SSV Robert C. Seamans. Assistance from L. Sala, M.D. Ohman, and E. Zettler made this project possible. methods : Floating debris items with attached gooseneck barnacles (Fig 1A) were opportunistically collected during the 2009 Scripps Environmental Accumulation of Plastic Expedition (SEAPLEX) and two 2012 Sea Education Association (SEA) research cruises onboard the SSV Robert C. Seamans: S242, an undergraduate voyage from Honolulu, HI to San Francisco, CA (mid-June to mid-July 2012), and S243, the Plastics at SEA: North Pacific Expedition from San Diego, CA to Honolulu, HI (early October to mid-November 2012; Fig 1A). Collection occurred by several means, including 1) from the vessel using a long-handled dip net (335 \u03bcm mesh, 0.5m diameter mouth); 2) incidentally during neuston net (335 \u03bcm mesh, 0.5 x 1.0m mouth) tows at the air-sea interface; and 3) from small boat surveys within 0.5 km of each vessel when sea conditions were calm. No specific permissions were required for these samples, since they were 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 PeerJ reviewing PDF | (v2013:07:631:1:0:NEW 25 Sep 2013) R ev ie w in g M an us cr ip t taken in international waters and did not involve protected species. Seven debris items were sampled on SEAPLEX and 29 by SEA (5 during S242 and 24 on S243). Stations within 8.5 km of each other were combined for a total of 19 sampling locations within in the northeastern Pacific Ocean (Fig 2, Table 1). During SEAPLEX, the entire piece of debris with attached barnacles was preserved in 5% Formalin buffered with sodium borate. When the item was too large to be preserved (e.g., a fishing buoy), barnacles were removed and preserved separately. On SEA cruises, as many barnacles as possible to a maximum of 50 were removed from the debris object and preserved in 10% ethanol. Where feasible, a fragment of the item itself was also sampled. In the laboratory, capitulum length was measured using a ruler and species identification (L. anatifera or L. pacifica) determined for all intact individuals (Fig 1B). Barnacles less than 0.8 cm were present, but not sampled in this study. Barnacles greater than approximately 0.8 cm in length were dissected and the contents of their stomach and intestinal tract examined under a dissecting microscope (6-25x magnification as needed). Barnacles were cut open with a scalpel, and the intestinal tract removed and placed in a separate section of the petri dish. The intestinal tract was opened lengthwise, and the contents examined systematically both visually and with forceps. To avoid cross-contamination between samples, each barnacle was dissected in unique, clean petri dish and the scalpel was thoroughly rinsed with deionized water between each samples. Only microplastic fragments and monofilament that were clearly present inside the intestine were considered. Fine microfibers were discounted, as they could not be distinguished from airborne contamination. Because the vast majority of microplastic found were relatively large degraded fragments (>0.5 mm in diameter), visual examination was sufficient to confirm that the microplastic was present in the intestine, and not a result of contamination (Fig 1C). Plastic particles found in the stomach or intestine were quantified, photographed digitally against a ruler for size assessment, rinsed with fresh water and stored in a glass vial for later 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 PeerJ reviewing PDF | (v2013:07:631:1:0:NEW 25 Sep 2013) R ev ie w in g M an us cr ip t analyses. The maximum diameter (feret diameter) and two-dimensional area of each particle were digitally measured with the software package NIH ImageJ (Rasband 2012). On the SEAPLEX cruise in 2009, we also measured the diameter and area of all plastic particles captured in surfacetowed plankton nets (number of particles=30,518) using NIH ImageJ-based tools in the Zooprocess software calibrated against manual measurements (Gilfillan et al. 2009, Gorsky et al. 2010). We identified the type of plastic recovered from a randomly selected subset of barnacles (Barnacles N=42; particles N=219). A Raman spectrometer (PeakSeeker Pro-785 with AmScope operated at 10 \u2013 50 mW and 5 \u2013 20 second integration time; Raman Systems MII, Inc/Agiltron, Inc., Woburn, MA) and associated RSIQ software were used to identify plastic type. The Raman spectrum for each plastic piece was compared to a reference library of known plastic types for identification. Particles of clear, white, gray and pale-colored (light blues and greens, oranges and yellows) plastics yielded high quality Raman spectra and were readily identifiable. Those that were darker (medium to dark blues, reds and greens as well as black; 35% of particles subjected to Raman spectroscopy) were heated by the laser beam and melted even at the lowest possible power and integration time settings, resulting in no usable spectra. We also identified a subset of the debris items to which the barnacles were attached. Fragments of 18 objects were collected for analysis, but 6 could not be identified due to darker pigmentation due to melting under the laser. Statistics and figures were generated with the R statistical environment, version R-2.15.1 (R Development Core Team 2012) and QuantumGIS, version 1.8.0-Lisboa (QGIS Development Team 2013). discussion : Our results show that 33.5% of lepadid barnacles collected from the NPSG ingested microplastic, and that the sizes and types of ingested particles were approximately representative of microplastic found on the NPSG surface. Plastic ingestion in these barnacles may therefore be explained by non-selective suspension feeding while exposed to high concentrations of microplastic. The percentage of barnacles observed with ingested plastic in this study is higher than the 9.2% found in NPSG micronektonic fishes (Davison and Asch 2011) and the 19-24.5% found in larger mesopelagic fishes (Jantz et al. 2013, Choy and Drazen 2013). It is likely that barnacles encounter microplastic more frequently than vertically migrating mesopelagic fishes due to the barnacles\u2019 consistent location at the air-sea interface. Since barnacles probably clear their guts in a matter of hours (Ritz 2008), it is likely that a higher percentage of the barnacle population interacts with microplastic than is presented here. Unfortunately, due to logistical considerations on both cruises, barnacle samples were not usually concurrent with neuston tows. Since neustonic microplastic is highly spatially heterogeneous (Ryan et al. 2009), we are thus unable directly compare neustonic microplastic concentrations with incidence of barnacle ingestion. However, the sampling area is known to contain high concentrations of neustonic microplastic (Moore et al. 2001, Goldstein et al. 2012). The objects to which the barnacles are attached may also shed microplastic particles, increasing the likelihood of those particles being ingested by the local rafting community. However, the microplastics ingested by individual barnacles in this study were of multiple plastic types and colors, suggesting they are taking in particles from the surrounding water rather than 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 PeerJ reviewing PDF | (v2013:07:631:1:0:NEW 25 Sep 2013) R ev ie w in g M an us cr ip t solely from the substrate to which they are attached. Lepadid barnacles are known to be very nonselective feeders. For example, L.anatifera opportunistically ingests a wide variety of zooplankton and even fills its gut with sand when stranded on the beach (Howard and Scott 1959). L. anatifera can also readily consume large prey items up to 5 mm in diameter, larger than the majority of microplastic debris (Patel 1959). Less is known about the feeding habits of L. pacifica, but it is presumed to have a similar feeding ecology as L. anatifera and other lepadid barnacles (Crisp and Southward 1961, Cheng and Lewin 1976). To avoid difficulties in identifying plastic with darker pigmentation, future studies might consider supplementing Raman spectroscopy with density analysis (Moret-Ferguson et al. 2010), or utilizing Fourier transform infrared spectroscopy when available (Rios et al. 2007, Goldstein 2012). We found only one barnacle with a capitulum length of less than 1.7 cm that had ingested plastic. This observation implies that barnacles may need to reach a certain size before plastic ingestion is possible, perhaps due to the size of the cirri or oral opening. However, our study used visual methods to identify microplastic in barnacle gut contents, and spectroscopic methods or chemical digestion of the tissue are needed to positively identify plastic particles smaller than approximately 300 \u03bcm (Claessens et al. 2011, Hidalgo-Ruz et al. 2012). It is therefore possible that plastic ingestion in the smaller barnacles was not detected in this study. Assessing the ecological significance of plastic ingestion in pelagic invertebrates and fishes remains a challenge. Even in relatively well-studied species, it has been difficult to link plastic ingestion to mortality. For example, two studies of Laysan and black-footed albatross chicks did not find a linkage between cause of death and plastic ingestion (Sileo et al. 1990, Sievert and Sileo 1993), though a third study linked plastic ingestion with lower body weight in adult birds (Spear et al. 1995). Most studies on invertebrates have been relatively short-term investigations that have not found acute negative effects (Thompson et al. 2004, Browne et al. 2008, Graham and Thompson 2009), with the exception of an inflammatory immune response in 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 PeerJ reviewing PDF | (v2013:07:631:1:0:NEW 25 Sep 2013) R ev ie w in g M an us cr ip t mussels (von Moos et al. 2012). In zooplankton, the presence of non-edible particles can reduce the rate of feeding on edible particles (Huntley et al. 1983, Ayukai 1987, Cole et al. 2013), and physical interference with sensory apparatus may occur in very high-plastic environments (Cole et al. 2013). The lepadid barnacles in this study did not show evidence of acute harm (e.g., intestinal blockage or ulceration), though negative long-term effects cannot be ruled out. Plastic ingestion may also lead to increased body loads of persistent organic pollutants in both vertebrates and invertebrates (Teuten et al. 2009, Yamashita et al. 2011, Gassel et al. 2013, Besseling et al. 2013), but it is not known whether this occurs in barnacles, or has population- level ramifications in any taxa (Gouin et al. 2011). For example, a modeling study based on lugworms (Arenicola marina) did not find a significant toxicological risk from plastic-adsorbed pollutants (Koelmans et al. 2013). Because L.anatifera appear to survive well in the laboratory (Patel 1959), more detailed studies may be possible. If barnacles are an important prey item, it is possible that their ingestion of plastic particles could transfer plastic or pollutants through the food web. Plastic particles found in fur seals (Eriksson and Burton 2003), piscivorous fishes (Davison and Asch 2011), and crabs (Farrell and Nelson 2013) have been linked to consumption of contaminated prey. The only documented predator of rafting Lepas spp. is the nudibranch Fiona pinnata (Bieri 1966), though it is probable that omnivorous rafting crabs also consume barnacles to some extent (Davenport 1992, Frick et al. 2011). Relatively low rates of predation on these barnacles may explain Lepas\u2019 place as one of the most abundant members of the North Pacific subtropical rafting community (Newman and Abbott 1980, Thiel and Gutow 2005b). For example, one study found that L. pacifica was excluded from nearshore kelp forests by the fish Oxyjulis californica, but was able to inhabit floating kelp paddies in high densities when O. californica was absent (Bernstein and Jung 1979). Studies of the diets of fishes associated with Fish Aggregating Devices (FADs) have found that fishes associated with floating objects rarely feed directly on the fouling community (Ibrahim et 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 PeerJ reviewing PDF | (v2013:07:631:1:0:NEW 25 Sep 2013) R ev ie w in g M an us cr ip t al. 1996, Nelson 2003, Vassilopoulou et al. 2004). The likelihood of predators ingesting plastic by feeding on barnacles may therefore be relatively low. While plastic ingestion in taxa such as sea turtles (Schuyler et al. 2013) and cetaceans (Jacobsen et al. 2010) is clearly detrimental, the implications of widespread plastic ingestion in Lepas remain uncertain. Since little is known about the trophic structure and connectivity of both the rafting and drifting components of the neuston, additional studies are necessary to determine the impacts of microplastic ingestion on the rafting community and the larger pelagic ecosystem. a) a dense aggregation of lepas spp. barnacles growing on a buoy and attached line, : collected in October 2012. B) Basic anatomy of Lepas denoting the capitulum, which includes the body and its enclosing plates, and the peduncle, the muscular stalk that attaches the barnacle to the substrate. C) Microplastic ingested by an individual barnacle. PeerJ reviewing PDF | (v2013:07:631:1:0:NEW 25 Sep 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:07:631:1:0:NEW 25 Sep 2013) R ev ie w in g M an us cr ip t Figure 2 Ingestion of microplastic by barnacles across the study area. Circles indicate sampling stations and dark fill indicates the proportion of barnacles that had ingested microplastic at each site. Station coordinates, sample sizes, and ingestion proportions are given in Table 1. PeerJ reviewing PDF | (v2013:07:631:1:0:NEW 25 Sep 2013) R ev ie w in g M an us cr ip t Figure 3 Number of microplastic particles ingested by barnacles. (N=385). B) Frequency distribution of ingestion by capitulum length (N=369; sample size is smaller than above since capitulum length was not measured for 16 barnacles). Black bars are the number of individual barnacles that ingested plastic and grey bars are the number of individual barnacles that did not ingest plastic. Bins of capitulum length are greater than the first value, and less than or equal to the second value (e.g., >0.5 cm and <=1.0 cm). Percentages of ingestion by size class are as follows: 6.7%, 0, 23.2%, 43.9%, 45.2%, 35.3%, 25.0%, 40.0%, 0. PeerJ reviewing PDF | (v2013:07:631:1:0:NEW 25 Sep 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:07:631:1:0:NEW 25 Sep 2013) R ev ie w in g M an us cr ip t Figure 4 Size of microplastic particles ingested by barnacles. Size\u2013frequency distributions for A) maximum diameter and B) two-dimensional surface area of particles ingested by barnacles (black; N=507) compared to of all microplastic particles collected in 2009 (grey; N=30,518). Note: 518 microplastic particles were recovered from barnacles, but 11 were lost before they could be photographed for this analysis. PeerJ reviewing PDF | (v2013:07:631:1:0:NEW 25 Sep 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:07:631:1:0:NEW 25 Sep 2013) R ev ie w in g M an us cr ip t Table 1(on next page) Station locations and proportion of microplastic ingestion. PeerJ reviewing PDF | (v2013:07:631:1:0:NEW 25 Sep 2013) R ev ie w in g M an us cr ip t Station ID Date of collection Latitude (\u00b0N) Longitude (\u00b0W) Total no. barnacles Proportion with plastic Proportion without plastic S242-021-DN 1-Jul-12 36.135 154.957 5 0.80 0.20 S242-023-DN 2-Jul-12 37.672 152.163 20 0.00 1.00 S243-083-DN 31-Oct-12 27.000 146.782 10 0.00 1.00 S243-069-DN 27-Oct-12 30.057 145.057 15 0.47 0.53 S243-055-057058-DN 24-Oct-12 30.140 141.220 80 0.68 0.33 S243-051-052DN 23-Oct-12 30.230 140.690 34 0.18 0.82 U39.F32 15-Aug-09 34.076 140.474 53 0.40 0.60 S243-046-DN 22-Oct-12 31.330 140.338 52 0.42 0.58 S3.F6 10-Aug-09 32.911 140.320 2 0.00 1.00 S242-031-NT 6-Jul-12 39.178 140.160 12 0.00 1.00 S4.F30-F26 14-Aug-09 34.090 139.870 9 0.33 0.67 S242-032-DN 6-Jul-12 39.270 139.570 10 0.10 0.90 S2.F22U40.F11 9-Aug-09 32.050 137.928 15 0.07 0.93 F13 9-Aug-09 32.075 137.223 1 1.00 0.00 S243-032-DN 16-Oct-12 33.563 135.432 17 0.59 0.41 S242-035-DN 8-Jul-12 39.717 135.325 10 0.00 1.00 S243-025-027DN 14-Oct-12 33.700 133.460 13 0.00 1.00 S243-023-DN 13-Oct-12 33.051 132.445 14 0.00 1.00 S243-011-DN 9-Oct-12 33.493 127.715 13 0.00 1.00 PeerJ reviewing PDF | (v2013:07:631:1:0:NEW 25 Sep 2013) R ev ie w in g M an us cr ip t barnacles and ingestion microplastic. :",
+    "v2_text": "results : Of the 385 barnacles examined, 129 individuals (33.5%) had ingested plastic (Fig 1). These included 243 Lepas anatifera and 85 Lepas pacifica (57 barnacles could not be identified to species), of which 90 L. anatifera, 34 L. pacifica, and 5 Lepas spp. contained plastic. Forty-one of the barnacles that ingested plastic had one plastic particle in their stomach or intestines, 26 individuals had two particles, and 57 individuals contained three or more particles, to a maximum of 30 particles (Fig 2a). Overall, the number of ingested particles was positively correlated to capitulum length (Fig 2b, Kendall\u2019s tau=0.099, p=0.015). However, when we considered only barnacles that had ingested plastic, the correlation between plastic ingestion and capitulum length was not 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 PeerJ reviewing PDF | (v2013:07:631:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t significant (Kendall\u2019s tau=-0.080, p=0.229). Individuals with a capitulum length between 2 and 3 cm consumed the greatest number plastic particles. With the exception of one individual, all the barnacles that consumed plastic had a capitulum length of 1.7 cm or greater. In total, 518 plastic particles were recovered from barnacle digestive tracts. Of these, 99% were degraded fragments and 1% were monofilament line. None of the pre-production pellets known as \u201cnurdles\u201d were found. The median diameter of ingested particles was 1.41 mm, and the median surface area 1.00 mm2, smaller than the median diameter of 1.78 mm and median surface area of 1.27 mm2 for all particles collected in nets during 2009 (Fig 3, Kolmogorov-Smirnov test p<0.001). The smallest particle ingested by barnacles had a maximum diameter of 0.609 mm and the largest (a long thin fragment) a maximum diameter of 6.770 mm. No blockage of the stomach or intestine was observed, and particles did not accumulate in any area of the digestive tract. All particles were of a plausible size to pass through the anus. Of the randomly selected subset of 219 ingested plastic particles that were analyzed for plastic type, 58.4% were polyethylene, 5.0 % were polypropylene, and 1.4% were polystyrene. As noted in the Methods section, we were unable to identify 35% of the subset due to darker pigmentation in these particles, which caused melting under the Raman spectrophotometer. Of the 29 barnacles that had ingested more than one piece of plastic, 66% contained more than one type of plastic. The plastic types of 12 floating debris items to which barnacles were attached were more diverse than those of ingested particles. Four substrates were polystyrene, 3 were polyethylene, 2 were polypropylene, 2 were polyethylene terephthalate, and one was tire rubber. acknowledgements : We thank the captain, crews, and students of the SEAPLEX cruise on the R/V New Horizon and Sea Education Association cruises S-242 and S-243 on the SSV Robert C. Seamans. Assistance from L. Sala, M.D. Ohman, and E. Zettler made this project possible. discussion : Our results show that 33.5% of lepadid barnacles collected from the NPSG ingested microplastic, and that the sizes and types of ingested particles were approximately representative of microplastic found on the NPSG surface. Plastic ingestion in these barnacles may therefore be 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 PeerJ reviewing PDF | (v2013:07:631:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t explained by non-selective suspension feeding while exposed to high concentrations of microplastic. The percentage of barnacles observed with ingested plastic in this study is higher than the 9.2% found in NPSG micronektonic fishes (Davison and Asch 2011) and the 19-24.5% found in larger mesopelagic fishes (Jantz et al. 2013, Choy and Drazen 2013). It is likely that barnacles encounter microplastic more frequently than vertically migrating mesopelagic fishes due to the barnacles\u2019 consistent location at the air-sea interface. The objects to which the barnacles are attached may also shed microplastic particles, increasing the likelihood of those particles being ingested by the local rafting community. However, the microplastics ingested by individual barnacles in this study were of multiple plastic types and colors, suggesting they are taking in particles from the surrounding water rather than solely from the substrate to which they are attached. Lepadid barnacles are known to be very nonselective feeders. For example, L.anatifera opportunistically ingests a wide variety of zooplankton and even fills its gut with sand when stranded on the beach (Howard and Scott 1959). L. anatifera can also readily consume large prey items up to 5 mm in diameter, larger than the majority of microplastic debris (Patel 1959). Less is known about the feeding habits of L. pacifica, but it is presumed to have a similar feeding ecology as L. anatifera and other lepadid barnacles (Crisp and Southward 1961, Cheng and Lewin 1976). We found only one barnacle with a capitulum length of less than 1.7 cm that had ingested plastic. This observation implies that barnacles may need to reach a certain size before plastic ingestion is possible, perhaps due to the size of the cirri or oral opening. However, our study used visual methods to identify microplastic in barnacle gut contents, and spectroscopic methods or chemical digestion of the tissue are needed to positively identify plastic particles smaller than approximately 300 \u03bcm (Claessens et al. 2011, Hidalgo-Ruz et al. 2012). It is therefore possible that plastic ingestion in the smaller barnacles was not detected in this study. 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 PeerJ reviewing PDF | (v2013:07:631:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t Assessing the ecological significance of plastic ingestion in pelagic invertebrates and fishes remains a challenge. Even in relatively well-studied species, it has been difficult to link plastic ingestion to mortality. For example, two studies of Laysan and black-footed albatross chicks did not find a linkage between cause of death and plastic ingestion (Sileo et al. 1990, Sievert and Sileo 1993), though a third study linked plastic ingestion with lower body weight in adult birds (Spear et al. 1995). Most studies on invertebrates have been relatively short-term investigations that have not found acute negative effects (Thompson et al. 2004, Browne et al. 2008, Graham and Thompson 2009), with the exception of an inflammatory immune response in mussels (von Moos et al. 2012). In zooplankton, the presence of non-edible particles can reduce the rate of feeding on edible particles (Huntley et al. 1983, Ayukai 1987, Cole et al. 2013), and physical interference with sensory apparatus may occur in very high-plastic environments (Cole et al. 2013). The lepadid barnacles in this study did not show evidence of acute harm (e.g., intestinal blockage or ulceration), though negative long-term effects cannot be ruled out. Plastic ingestion may also lead to increased body loads of persistent organic pollutants in both vertebrates and invertebrates (Teuten et al. 2009, Yamashita et al. 2011, Gassel et al. 2013, Besseling et al. 2013), but it is not known whether this occurs in barnacles, or has population-level ramifications in any taxa (Gouin et al. 2011). For example, a modeling study based on lugworms (Arenicola marina) did not find a significant toxicological risk from plastic-adsorbed pollutants (Koelmans et al. 2013). Because L.anatifera appear to survive well in the laboratory (Patel 1959), more detailed studies may be possible. If barnacles are an important prey item, it is possible that their ingestion of plastic particles could transfer plastic or pollutants through the food web. Plastic particles found in fur seals (Eriksson and Burton 2003), piscivorous fishes (Davison and Asch 2011), and crabs (Farrell and Nelson 2013) have been linked to consumption of contaminated prey. The only documented 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 PeerJ reviewing PDF | (v2013:07:631:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t predator of rafting Lepas spp. is the nudibranch Fiona pinnata, (Bieri 1966), though it is probable that omnivorous rafting crabs also consume barnacles to some extent (Davenport 1992, Frick et al. 2011). Relatively low rates of predation on these barnacles may explain Lepas\u2019 place as one of the most abundant members of the North Pacific subtropical rafting community (Newman and Abbott 1980, Thiel and Gutow 2005b). For example, one study found that L. pacifica was excluded from nearshore kelp forests by the fish Oxyjulis californica, but was able to inhabit floating kelp paddies in high densities when O. californica was absent (Bernstein and Jung 1979). The likelihood of predators ingesting plastic by feeding on barnacles may therefore be relatively low. While plastic ingestion in taxa such as sea turtles (Bugoni et al. 2001) and cetaceans (Jacobsen et al. 2010) is clearly detrimental, the implications of widespread plastic ingestion in Lepas remain uncertain. Since little is known about the trophic structure and connectivity of both the rafting and drifting components of the neuston, additional studies are necessary to determine the impacts of microplastic ingestion on the rafting community and the larger pelagic ecosystem. methods : Floating debris items with attached gooseneck barnacles were opportunistically collected during the 2009 Scripps Environmental Accumulation of Plastic Expedition (SEAPLEX) and two 2012 Sea Education Association (SEA) research cruises onboard the SSV Robert C. Seamans: S242, an undergraduate voyage from Honolulu, HI to San Francisco, CA (mid-June to mid-July 2012), and S243, the Plastics at SEA: North Pacific Expedition from San Diego, CA to Honolulu, HI (early October to mid-November 2012). Collection occurred by several means, including 1) from the vessel using a long-handled dip net (335 \u03bcm mesh, 0.5m diameter mouth); 2) incidentally during neuston net (335 \u03bcm mesh, 0.5 x 1.0m mouth) tows at the air-sea interface; 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 PeerJ reviewing PDF | (v2013:07:631:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t and 3) from small boat surveys within 0.5 km of each vessel when sea conditions were calm. No specific permissions were required for these samples, since they were taken in international waters and did not involve protected species. Seven debris items were sampled on SEAPLEX and 29 by SEA (5 during S242 and 24 on S243). Stations within 8.5 km of each other were combined for a total of 19 sampling locations within in the northeastern Pacific Ocean (Fig 1, Table 1). During SEAPLEX, the entire piece of debris with attached barnacles was preserved in 5% Formalin buffered with sodium borate. When the item was too large to be preserved (e.g., a fishing buoy), barnacles were removed and preserved separately. On SEA cruises, as many barnacles as possible to a maximum of 50 were removed from the debris object and preserved in 10% ethanol. Where feasible, a fragment of the item itself was also sampled. In the laboratory, capitulum length was measured using a ruler and species identification (L. anitifera or L. pacifica) determined for all intact individuals. Barnacles greater than approximately 0.8 cm in length were dissected and the contents of their stomach and intestinal tract examined under a dissecting microscope. Barnacles less than 0.8 cm were present, but not sampled in this study. Plastic particles found in the stomach or intestine were quantified, photographed digitally against a ruler for size assessment, rinsed with fresh water and stored in a glass vial for later analyses. The maximum diameter (feret diameter) and two-dimensional area of each particle were digitally measured with the software package NIH ImageJ (Rasband 2012). On the SEAPLEX cruise in 2009, we also measured the diameter and area of all plastic particles captured in surface-towed plankton nets (N=30,518) using NIH ImageJ-based tools in the Zooprocess software, and calibrated against manual measurements (Gilfillan et al. 2009, Gorsky et al. 2010, Goldstein et al. 2012b). We identified the type of plastic recovered from a randomly selected subset of barnacles (Barnacles N=42; particles N=219). A Raman spectrometer (PeakSeeker Pro-785 with AmScope 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 PeerJ reviewing PDF | (v2013:07:631:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t operated at 10 \u2013 50 mW and 5 \u2013 20 second integration time; Raman Systems MII, Inc/Agiltron, Inc., Woburn, MA) and associated RSIQ software were used to identify plastic type. The Raman spectrum for each plastic piece was compared to a reference library of known plastic types for identification. Particles of clear, white, gray and pale-colored (light blues and greens, oranges and yellows) plastics yielded high quality Raman spectra and were readily identifiable. Those that were darker (medium to dark blues, reds and greens as well as black; 35% of particles subjected to Raman spectroscopy) were heated by the laser beam and melted even at the lowest possible power and integration time settings, resulting in no usable spectra. We also identified a subset of the debris items to which the barnacles were attached. Fragments of 18 objects were collected for analysis, but 6 could not be identified due to darker pigmentation due to melting under the laser. Statistics and figures were generated with the R statistical environment, version R-2.15.1 (R Development Core Team 2012) and QuantumGIS, version 1.8.0-Lisboa (QGIS Development Team 2013).",
+    "url": "https://peerj.com/articles/185/reviews/",
+    "review_1": "Tomas Perez-Acle \u00b7 Oct 2, 2013 \u00b7 Academic Editor\nACCEPT\nDear authors,\nThank you very much for your efforts to include the comments and solicitations of the reviewing panel. It is now our pleasure to publish your paper in accordance to the previous decision. Thanks a lot for considering PeerJ to be the showcase of this scientific contribution. Looking forward for your future contributions in this field.",
+    "review_2": "Tomas Perez-Acle \u00b7 Aug 12, 2013 \u00b7 Academic Editor\nMAJOR REVISIONS\nDear authors,\nDespite that I have made extensive efforts to find additional reviewers for your article entitled \u201cProtein signatures using electrostatic molecular surfaces in harmonic space\u201d, I haven't succeeded. In order to speed up the evaluation process of your article and considering my expertise in structure-function relationships in protein structure, I have acted as the second reviewer. As a whole, my overall decision is that your paper should be accepted for publication at PeerJ once all issues raised by the review process are properly addressed. Please find my review below (and you will find the Comments of Reviewer 1 below that).\n\nBasic Reporting:\nThe article entitled \u201cProtein signatures using electrostatic molecular surfaces in harmonic space\u201d from authors Carvalho, Vlachakis, Tsiliki, Vasileios, Megalooikonomou and Kossida, presents a novel approach to use the fourier spectrum computed from the electrostatic potential derived from protein surfaces as a comparative value among protein structures. According to the authors, this method reduces the 3D information available at the protein surface to 1D information where a specific distance metric in the harmonic space may be used to compare protein structures.\n\nExperimental design:\nBy relying on the Wiener-Khinchin theorem, the authors of this paper correlate different points of the electrostatic field computed from an APBS calculation by applying a Fourier transform of the power spectrum inferred over a regular grid over the protein surface. The authors evaluated their method by using a training set composed by 3 helicase proteins from HCV, including 2 crystal structures and 1 comparative model. They also included as decoys 1 HCV polymerase and 1 cAMP-dependant kinase.\n\nValidity of the findings:\nMinor concerns\n- Despite that the authors succeeded presenting a method to compare protein structures relying on 1D information, it is unclear how this method can be used to extract functional information from proteins. Moreover, one of the authors\u2019 claims is that this method allows \"a fast similarity search\". Though comparing 1D information is faster than 3D, the authors do not consider that their method requires a previous calculation of the electrostatic potential of every protein structure. Is the time needed to compute the electrostatic potential being considered as part of the method? On the other hand, in order to better evaluate the performance of their method, the authors should compare their results with well-established methods such as VASP and others.\n- Notoriously, the elegant approach by which the authors use the Fourier transform of the power spectrum computed on the electrostatic potential of protein surfaces, heavily contrast with the use of a simple \u201cdistance metric\u201d by which they compare the power spectrum of different proteins. The inclusion of a more robust method such as SVM or Kohonen maps to compare proteins, could improve their findings.\n- In order to simplify their calculations, the authors have assumed that electrostatic field is isotropic so that the power spectrum depends only on the distance between each pair of points. However, it is well known that in the surface of proteins is where the side chain of charged residues is actually located. At physiological pH, many residues can be charged forming dipoles that are sensitive to the charge flow. Therefore, It is unclear how this simplification can be used to account for changes in electrostatic potential due to flow of charges as in the case of MD simulations.\n\nMajor concerns:\n- Overall, the major concern behind this interesting work is the scarcity of the protein structure database that was used as test set. Despite that the inclusion of decoys is an elegant way to approximate to the real values of precision and recall, the authors should make a compelling effort to increase the size of both the training and the test set used to evaluate this study. They should also improve the statistics behind the evaluation of their method by taking a look at the actual predictive value (precision and recall).",
+    "review_3": "Reviewer 1 \u00b7 Jul 22, 2013\nBasic reporting\nThe contribution report a new method of analysis of protein molecular surface based on Fourier analysis. The authors calculate a power spectrum of surfaces of the molecular electrostatic potential, to compare with other molecules in a structural database. This method allows a reduction of the three-dimensional information to one-dimensional information. They use a training set of the Hepatitis C viral proteins.\nThis approach shows a contribution to speed-up the comparison and classification of proteins. The method is mathematically correct and allows the calculation a power spectrum that can easily be comparing between different molecules.\nExperimental design\nThe training set is small to can show the real meaningful of the methods, in fact the author use only to structure, Helicase A and Helicase B, all the other structures were derivatives of one of them, so they present only small differences with the original, and in this case are not candidate to show the advantages of the methods. In fact, one of them present a power spectrum different to the non related protein (Polymerase) but the other present a power spectrum similar to the other non related protein ( kinase).\nValidity of the findings\nThe findings are based in a trainning set not robust, I will suggest use a bigger and robust training set, to show the real effect of the methods.\nAdditional comments\nThis a good contribution that need inclrese the training set and demeostrate tha capability of the powwr spectrum to differenciate non related protein.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Protein signatures using electrostatic molecular surfaces in harmonic space (v0.1)\". PeerJ https://doi.org/10.7287/peerj.185v0.1/reviews/1",
+    "pdf_1": "https://peerj.com/articles/185v0.2/submission",
+    "pdf_2": "https://peerj.com/articles/185v0.1/submission",
+    "all_reviews": "Review 1: Tomas Perez-Acle \u00b7 Oct 2, 2013 \u00b7 Academic Editor\nACCEPT\nDear authors,\nThank you very much for your efforts to include the comments and solicitations of the reviewing panel. It is now our pleasure to publish your paper in accordance to the previous decision. Thanks a lot for considering PeerJ to be the showcase of this scientific contribution. Looking forward for your future contributions in this field.\nReview 2: Tomas Perez-Acle \u00b7 Aug 12, 2013 \u00b7 Academic Editor\nMAJOR REVISIONS\nDear authors,\nDespite that I have made extensive efforts to find additional reviewers for your article entitled \u201cProtein signatures using electrostatic molecular surfaces in harmonic space\u201d, I haven't succeeded. In order to speed up the evaluation process of your article and considering my expertise in structure-function relationships in protein structure, I have acted as the second reviewer. As a whole, my overall decision is that your paper should be accepted for publication at PeerJ once all issues raised by the review process are properly addressed. Please find my review below (and you will find the Comments of Reviewer 1 below that).\n\nBasic Reporting:\nThe article entitled \u201cProtein signatures using electrostatic molecular surfaces in harmonic space\u201d from authors Carvalho, Vlachakis, Tsiliki, Vasileios, Megalooikonomou and Kossida, presents a novel approach to use the fourier spectrum computed from the electrostatic potential derived from protein surfaces as a comparative value among protein structures. According to the authors, this method reduces the 3D information available at the protein surface to 1D information where a specific distance metric in the harmonic space may be used to compare protein structures.\n\nExperimental design:\nBy relying on the Wiener-Khinchin theorem, the authors of this paper correlate different points of the electrostatic field computed from an APBS calculation by applying a Fourier transform of the power spectrum inferred over a regular grid over the protein surface. The authors evaluated their method by using a training set composed by 3 helicase proteins from HCV, including 2 crystal structures and 1 comparative model. They also included as decoys 1 HCV polymerase and 1 cAMP-dependant kinase.\n\nValidity of the findings:\nMinor concerns\n- Despite that the authors succeeded presenting a method to compare protein structures relying on 1D information, it is unclear how this method can be used to extract functional information from proteins. Moreover, one of the authors\u2019 claims is that this method allows \"a fast similarity search\". Though comparing 1D information is faster than 3D, the authors do not consider that their method requires a previous calculation of the electrostatic potential of every protein structure. Is the time needed to compute the electrostatic potential being considered as part of the method? On the other hand, in order to better evaluate the performance of their method, the authors should compare their results with well-established methods such as VASP and others.\n- Notoriously, the elegant approach by which the authors use the Fourier transform of the power spectrum computed on the electrostatic potential of protein surfaces, heavily contrast with the use of a simple \u201cdistance metric\u201d by which they compare the power spectrum of different proteins. The inclusion of a more robust method such as SVM or Kohonen maps to compare proteins, could improve their findings.\n- In order to simplify their calculations, the authors have assumed that electrostatic field is isotropic so that the power spectrum depends only on the distance between each pair of points. However, it is well known that in the surface of proteins is where the side chain of charged residues is actually located. At physiological pH, many residues can be charged forming dipoles that are sensitive to the charge flow. Therefore, It is unclear how this simplification can be used to account for changes in electrostatic potential due to flow of charges as in the case of MD simulations.\n\nMajor concerns:\n- Overall, the major concern behind this interesting work is the scarcity of the protein structure database that was used as test set. Despite that the inclusion of decoys is an elegant way to approximate to the real values of precision and recall, the authors should make a compelling effort to increase the size of both the training and the test set used to evaluate this study. They should also improve the statistics behind the evaluation of their method by taking a look at the actual predictive value (precision and recall).\nReview 3: Reviewer 1 \u00b7 Jul 22, 2013\nBasic reporting\nThe contribution report a new method of analysis of protein molecular surface based on Fourier analysis. The authors calculate a power spectrum of surfaces of the molecular electrostatic potential, to compare with other molecules in a structural database. This method allows a reduction of the three-dimensional information to one-dimensional information. They use a training set of the Hepatitis C viral proteins.\nThis approach shows a contribution to speed-up the comparison and classification of proteins. The method is mathematically correct and allows the calculation a power spectrum that can easily be comparing between different molecules.\nExperimental design\nThe training set is small to can show the real meaningful of the methods, in fact the author use only to structure, Helicase A and Helicase B, all the other structures were derivatives of one of them, so they present only small differences with the original, and in this case are not candidate to show the advantages of the methods. In fact, one of them present a power spectrum different to the non related protein (Polymerase) but the other present a power spectrum similar to the other non related protein ( kinase).\nValidity of the findings\nThe findings are based in a trainning set not robust, I will suggest use a bigger and robust training set, to show the real effect of the methods.\nAdditional comments\nThis a good contribution that need inclrese the training set and demeostrate tha capability of the powwr spectrum to differenciate non related protein.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Protein signatures using electrostatic molecular surfaces in harmonic space (v0.1)\". PeerJ https://doi.org/10.7287/peerj.185v0.1/reviews/1\nReview 4: nan\nReview 5: nan\nReview 6: \n",
+    "LLM_extracted_review": "[\"Review 1: The paper is accepted for publication, and the authors are thanked for their efforts in addressing the comments from the reviewing panel.\\n\\nReview 2: \\n- Minor concerns include the unclear extraction of functional information from the method, the need to consider the time required for electrostatic potential calculations, and the suggestion to compare results with established methods like VASP. \\n- The use of a simple distance metric for comparing power spectra is criticized, with a recommendation to use more robust methods like SVM or Kohonen maps. \\n- The assumption of isotropy in the electrostatic field is questioned, particularly regarding its implications for charge flow in proteins. \\n- Major concerns highlight the small size of the protein structure database used for testing, suggesting the need for a larger training and test set and improved statistical evaluation of the method.\\n\\nReview 3: \\n- The method is mathematically correct and allows for the comparison of protein structures, but the training set is too small to demonstrate the method's effectiveness. \\n- The limited diversity in the training set, primarily consisting of derivatives of two helicase structures, raises concerns about the robustness of the findings. \\n- A larger and more diverse training set is recommended to better showcase the method's capabilities in differentiating unrelated proteins.\"]"
+}

peerj_json_files/PeerJ_Json_136.json ADDED Viewed

	@@ -0,0 +1,17 @@

+{
+    "v1_Abstract": "We developed a novel method based on the Fourier analysis of protein molecular surfaces to speed up the analysis of the vast structural data generated in the post--genomic era. This method computes the power spectrum of surfaces of the molecular electrostatic potential, whose three--dimensional coordinates have been either experimentally or theoretically determined. Thus we achieve a reduction of the initial three--dimensional information on the molecular surface to the one--dimensional information on pairs of points at a fixed scale apart. Consequently, the similarity search in our method is computationally less demanding and significantly faster than shape comparison methods. As proof of principle, we applied our method to a training set of viral proteins that are involved in major diseases such as Hepatitis C, Dengue fever, Yellow fever, Bovine viral diarrhea and West Nile fever. The training set contains proteins of four different protein families, as well as a mammalian representative enzyme. We found that the power spectrum successfully assigns a unique signature to each protein included in our training set, thus providing a direct probe of functional similarity among proteins. The results agree with established biological data from conventional structural biochemistry analyses.",
+    "v1_col_introduction": "introduction  : The spatial structure of proteins encodes information on their function, which is essential for a successful drug design. In the post\u2013genomic era, the search for functional similarities among proteins is based mostly on identity and/or similarity of genomic sequences rather than on their spatial structure. An approach that has been widely used is the application of self-organizing maps to the protein amino acid sequence in order to predict the protein shape and to infer the protein function [1, 2]. This approach searches for local similarities in the amino acid sequence and is based on the assumption that the proteins have the same size and that the amino acid sequence is a determinant of the protein structure. However, there are many examples of proteins where sequence\u2013based searches are insufficient to describe their biological function [3]. While self-organizing maps can classify proteins into families, they fail at predicting the structure. Since structure is more conserved than sequence, evolutionary relationships among proteins, protein structure\u2013function predictions and comparative modelling should be based on structural information, rather than on primary amino acid or genomic sequence [4].\nOther approaches have been developed that search for functional similarities using the complete three\u2013 dimensional information encoded in the spatial coordinates of all the atoms within the protein structure, which have been derived from X\u2013ray or nuclear magnetic res-\nonance experiments. In these approaches, the three\u2013 dimensional protein structure is modelled by a representation (or descriptor) based on topological characteristics or structure elements (see e.g. Ref. [5] and references therein). Although the structure\u2013based approaches are more informative on the function than the sequence\u2013 based ones, structure comparison methods are too slow and are thus rendered impractical to use in large\u2013scale experiments and real\u2013life applications [6\u20138].\nOther approaches use the protein solvent\u2013accessible surface, since it is a stronger determinant of the protein function than sequence or structure [9]. Shape descriptors have been developed based on spatial symmetries, where the search for similarities consists of shape comparison [10\u201312]. However, surface comparison methods are computationally challenging, largely because they suffer from ambiguity in spatial orientation and require that the surfaces be aligned for an optimal matching (see Ref. [9] and references therein).\nBioinformatics has become the new biomedical informatics bottleneck, as the cost of genome sequencing and the sheer quantity of genomic data has recently skyrocketed. It has been estimated that the unprocessed data generated per sequencing machine can be of order at least 30 Gbs per day, which can scale up by a significant factor in the case of mapped/processed data. There is a clear requirement for fast and efficient analysis of the entire genome/proteome sequencing data in the up\u2013coming era of personalized medicine. Due to the continuous improvements in sequencing technologies and proteomic method-\nPeerJ reviewing PDF | (v2013:06:567:1:0:NEW 1 Oct 2013)\nR ev ie w in g M an\nus cr ip t\n2 ologies, the current scaling of available computing, storage and analysis throughput is far lower than the scaling of the data generation rate. The induced lag between the processing potential and the processing requirements already poses problems to researchers and companies in the bioinformatics field. Since it is impossible to constantly upgrade computer hardware to keep up with the increasing data production rate, the only feasible solution is to devise algorithms that can offer a competing processing scaling using the existing hardware at its full potential.\nHere, we propose a new approach to search for functional similarities among proteins using their molecular surfaces [13]. Protein molecular surfaces are determinant of the protein biological activity, with different types of molecular surfaces encoding different information about the protein function. We choose to use surfaces of the molecular electrostatic potential due to the importance of the charge distribution in the protein\u2013protein interactions. Protein\u2013protein interactions are essential for cell signalling and cell function [14, 15]. These processes require a correct and fast molecular recognition, in which interactions among electrostatic charges intervene. Disturbances in these processes are in the origin of almost every major disorder [16] and may lead to severe diseases such as cancer [17\u201319]. Therefore the electrostatic potential distribution on the protein molecular surfaces is crucial to virtually all biological macromolecules involved in key biochemical pathways [20\u201322].\nOnce we calculate the molecular surface for a particular filter, we proceed to measure the signal off of the molecular surface. For our signal analysis, we propose a method based on the Fourier analysis of molecular surfaces. An advantage of Fourier analysis is that it most easily separates large from small scales. The signal at each point can be regarded as a realization of a distribution of fluctuations around an average value of the molecular surface [23, 24]. Instead of measuring information on the individual points over the surface as shape descriptors do, we measure information on the correlations among the points, thus waiving the need that the surfaces be aligned. The simplest statistic is the two\u2013point correlation function in Fourier space, which averages the signal over the whole volume and measures the variance in the distribution. Hence our approach transforms three\u2013 dimensional spatial data into one\u2013dimensional frequency data.\nThe manuscript is organized as follows. First we present the selected proteins and how we synthesise the corresponding molecular surfaces. Then we describe our proposed method to extract functional information, based on the Fourier analysis of molecular surfaces and on a dimensionality reduction of the usable information. Then we present the results and discuss further improvements in the robustness of this method. Finally we outline an integrated solution for a functional similarity search among proteins, which progresses towards a\nFigure 1: Surfaces of the electrostatic molecular potential. Left panel: Hepatitis C helicase protein, Right panel: Hepatitis C polymerase protein. The electrostatic potential is measured in eV, with range as shown in the corresponding colour bar.\ndimensionality increase of the usable information and a reduction of the protein sample size.",
+    "v2_Abstract": "We developed a novel method based on the Fourier analysis of protein molecular surfaces to speed up the analysis of the vast structural data generated in the post{genomic era. This method computes the power spectrum of surfaces of the molecular electrostatic potential, whose three-dimensional coordinates have been either experimentally or theoretically determined. Thus we achieve a reduction of the initial three{dimensional information on the molecular surface to the one-dimensional information on pairs of points at a fixed scale apart. Consequently, the similarity search in our method is computationally less demanding and significantly faster than shape comparison methods. As proof of principle, we applied our method to a training set of the Hepatitis C viral proteins with similar and dissimilar functions, as well as to a mammalian representative enzyme. We found that the power spectrum successfully assigns a unique signature to each protein included in our training set, thus providing a direct probe of functional similarity among proteins. The results agree with established biological data from conventional structural biochemistry analyses.",
+    "v2_col_introduction": "introduction  : The spatial structure of proteins encodes information on their function, which is essential for a successful drug design. In the post\u2013genomic era, the search for functional similarities among proteins is based mostly on identity and/or similarity of genomic sequences rather than on their spatial structure. There are many examples of proteins where sequence\u2013based searches are insufficient to describe their biological function [1]. Since structure is more conserved than sequence, evolutionary relationships among proteins, protein structure\u2013function predictions and comparative modelling should be based on structural information, rather than on primary amino acid or genomic sequence [2].\nApproaches have been developed that search for functional similarities using the complete three\u2013dimensional information encoded in the spatial coordinates of all the atoms within the protein structure, which have been derived from X\u2013ray or nuclear magnetic resonance experiments. In these approaches, the three\u2013dimensional protein structure is modelled by a representation (or descriptor) based on topological characteristics or structure elements (see e.g. Ref. [3] and references therein). Although the structure\u2013based approaches are more informative on the function than the sequence\u2013based ones, structure comparison methods are too slow and are thus rendered impractical to use in large\u2013scale experiments and real\u2013life applications [4\u20136].\nOther approaches use the protein solvent\u2013accessible surface, since it is a stronger determinant of the protein\nfunction than sequence or structure [7]. Shape descriptors have been developed based on spatial symmetries, where the search for similarities consists of shape comparison [8\u201310]. However, surface comparison methods are computationally challenging, largely because they suffer from ambiguity in spatial orientation and require that the surfaces be aligned for an optimal matching (see Ref. [7] and references therein).\nBioinformatics has become the new biomedical informatics bottleneck, as the cost of genome sequencing and the sheer quantity of genomic data has recently skyrocketed. It has been estimated that the unprocessed data generated per sequencing machine can be of order at least 30 Gbs per day, which can scale up by a significant factor in the case of mapped/processed data. There is a clear requirement for fast and efficient analysis of the entire genome/proteome sequencing data in the up\u2013coming era of personalized medicine. Due to the continuous improvements in sequencing technologies and proteomic methodologies, the current scaling of available computing, storage and analysis throughput is far lower than the scaling of the data generation rate. The induced lag between the processing potential and the processing requirements already poses problems to researchers and companies in the bioinformatics field. Since it is impossible to constantly upgrade computer hardware to keep up with the increasing data production rate, the only feasible solution is to devise algorithms that can offer a competing processing scaling using the existing hardware at its full potential.\nHere, we propose a new approach to search for functional similarities among proteins using their molecular\nPeerJ reviewing PDF | (v2013:06:567:0:1:NEW 7 Jun 2013)\nR ev ie w in g M an\nus cr ip t\n2 surfaces [11]. Protein molecular surfaces are determinant of the protein biological activity, with different types of molecular surfaces encoding different information about the protein function. We choose to use surfaces of the molecular electrostatic potential due to the importance of the charge distribution in the protein\u2013protein interactions. Protein\u2013protein interactions are essential for cell signalling and cell function [12, 13]. These processes require a correct and fast molecular recognition, in which interactions among electrostatic charges intervene. Disturbances in these processes are in the origin of almost every major disorder [14] and may lead to severe diseases such as cancer [15\u201317]. Therefore the electrostatic potential distribution on the protein molecular surfaces is crucial to virtually all biological macromolecules involved in key biochemical pathways [18\u201320].\nOnce we calculate the molecular surface for a particular filter, we proceed to measure the signal off of the molecular surface. For our signal analysis, we propose a method based on the Fourier analysis of molecular surfaces. An advantage of Fourier analysis is that it most easily separates large from small scales. The signal at each point can be regarded as a realization of a distribution of fluctuations around an average value of the molecular surface [21, 22]. Instead of measuring information on the individual points over the surface as shape descriptors do, we measure information on the correlations among the points, thus waiving the need that the surfaces be aligned. The simplest statistic is the two\u2013point correlation function in Fourier space, which averages the signal over the whole volume and measures the variance in the distribution. Hence our approach transforms three\u2013 dimensional spatial data into one\u2013dimensional frequency data.\nThe manuscript is organized as follows. First we describe our proposed method to extract functional information, based on the Fourier analysis of molecular surfaces and on a dimensionality reduction of the usable information. Then we present the results and discuss further improvements in the robustness of this method. Finally we outline an integrated solution for a functional similarity search among proteins, which progresses towards a dimensionality increase of the usable information and a reduction of the protein sample size.",
+    "v1_text": "results : Power spectrum of the molecular surfaces of the selected proteins To test our method, we used for training set the protein simulations described above, containing four different protein families. For each molecular surface of the electrostatic potential, we computed its power spectrum and the corresponding white-noise power spectrum. The white\u2013noise power spectrum was computed from a surface synthesised as a Gaussian distribution N(0, 1) times the mean value of the corresponding molecular surface. We observe that the power spectra of all molecular surfaces have comparable magnitudes, stabilizing around 10\u22126 for sufficiently large k (not shown), whereas the white\u2013noise power spectra have magnitudes that range from 10\u221211 to 10\u22127 (Fig. 3). This range is populated by the HCV polymerase at the top, followed by the HCV helicaseStrB and HCV helicaseHM in the intermediary range, and finally the 1A1Vs helicase HCV helicaseStrA and its models HCV helicaseEM and HCV helicaseMD at the bottom. Hence the information derived from the mean value alone, assuming an underlining Gaussian distribution, suggests a coarse clustering of the proteins in helicases, polymerases and a mixed PeerJ reviewing PDF | (v2013:06:567:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t 6 Figure 8: Power spectrum of the molecular surfaces of the selected methyltransferase proteins. Power spectra of the molecular surfaces divided by the power spectra of the corresponding white-noise molecular surfaces. The symbols depict the power spectra and the error bars depict the error associated with the measurement. The values of k are measured in nm\u22121. cluster containing helicases and non\u2013helicases. For an easier comparison of the results, we divided the power spectra of molecular surfaces by the mean of the corresponding white\u2013noise power spectra (Figs. 4, 5, 6, 7, 8, 9). For each molecular surface, the power at each k is one realization of a distribution, hence the power spectrum is noisy. This noise was estimated by the uncertainty of the Fourier coefficients at each k, given by \u2206P (k) = P (k) \u221a 2/k2, which we used to compute the error bars. We also included the power spectrum of Mouse kinase in all plots, which shows a nearly flat spectrum punctuated by irregular peaks. First we analyse the HCV helicase protein set, which illustrates how our method performs at distinguishing different treatments and strains of the same protein. We plotted the power spectra of the HCV helicase proteins in Fig. 4. We observe that the power spectra of HCV helicaseStrA, HCV helicaseEM, HCV helicaseMD and HCV helicaseStrB exhibit a similar pattern up to k \u2248 10 nm\u22121 compatible with that of HCV helicaseHM. A further inspection reveals details that distinguish among the helicases. In particular, we observe that the power spectra of the models HCV helicaseEM and HCV helicaseMD exhibit very similar patterns of peaks attesting to their similar binding state. Although HCV helicaseStrA is in a different binding state, its power spectrum exhibits the same level of similarities with both HCV models, with an anticipated HCV-like grouping of peaks specific to our data. For k > 1 nm\u22121, these three helicase proteins exhibit three strong peaks at k \u2248 2.3, 4.6, 7.3 nm\u22121. From the distance between peaks, we infer an average wavelength of \u03bb \u2248 2.5 nm. The power spectrum of HCV helicaseHM follows the same pattern as that of HCV helicaseStrA shifted to smaller k with a varying relative phase which most of the time is close to \u03c0, with strong peaks at k \u2248 1.8, 3.6 nm\u22121. In comparison with the HCV models, the power spectrum of HCV helicaseStrB exhibits differences in the position of the peaks (found at k \u2248 2.7, 3.6, 5.5 nm\u22121) and in their amplitude ratios, which attest to the different treatment in HCV helicaseStrB from that in the HCV models. As k increases, we observe a gradual damping of the power of the helicases and an emerging tail reminiscent of shot noise in a Poisson power spectrum, more prominent in HCV helicaseHM and HCV helicaseStrB, PeerJ reviewing PDF | (v2013:06:567:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t 7 which indicates the damping of the fluctuations about the mean value and thus the vanishing of the structural signal. This damping is most visible for k > 6 nm\u22121. This agrees with the observation above that sets the upper limit of k to the size of a typical cluster of aminoacids and hence sets the minimum distance below which correlations are not of biological interest nor can be reliably probed by X\u2013ray/NMR experiments. We now proceed to analyse the remaining helicase strains, which illustrates how our method performs at distinguishing strains of the same family. We plotted the power spectra of the Dengue virus (DEN) helicase proteins in Fig. 5 and the Yellow fever virus (YF) helicase proteins in Fig. 6. The power spectra of both the YF helicase proteins and the DEN helicase proteins exhibit a similar pattern with a varying relative phase among the proteins of each strain, with the difference between the two strains being in the typical wavelength and amplitude. In particular, we observe that the DEN helicases have an underlying flat spectrum punctuate by peaks at k \u2248 2.0, 3.8, 5.5 nm\u22121 (DEN 2BHR), k \u2248 3.0, 5.0, 6.0 nm\u22121 (DN 2JLQ) and k \u2248 3.5, 5.0, 6.5 nm\u22121 (DN 2JLU), that yield an average \u03bb \u2248 4.2 nm. The DEN 2BMF shows a different pattern characterized by a decreasing power law up to k \u2248 5, with superposed peaks k \u2248 5.5, 7.0, 8.5 nm\u22121. In contrast, the YF helicases have a nearly flat spectrum punctuated by small peaks at k \u2248 1.0, 3.5, 5.5 nm\u22121 (YF 1YKS) and k \u2248 1.5, 5.5 nm\u22121 (YF 1YMF), that yield an average \u03bb \u2248 2.0 nm. The YF 2V80 shows a nearly flat spectrum with barely no peaks, indicating a predominantly isotropic distribution of power. These families have a similar pattern with the HCV helicases but the features have smaller amplitudes. The global pattern attests to the fact that these proteins are also helicases and have the same treatment as the HCV, whereas the differences in amplitude attest to the fact that are of different strains. We now proceed to analyse the non-helicase families, which illustrates how our method performs at distinguishing protein families. We plotted the power spectra of the polymerase proteins in Fig. 7. We observe that all the polimerases have the same pattern characterized by an underlying decreasing power law with superposed peaks. In particular, the West Nile strains have the same pattern at all scales and a peak at k \u2248 8 nm\u22121, i.e. close to the smallest scale accessible. The BVDV strain has a very similar power law behaviour to the WN strains but is punctuated by regular peaks namely at k \u2248 1.5, 3.0, 4.5, 6.0, 8.0 nm\u22121, corresponding to an average \u03bb \u2248 4 nm. The HCV polymerase has the steepest decreasing power law behaviour and peaks at k \u2248 5.5, 7.5 nm\u22121. We plotted the power spectra of the methyltransferase proteins in Fig. 8. We observe that all the YF methyltransferase have similar patterns characterized by a nearly flat, featureless power spectrum punctuated by Figure 10: Power spectra of the molecular surfaces of the HCV helicaseEM after being subject to molecular dynamics simulations for 100 ps. Power spectra of the molecular surfaces divided by the power spectra of the corresponding white-noise molecular surfaces. The symbols depict the power spectra and the error bars depict the error associated with the measurement. The values of k are measured in nm\u22121. irregular low\u2013amplitude peaks. Finally, we plotted the power spectra of the glycoproteins in Fig. 9. We observe that all the BVDV glycoproteins have the same pattern characterized by an underlying a convex quadratic function with superposed peaks. In particular both the strains 4DVN and 4DW3 have a single peak at k \u2248 7.5 nm\u22121, whereas the 4DW4 have peaks at k \u2248 2.0, 4.5, 6.0, 8.0 nm\u22121, corresponding to an average \u03bb \u2248 3.3 nm. Power Spectrum of a dynamical simulation To further test our method, we used the 1A1V template energetically minimized up to a gradient of 10\u22125 to generate ten dynamical realizations captured in ten time frames separated by 10 ps. We then energetically minimized the tenth frame up to a gradient of 10\u22125 [38\u201340] We computed the power spectrum of each frame, generated the corresponding white noise surface and plotted the results in Fig. 10. The purpose of this test is to show how our method behaves when applied to controlled simulations. We observe that there is no significant difference among the different frames. This observation supports the fact that the surfaces do not change over time after energy minimization (EM). Also we observe that the two simulations energetically minimized up to a gradient 10\u22125 are in phase, whereas the simulation with up to a gradient 5 \u00d7 10\u22122 is visibly out of phase with the former. This observation supports the fact that there is a difference between crude and fine EM. PeerJ reviewing PDF | (v2013:06:567:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t 8 conclusions : We presented a new method based on the Fourier analysis of protein molecular surfaces to extract functional information on proteins. For a selected set of proteins of HCV with different structural features, we first produced surfaces of the molecular electrostatic potential, as well as the corresponding white\u2013noise surfaces, and then computed their two-point correlation function in harmonic space (the power spectrum). We found that this method can distinguish different functional protein groups. More specifically, in this manuscript we established a helicase, a polymerase, a methyltransferase and a glycoprotein group. We also tested this method on dynamical simulations after energy minimization. An immediate extension of this work is the application of this method to isolated structural subunits that form larger structures within proteins. Similarly sized subunits will have a strong signal in the same frequency range, which will add up in the protein power spectrum. Hence, we must first measure the contribution of each subunit separately and produce a catalogue of subunit signatures, so we can distinguish them in the combined signal when running similarity searches. By reducing the initial three\u2013dimensional information on the molecular surface to the one\u2013dimensional information on pairs of points at a fixed scale apart, this method allows for a fast similarity search. Further refinements in the similarity search will require methods that use information from higher\u2013order correlation functions, such as the correlation among three points at a fixed triangular configuration or its Fourier\u2013transformed (the bispectrum). (See Fig. 2 right panel for an illustration.) The bispectrum measures phase correlations among the modes and thus deviations from a Gaussian distribution. Our ultimate goal is to integrate higher\u2013order correlations and to apply the resulting method to the RCSB database so as to provide the biopharmaceutical and structural research communities with a novel and easily searchable reference without the three\u2013dimensional information compromising the speed of the calculation. This method aims to coalesce techniques, which have been extensively tested and used in other fields such as cosmology, into a fast and robust pipeline for the analysis and processing of very large, three\u2013dimensional biological datasets in an effort to speed up protein similarity searches. Acknowledgments CSC is funded by the FCT\u2013Lisbon, Grant no. SFRH/BPD/65993/2009. This work was supported in part by the European Union (European Social Fund - ESF) and Greek national funds through the Operational Program \u201dEducation and Lifelong Learning\u201d of the National Strategic Reference Framework (NSRF) - Research Funding Program: \u201cThales\u201d. Investing in knowledge society through the European Social Fund. \u2217 Electronic address: cscarvalho@oal.ul.pt \u2020 Electronic address: skossida@bioacademy.gr [1] Kohonen T, (1982) Biol Cybern 43, 59\u201369 [2] Andrade MA, Casari G, Sander C and Valencia A (1997) Biol Cybern 76, 441\u2013450 [3] Dobson PD, Cai YD, Stapley BJ, Doig AJ (2004) Curr Med Chem 11: 2135\u20132142 [4] Illerg\u030aard K, Ardell DH, Elofsson A (2009) Proteins 15; 77(3): 499\u2013508. [5] Venkatraman V, Sael L and Kihara D (2009), Cell Biochem Biophys 54: 23\u201332. [6] Kolodny R, Koehl P, Levitt M (2005) J Mol Biol 346: 1173\u20131188. [7] Mayr G, Domingues FS, Lackner P (2007) BMC Struct Biol 7: 50. [8] Berbalk C, Schwaiger CS, Lackner P (2009) Protein Sci 18: 2027\u20132035. [9] Via A, Ferre\u0300 F, Brannetti B and Helmer-Citterich M (2000), Protein surface similarities: a survey methods to describe and compare protein surfaces [10] Kazhdan M, Funkhouser T and Rusinkiewicz S (2003), Proc 2003 Eurographics, 43: 156\u2013164. [11] Ritchie DW, Kozakov D and Vajda S (2008), Bioinformatics 24(17): 1865\u20131873. [12] Venkatraman V, Chakravarthy PR, and Kihara D (2009), J Cheminformatics 1: 19. [13] Vlachakis D, Tsiliki G, Tsagkrasoulis D, Carvalho CS, Megalooikonomou V, Kossida S (2012) EMBnet J. 18(1): 6\u20139. [14] Przytycka TM, Singh M, Slonim DK (2010) Brief Bioinform. 11 (1): 15\u201329. [15] Berger-Wolf TY, Przytycka TM, Singh M, Slonim DK (2010) Pac Symp Biocomput. 15: 120\u2013122. [16] Gire SK, Stremlau M, Andersen KG, Schaffner SF, Bjornson Z, Rubins K, Hensley L, McCormick JB, Lander ES, Garry RF, Happi C, Sabeti PC (2012) Science 9, 338(6108): 750\u2013752. [17] Elcock AH, Gabdoulline RR, Wade RC, McCammon JA. (1999) J Mol Biol. 291(1): 149\u2013162. [18] Sept D, Elcock AH, McCammon JA. (1999) J Mol Biol. 294(5): 1181\u20131189. [19] Wlodek ST, Shen T, McCammon JA. (2000) Biopolymers. 53(3): 265\u2013271. [20] Honig B, Nicholls A. (1995) Science 268(5214): 1144\u2013 1149. [21] Wong GC, Pollack L. (2010) Annu Rev Phys Chem. 61: 171\u2013189. [22] McCammon JA. (2009) Proc Natl Acad Sci USA. 106(19): 7683\u20137684. [23] Vlachakis D, Champeris\u2013Tsaniras S, Kossida S. (2012) Mol Biochem. 1(3): 144\u2013149. [24] Kandil S, Biondaro S, Vlachakis D, Cummins AC, Coluccia A, Berry C, Leyssen P, Neyts J and Brancale A. (2009) J Bioorg Med Chem Lett. 1; 19(11): 2935\u20132937. [25] Konecny R, Baker NA, McCammon JA (2012) Comput Sci Discov 26; 5(1) 015005. [26] Unni S, Huang Y, Hanson RM, Tobias M, Krishnan S, Li WW, Nielsen JE, Baker NA (2011) J Comput Chem. 32(7): 1488\u20131491. [27] www.poissonboltzmann.org/apbs [28] www.rcsb.org PeerJ reviewing PDF | (v2013:06:567:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t 9 [29] Hess B, Kutzner C, van der Spoel D, Lindahl E (2008) J Chem Theory Comput 4: 435\u2013447. [30] Lindahl E, Hess B, van der Spoel D (2001) J Mol Model 7: 306\u2013317. [31] van der Spoel D, Lindahl E, Hess B, Groenhof G, Mark AE, et al. (2005) J Comput Chem 26: 1701\u201318. [32] Sellis D, Vlachakis D, Vlassi M (2009) Bioinform Biol Insights 3: 99\u2013102. [33] S\u0306ali A, Blundell TL (1993) J. Mol. Biol. 234: 779\u2013815. [34] Eswar N, John B, Mirkovic N, Fiser A, Ilyin VA, Pieper U, Stuart AC, Mart\u0301\u0131-Renom MA, Madhusudhan MS, Yerkovich B, S\u0306ali A (2003) Nucl Acids Res 31: 3375\u2013 3380. [35] Laskowski RA, Rullmannn JA, MacArthur MW, Kaptein R, Thornton JM (1996) J. Biomol. NMR 8: 477\u2013486. [36] Connolly ML (1983), J. Appl. Cryst. 16: 548\u2013558. [37] Peacock JA (1999) Cosmological Physics, CUP. [38] Vangelatos I, Vlachakis D, Sophianopoulou V, Diallinas G (2009) Molecular membrane biology 26 (5-7): 356-370 [39] Sellis D, Drosou D, Vlachakis D, Voukkalis D, Gian- nakouros T, Vlassi M (2012) Biochimica et Biophysica Acta 1820 (1): 44\u201355 [40] Vlachakis D, Tsagrasoulis D, Megalooikonomou V, Kossida S (2013) Bioinformatics 29 (1): 126-128 PeerJ reviewing PDF | (v2013:06:567:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t method : selected proteins : For our training set, we selected four distinct protein families, which include twelve helicase proteins, six methyltransferase proteins, four polymerase proteins and four glycoproteins. These proteins are mainly viral components that are involved in major diseases such as Hepatitis C, Dengue fever, Yellow fever, Bovine viral diarrhea and West Nile fever. We use the Mouse kinase protein as a decoy, since it has a very different function from all the other proteins. Helicases are responsible for the unwinding of double stranded DNA or RNA during viral replication. Polymerases are key enzymes that are used for copying the viral genetic material. Methyltransferases or methylases are transferase enzymes that are responsible for transferring methyl groups from a donor to an acceptor. Finally glycoproteins are used for molecular recognition by viruses. Protein treatments vary depending on the needs of each comparison chart. The main treatment is the default X\u2013ray crystallography protein conformation as it is deposited in the RCSB database [28]. The selected unedited proteins were the following. Among the helicases, we selected: a) 1A1V and 8OHM of the Hepatitis C virus (HCV), b) 1YMF, 1YKS and 2V80 of the Yellow fever virus (denoted by YF 1YMF, YF 1YKS and YF 2V80 respectively), and c) 2JLU, 2BHR, 2BMF and 2JLQ of the Dengue fever virus (denoted by DEN 2JLU, DEN 2BHR, DEN 2BMF and DEN 2JLQ respectively). Among the polymerases, we selected 2CJQ, 2HCS and 2HCN of the West Nile fever virus (denoted by WN 2CJQ, WN 2HCS and WN 2HCN respectively). Among the methyltransferase, we selected PeerJ reviewing PDF | (v2013:06:567:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t 3 3EVA, 3EVB, 3EVC, 3EVD, 3EVE and 3EVF of the Yellow fever virus (denoted by YF 3EVA, YF 3EVB, YF 3EVC, YF 3EVD, YF 3EVE and YF 3EVF respectively). Among the glycoproteins, we selected 1NB7, 4DVN, 4DW4 and 4DW3 of the bovine diarrhea virus (denoted by BVDV 1NB7, BVDV 4DVN, BVDV 4DW4 and BVDV 4DW3 respectively). In the Hepatitis C viral protein family, we considered two HCV helicase proteins, namely the HCV helicase strain A (the 1A1V entry, denoted by HCV helicaseStrA) and the HCV helicase strain B (the 8OHM entry, denoted by HCV helicaseStrB), whose three-dimensional coordinates were obtained from the RCSB database [28] of X\u2013ray protein crystallography structures. Furthermore, we generated two simulations of the 1A1V protein crystal, namely the energy\u2013minimized version (denoted by HCV helicaseEM) and the molecular dynamics version (denoted by HCV helicaseMD). Both the HCV helicaseEM and the HCV helicaseMD have been energetically minimized up to a gradient of 0.05. The HCV helicaseMD has additionally been subject to a molecular dynamics simulation. We also established a homology model of the HCV helicase (denoted by HCV helicaseHM) so that the in silico three\u2013dimensional model of HCV was included in our training set. We also included an example of a non\u2013helicase HCV viral protein, namely the 1NB7 structure of the HCV polymerase (the 1NB7 entry, denoted by HCV polymerase). Molecular surfaces of the selected proteins Surfaces of the molecular electrostatic potential follow the nonlinear Poisson\u2013Boltzmann equation [25, 26]. We solved numerically for the electrostatic potential using the finite\u2013difference method as implemented in the APBS Software [27]. The potential was calculated on a regular grid of size (65, 65, 65)1, with the grid\u2013fill\u2013by\u2013 solute parameter set to 80%. The dielectric constants of the solvent and the solute were set to 80.0 and 2.0, respectively. An ionic exclusion radius of 2.0 A\u030a, a solvent radius of 1.4 A\u030a and a solvent ionic strength of 0.145 M were applied. Default APBS charges and atomic radii were used. Energy minimization (EM) removes any residual geometrical strain from each molecular system, whereas molecular dynamics (MD) simulates a periodic cytoplasm\u2013like aqueous environment. Both EM and MD were performed with the Gromacs suite [29\u201331] through 1 The size of the grid was kept small in order to speed up the calculation and reduce the computational load. It was tested to be suitable for this study, as higher detail would not change the surface by much, while it would increase the computational load significantly. our previously developed graphical interface [32]. Molecular dynamics took place in a periodic environment, which was subsequently solvated with the simple point\u2013 charge water model using the truncated octahedron box extending to 7 A\u030a from each molecule. Partial charges were applied and the molecular systems neutralized with counter\u2013ions as required. The temperature was set to 300 K, the pressure to 1 atm and the step size to 2 fs. The total time elapsed at each molecular complex run was 50 ns, using constant number of atoms, volume and temperature (NVT) throughout the calculation in a canonical environment. The results of the MD simulations were collected in a molecular trajectory database for further analysis. The homology model was produced using Modeller [33, 34] and was evaluated using the Procheck utility [35]. This model was designed in order to include a computer modelled structure in our training set, which however shares high sequence identity with its template structure (approximately 90%). The RCSB/PDB entries of the selected proteins are summarized in Table I. In Fig. 1 we show surfaces of the electrostatic molecular potential for two HCV proteins, namely the helicase and the polymerase. The electrostatic potential is measured in eV. In these manuscript, we used the Connolly representation for the molecular surfaces [36]. Power spectrum of molecular surfaces Molecular surfaces contain information on a property of proteins along the three spatial dimensions. This property, in this case the values of the electrostatic potential, can be regarded as a field F (x) defined over points x on the surface. Functional information is encoded not only in the positions of the points but also in the correlations among points. The simplest correlation function that we can measure is that between pairs of points. The twopoint correlation function \u03be of the field F measures the convolution of the field over its complex conjugate (see e.g. Ref. [37]) \u03be(r) \u2261 \u3008F \u2217(x)F (x + r)\u3009 = 1 L3 \u222b d3x F \u2217(x)F (x + r).(1) The angle brackets indicate an averaging over the normalization volume, which here we take as the volume of the molecular surface, L3. We assume that the field has a flat geometry and can be decomposed in a Fourier expansion of plane waves F (x) = \u2211 k Fk exp[\u2212ik \u00b7 x], (2) where the wavenumber k relates with the frequency \u03bd by k = 2\u03c0/\u03bd. If the field has a curved geometry, then a PeerJ reviewing PDF | (v2013:06:567:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t 4 Family Protein Size=[lx, ly, lz] Helicase HCV helicaseStrA [72.4, 64.8, 55.1] HCV helicaseEM [72.8, 65.1, 55.5] HCV helicaseMD [72.3, 65.5, 56.3] HCV helicaseHM [71.5, 65.7, 55.9] HCV helicaseStrB [61.9, 69.6, 61.7] DEN 2BHR [93.5, 101.4, 76.8] DEN 2BMF [84.4, 111.7, 106.0] DEN 2JLQ [66.8, 69.6, 77.0] DEN 2JLU [80.4, 95.0, 85.0] YF 1YKS [62.2, 58.4, 67.8] YF 1YMF [63.6, 58.2, 67.8] YF 2V8O [49.2, 69.6, 67.6] Polymerase HCV polymerase [59.0, 77.7, 65.0] BVDV 2CJQ [74.4, 69.5, 64.5 WN 2HCN [78.5, 75.4, 61.1] WN 2HCS [77.2, 75.2, 63.2] Methyltransferase YF 3EVA [43.9, 56.2, 62.7] YF 3EVB [43.6, 56.4, 62.6] YF 3EVC [43.8, 56.0, 62.2] YF 3EVD [44.1, 55.5, 63.3] YF 3EVE [44.6, 55.9, 65.2] YF 3EVF [43.9, 56.1, 64.2] Glycoproteins BVDV 4DVN [46.2, 67.2, 68.0] BVDV 4DW3 [46.3, 68.5, 67.9] BVDV 4DW4 [46.8, 73.4, 67.6] Kinase Mouse kinase [52.5, 69.0, 48.8] Table I: Input data. Protein families, protein PDB names and sizes of the corresponding molecular surfaces along the [x,y,z]\u2013directions, measured in A\u030a. Fourier expansion in spherical harmonics should be used instead. However, the difference between the two expansions only matters in scales of order the size of the molecular surface, which correspond to the smallest frequency. The smallest frequency is the zero\u2013mode in the Fourier expansion and describes a global offset. The two-point correlation function becomes \u03be(r) = \u2329\u2211 k \u2211 k\u2032 F \u2217kFk\u2032 exp[i(k \u2212 k \u2032) \u00b7 x] exp[\u2212ik\u2032 \u00b7 r] \u232a .(3) Since the molecular surface is closed, the field is periodic within the size of the surface, which restricts the allowed wavenumbers to the harmonic boundary condition kn = (n2\u03c0/L)e\u0302k, where n \u2208 {0, 1, ...} is the order of the Fourier modes. As a consequence, all the cross terms with k\u2032 6= k average to zero and the remaining sum is \u03be(r) = ( L 2\u03c0 )3 \u222b d3k |Fk|2 exp[\u2212ik \u00b7 r]. (4) Hence the correlation function is the Fourier transform of the power spectrum P (k) = |Fk|2. This relationship is known as the Wiener-Khinchin theorem. The power spectrum measures amplitude correlations among the modes, discarding however information on the phase. We proceed to compute the Fourier transform Fk of 3 1 k P 1 PP 2 2 k12 12 k31 k23 P P Figure 2: Schematic representation of point configurations for correlations in harmonic space. Left panel: The configuration of the two-point correlation function contains one free parameter, k12, which is the distance in harmonic space between the two points P1 and P2. Right panel: The configuration of the three-point correlation function contains two free parameters, e.g. k12 and k23, describing the distances in harmonic space respectively between P1 and P2, and between P2 and P3. The third parameter k13 is related to the former two by the triangle condition k12 + k23 + k31 = 0. the molecular surface inferred over a regular grid. The Fourier\u2013transformed surface measures the amplitude of the plane waves whose combination reproduces the information on the original surface. The frequencies of the plane waves range from the frequency corresponding to the extension of the surface (i.e. to n = 1), up to the Nyquist frequency corresponding to twice the bin size of the grid (i.e. to n = N/2, where N is the number of bins along a direction of the grid). The size of the molecular surfaces ranges between 5\u22127 nm (Table I). The smallest spatial scale of biological interest is the size of a typical cluster of aminoacids, which is of order xball \u223c 0.3 nm. We choose this spatial scale for the size of the grid, so that the largest frequency scale that can be probed is of order kball \u223c 10 nm\u22121. Furthermore, we assume that the field is isotropic, i.e. that it does not have a preferential direction, so that the power spectrum depends only on the distance between each pair of points. (See Fig. 2 left panel for an illustration.) By assuming isotropy, we are discarding information on the direction. We proceed to take the ensemble average of P (k) so that the power at the mode k is the sum of the power at all the points on a sphere of radius k from the zero\u2013mode, resulting in a one-dimensional function P (k). In this way, we collapse the information on the three-dimensional field over the molecular surface onto a one-dimensional power spectrum over the wavenumbers of the Fourier\u2013transformed molecular surface. For a given k, we are sampling a distribution, which we assume to be Gaussian with mean value \u3008Fk\u3009 and variance \u2329 |Fk|2 \u232a = P (k), from which the Fourier coefficients Fk are drawn. Hence there is a fundamental uncertainty about the underlying variance, which depends on the number of coefficients sampled at a given k. Since the number of k\u2019s on a sphere of radius k scales as k2 and for any real field it holds that F\u2212k = Fk\u2217, where the asterisk stands for the complex conjugate, then the uncertainty scales as \u2206P (k)/P (k) = \u221a 2/k2. PeerJ reviewing PDF | (v2013:06:567:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t 5 Figure 4: Power spectrum of the molecular surfaces of the selected HCV helicase proteins. Power spectra of the molecular surfaces divided by the power spectra of the corresponding white-noise molecular surfaces. The symbols depict the power spectra and the error bars depict the error associated with the measurement. The values of k are measured in nm\u22121.",
+    "v2_text": "results : To test our method, we used for training set the protein simulations described above, containing helicase and non-helicase proteins. For each molecular surface of the electrostatic potential, we computed its power spectrum and the corresponding white-noise power spectrum. The white\u2013noise power spectrum was computed from a surface synthesised as a Gaussian distribution N(0, 1) times the mean value of the corresponding molecular surface. We observe that the power spectra of all molecular surfaces have comparable magnitudes, stabilizing around 10\u22126 for sufficiently large k (not shown), whereas the white\u2013noise power spectra have magnitudes which range from 10\u221211 to 10\u22127 (Fig. 3). This range is populated by the HCV polymerase at the top, fol- Figure 3: Power spectrum of the molecular surfaces of the selected proteins. Power spectra of the corresponding white-noise molecular surfaces. The values of k are measured in nm\u22121. lowed by the HCV helicaseStrB and HCV helicaseHM in the intermediary range, and finally the 1A1Vs helicase HCV helicaseStrA and its models HCV helicaseEM and HCV helicaseMD at the bottom. Hence the information derived from the mean value alone, assuming an underlining Gaussian distribution, suggests a coarse clustering of the proteins in helicases, polymerases and a mixed cluster containing helicases and non\u2013helicases. For an easier comparison of the results, we divided the power spectra of molecular surfaces by the mean of the corresponding white\u2013noise power spectra (Fig. 4). For each molecular surface, the power at each k is one realization of a distribution, hence the power spectrum is noisy. This noise was estimated by the uncertainty of the Fourier coefficients at each k, given by \u2206P (k) = P (k) \u221a 2/k2, which we used to compute the error bars. We observe that the power spectra of HCV helicaseStrA, HCV helicaseEM, HCV helicaseMD and HCV helicaseStrB exhibit a similar pattern up to k \u2248 10 nm\u22121 compatible with that of HCV helicaseHM, whereas the power spectrum of HCV polymerase exhibits a different pattern from that of the other proteins compatible with that of Mouse kinase. This global pattern attests to the fact that the former proteins are helicases whereas the latter is a polymerase. A further inspection reveals details that distinguish among the helicases. In particular, we observe that the power spectra of the models HCV helicaseEM and HCV helicaseMD exhibit very similar patterns of peaks attesting to their similar binding state. Although HCV helicaseStrA is in a different binding state, its power spectrum exhibits the same level of similarities with both HCV models, with an anticipated HCV-like grouping of peaks specific to our data. For k > 1 nm\u22121, these three helicase proteins exhibit three strong peaks at k \u2248 2.3, 4.6, 7.3 nm\u22121. From the distance between peaks, we infer an average wavelength of \u03bb \u2248 2.5 nm. The power spectrum of HCV helicaseHM follows the same pattern PeerJ reviewing PDF | (v2013:06:567:0:1:NEW 7 Jun 2013) R ev ie w in g M an us cr ip t 5 Figure 4: Power spectrum of the molecular surfaces of the selected proteins. Power spectra of the molecular surfaces divided by the power spectra of the corresponding white-noise molecular surfaces. The symbols depict the power spectra and the error bars depict the error associated with the measurement. Top panel: Helicase proteins. Bottom panel: Non\u2013helicase proteins. The values of k are measured in nm\u22121. as that of HCV helicaseStrA shifted to smaller k with a varying relative phase which most of the time is close to \u03c0, with strong peaks at k \u2248 1.8, 3.6 nm\u22121. In comparison with the HCV models, the power spectrum of HCV helicaseStrB exhibits differences in the position of the peaks (found at k \u2248 2.7, 3.6, 5.5 nm\u22121) and in their amplitude ratios, which attest to the different treatment in HCV helicaseStrB from that in the HCV models. As k increases, we observe a gradual damping of the power of the helicases and an emerging tail reminiscent of shot noise in a Poisson power spectrum, more prominent in HCV helicaseHM and HCV helicaseStrB, which indicates the damping of the fluctuations about the mean value and thus the vanishing of the structural signal. This damping is most visible for k > 6 nm\u22121. This agrees with the observation above that sets the upper limit of k to the size of a typical cluster of aminoacids and hence sets the minimum distance below which correlations are not of biological interest nor can be reliably probed by X\u2013ray/NMR experiments. Conversely, the polymerases exhibit a predomi- nantly structureless power spectrum. In particular, Mouse kinase shows a flat spectrum which indicates an isotropic distribution of the power, whereas HCV polymerase shows an unstable area for 2 < k < 4 nm\u22121 and a strong peak at k \u2248 5.5 nm\u22121, then settling in a flat spectrum. conclusions : We presented a new method based on the Fourier analysis of protein molecular surfaces to extract functional information on proteins. For a selected set of proteins of HCV with different structural features, we first produced surfaces of the molecular electrostatic potential, as well as the corresponding white\u2013noise surfaces, and then computed their two-point correlation function in harmonic space (the power spectrum). We found that this method can distinguish between helicases and polymerases, as well as between different treatments in helicases. However, for the same treatment, this method failed to distinguish between different bound states in helicases. An immediate extension of this work is the application of this method to isolated structural subunits that form larger structures within proteins. Similarly sized subunits will have a strong signal in the same frequency range, which will add up in the protein power spectrum. Hence, we must first measure the contribution of each subunit separately and produce a catalogue of subunit signatures, so we can distinguish them in the combined signal when running similarity searches. By reducing the initial three\u2013dimensional information on the molecular surface to the one\u2013dimensional information on pairs of points at a fixed scale apart, this method allows for a fast similarity search. Further refinements in the similarity search will require methods that use information from higher\u2013order correlation functions, such as the correlation among three points at a fixed triangular configuration or its Fourier\u2013transformed (the bispectrum). (See Fig. 2 right panel for an illustration.) The bispectrum measures phase correlations among the modes and thus deviations from a Gaussian distribution. Our ultimate goal is to integrate higher\u2013order correlations and to apply the resulting method to the RCSB database so as to provide the biopharmaceutical and structural research communities with a novel and easily searchable reference without the three\u2013dimensional information compromising the speed of the calculation. This method aims to coalesce techniques, which have been extensively tested and used in other fields such as cosmology, into a fast and robust pipeline for the analysis and processing of very large, three\u2013dimensional biological datasets in an effort to speed up protein similarity searches. Acknowledgments CSC is funded by the FCT\u2013Lisbon, Grant no. SFRH/BPD/65993/2009. This work was sup- PeerJ reviewing PDF | (v2013:06:567:0:1:NEW 7 Jun 2013) R ev ie w in g M an us cr ip t 6 ported in part by the European Union (European Social Fund - ESF) and Greek national funds through the Operational Program \u201dEducation and Lifelong Learning\u201d of the National Strategic Reference Framework (NSRF) - Research Funding Program: \u201cThales\u201d. Investing in knowledge society through the European Social Fund. \u2217 Electronic address: cscarvalho@oal.ul.pt \u2020 Electronic address: dvlachakis@bioacademy.gr \u2021 Electronic address: gtsiliki@bioacademy.gr \u00a7 Electronic address: vasilis@ceid.upatras.gr \u00b6 Electronic address: skossida@bioacademy.gr [1] Dobson PD, Cai YD, Stapley BJ, Doig AJ (2004) Curr Med Chem 11: 2135\u20132142 [2] Illerg\u030aard K, Ardell DH, Elofsson A (2009) Proteins 15; 77(3): 499\u2013508. [3] Venkatraman V, Sael L and Kihara D (2009), Cell Biochem Biophys 54: 23\u201332. [4] Kolodny R, Koehl P, Levitt M (2005) J Mol Biol 346: 1173\u20131188. [5] Mayr G, Domingues FS, Lackner P (2007) BMC Struct Biol 7: 50. [6] Berbalk C, Schwaiger CS, Lackner P (2009) Protein Sci 18: 2027\u20132035. [7] Via A, Ferre\u0300 F, Brannetti B and Helmer-Citterich M (2000), Protein surface similarities: a survey methods to describe and compare protein surfaces [8] Kazhdan M, Funkhouser T and Rusinkiewicz S (2003), Proc 2003 Eurographics, 43: 156\u2013164. [9] Ritchie DW, Kozakov D and Vajda S (2008), Bioinformatics 24(17): 1865?1873. [10] Venkatraman V, Chakravarthy PR, and Kihara D (2009), J Cheminformatics 1: 19. [11] Vlachakis D, Tsiliki G, Tsagkrasoulis D, Carvalho CS, Megalooikonomou V, Kossida S (2012) EMBnet J. 18(1): 6\u20139. [12] Przytycka TM, Singh M, Slonim DK (2010) Brief Bioinform. 11 (1): 15\u201329. [13] Berger-Wolf TY, Przytycka TM, Singh M, Slonim DK (2010) Pac Symp Biocomput. 15: 120\u2013122. [14] Gire SK, Stremlau M, Andersen KG, Schaffner SF, Bjornson Z, Rubins K, Hensley L, McCormick JB, Lander ES, Garry RF, Happi C, Sabeti PC (2012) Science 9; 338(6108); 750\u2013752. [15] Elcock AH, Gabdoulline RR, Wade RC, McCammon JA. (1999) J Mol Biol. 291(1): 149\u2013162. [16] Sept D, Elcock AH, McCammon JA. (1999) J Mol Biol. 294(5): 1181\u20131189. [17] Wlodek ST, Shen T, McCammon JA. (2000) Biopolymers. 53(3): 265\u2013271. [18] Honig B, Nicholls A. (1995) Science 268(5214): 1144\u2013 1149. [19] Wong GC, Pollack L. (2010) Annu Rev Phys Chem. 61: 171\u2013189. [20] McCammon JA. (2009) Proc Natl Acad Sci USA. 106(19): 7683\u20137684. [21] Vlachakis D, Champeris\u2013Tsaniras S, Kossida S. (2012) Mol Biochem. 1(3): 144\u2013149. [22] Kandil S, Biondaro S, Vlachakis D, Cummins AC, Coluccia A, Berry C, Leyssen P, Neyts J and Brancale A. (2009) J Bioorg Med Chem Lett. 1; 19(11): 2935\u20132937. [23] Konecny R, Baker NA, McCammon JA (2012) Comput Sci Discov 26; 5(1) 015005. [24] Unni S, Huang Y, Hanson RM, Tobias M, Krishnan S, Li WW, Nielsen JE, Baker NA (2011) J Comput Chem. 32(7): 1488\u20131491. [25] www.poissonboltzmann.org/apbs [26] www.rcsb.org [27] Hess B, Kutzner C, van der Spoel D, Lindahl E (2008) J Chem Theory Comput 4: 435\u2013447. [28] Lindahl E, Hess B, van der Spoel D (2001) J Mol Model 7: 306\u2013317. [29] van der Spoel D, Lindahl E, Hess B, Groenhof G, Mark AE, et al. (2005) J Comput Chem 26: 1701\u201318. [30] Sellis D, Vlachakis D, Vlassi M (2009) Bioinform Biol Insights 3: 99\u2013102. [31] S\u0306ali A, Blundell TL (1993) J. Mol. Biol. 234, 779\u2013815. [32] Eswar N, John B, Mirkovic N, Fiser A, Ilyin VA, Pieper U, Stuart AC, Mart\u0301\u0131-Renom MA, Madhusudhan MS, Yerkovich B, S\u0306ali A (2003) Nucl Acids Res 31, 3375\u2013 3380. [33] Laskowski RA, Rullmannn JA, MacArthur MW, Kaptein R, Thornton JM (1996) J. Biomol. NMR 8, 477\u2013486. [34] Connolly ML (1983), J. Appl. Cryst. 16, 548\u2013558. [35] Peacock JA (1999) Cosmological Physics, CUP. PeerJ reviewing PDF | (v2013:06:567:0:1:NEW 7 Jun 2013) R ev ie w in g M an us cr ip t Figure 1 Figure 1: Surfaces of the electrostatic molecular po-tential. Left panel: Hepatitis C helicase protein PeerJ reviewing PDF | (v2013:06:567:0:1:NEW 7 Jun 2013) R ev ie w in g M an us cr ip t Figure 2 Figure 1: Surfaces of the electrostatic molecular potential. Righ panel: Hepatitis C polymerase protein. PeerJ reviewing PDF | (v2013:06:567:0:1:NEW 7 Jun 2013) R ev ie w in g M an us cr ip t Figure 3 Figure 2: Schematic representation of point con PeerJ reviewing PDF | (v2013:06:567:0:1:NEW 7 Jun 2013) R ev ie w in g M an us cr ip t gu-rations for correlations in harmonic space. Left panel. PeerJ reviewing PDF | (v2013:06:567:0:1:NEW 7 Jun 2013) R ev ie w in g M an us cr ip t Figure 4 Figure 2: Schematic representation of point con PeerJ reviewing PDF | (v2013:06:567:0:1:NEW 7 Jun 2013) R ev ie w in g M an us cr ip t gu-rations for correlations in harmonic space. Right panel. PeerJ reviewing PDF | (v2013:06:567:0:1:NEW 7 Jun 2013) R ev ie w in g M an us cr ip t Figure 5 Figure 3: Power spectrum of the molecular surfaces ofthe selected proteins. PeerJ reviewing PDF | (v2013:06:567:0:1:NEW 7 Jun 2013) R ev ie w in g M an us cr ip t Figure 6 Figure 4: Power spectrum of the molecular surfacesof the selected proteins.Upper. PeerJ reviewing PDF | (v2013:06:567:0:1:NEW 7 Jun 2013) R ev ie w in g M an us cr ip t Figure 7 Figure 4: Power spectrum of the molecular surfacesof the selected proteins. Lower. PeerJ reviewing PDF | (v2013:06:567:0:1:NEW 7 Jun 2013) R ev ie w in g M an us cr ip t molecular surfaces : Surfaces of the molecular electrostatic potential follow the nonlinear Poisson\u2013Boltzmann equation [23, 24]. We solved numerically for the electrostatic potential using the finite\u2013difference method as implemented in the APBS Software [25]. The potential was calculated on a Figure 1: Surfaces of the electrostatic molecular potential. Left panel: Hepatitis C helicase protein, Righ panel: Hepatitis C polymerase protein. The electrostatic potential is measured in eV, with range as shown in the corresponding colour bar. regular grid of size (65, 65, 65)1, with the grid\u2013fill\u2013by\u2013 solute parameter set to 80%. The dielectric constants of the solvent and the solute were set to 80.0 and 2.0, respectively. An ionic exclusion radius of 2.0 A\u030a, a solvent radius of 1.4 A\u030a and a solvent ionic strength of 0.145 M were applied. Default APBS charges and atomic radii were used. For our training set, we selected Hepatitis C viral (HCV) proteins with similar and dissimilar functions. We considered two HCV helicase proteins, namely the HCV helicase strain A (denoted by HCV helicaseStrA) and the HCV helicase strain B (denoted by HCV helicaseStrB), whose three-dimensional coordinates were obtained from the RCSB database [26] of X\u2013ray protein crystallography structures, with entry codes 1A1V and 8OHM respectively. Furthermore, we generated two simulations of the 1A1V protein crystal, namely the energy\u2013minimized version (denoted by HCV helicaseEM) and the molecular dynamics version (denoted by HCV helicaseMD). We also established a homology model of the HCV helicase (denoted by HCV helicaseHM) so that the in silico three\u2013 dimensional model of HCV were included in our training set. Energy minimization (EM) removes any residual geometrical strain from each molecular system, whereas molecular dynamics (MD) simulates a periodic cytoplasm\u2013like aqueous environment. Both EM and MD were performed with the Gromacs suite [27\u201329] through our previously developed graphical interface [30]. Molecular dynamics took place in a periodic environment, which was subsequently solvated with the simple point\u2013 charge water model using the truncated octahedron box 1 The size of the grid was kept small in order to speed up the calculation and reduce the computational load. It was tested to be suitable for this study, as higher detail would not change the surface by much while it would increase the computational load significantly. PeerJ reviewing PDF | (v2013:06:567:0:1:NEW 7 Jun 2013) R ev ie w in g M an us cr ip t 3 extending to 7 A\u030a from each molecule. Partial charges were applied and the molecular systems neutralized with counter\u2013ions as required. The temperature was set to 300 K, the pressure to 1 atm and the step size to 2 fs. The total run of each molecular complex was 50 ns, using constant number of atoms, volume and temperature (NVT) throughout the calculation in a canonical environment. The results of the MD simulations were collected in a molecular trajectory database for further analysis. The homology model was produced using Modeller [31, 32] and was evaluated using the Procheck utility [33]. This model was designed in order to include a computer modelled structure in our training set, which however shares high sequence identity with its template structure (approximately 90%). We also included two examples of non\u2013helicase proteins, namely the 1NB7 structure of the HCV polymerase (denoted by HCV polymerase) and the camp\u2013 dependent protein kinase (denoted by Mouse kinase), with RCSB database entry codes 1NB7 and 1ATP respectively. Whereas the HCV polymerase is a virus protein of the same species as the selected helicases, the Mouse kinase is a mouse protein which has a very different function both from the HCV helicase and from the HCV polymerase. In Fig. 1 we show surfaces of the electrostatic molecular potential for two HCV proteins, namely the helicase and the polymerase. The electrostatic potential is measured in eV. In these manuscript, we used the Connolly representation for the molecular surfaces [34]. Power function of the molecular surfaces Molecular surfaces contain information on a property of proteins along the three spatial dimensions. This property, in this case the values of the electrostatic potential, can be regarded as a field F (x) defined over points x on the surface. Functional information is encoded not only in the positions of the points but also in the correlations among points. The simplest correlation function that we can measure is that between pairs of points. The twopoint correlation function \u03be of the field F measures the convolution of the field over its complex conjugate (see e.g. Ref. [35]) \u03be(r) \u2261 \u3008F \u2217(x)F (x + r)\u3009 = 1 L3 \u222b d3x F \u2217(x)F (x + r).(1) The angle brackets indicate an averaging over the normalization volume, which here we take as the volume of the molecular surface, L3. We assume that the field has a flat geometry and can be decomposed in a Fourier expansion of plane waves F (x) = \u2211 k Fk exp[\u2212ik \u00b7 x], (2) where the wavenumber k relates with the frequency \u03bd by k = 2\u03c0/\u03bd. If the field has a curved geometry, then a Fourier expansion in spherical harmonics should be used instead. However, the difference between the two expansions only matters in scales of order the size of the molecular surface, which correspond to the smallest frequency. The smallest frequency is the zero\u2013mode in the Fourier expansion and describes a global offset. The two-point correlation function becomes \u03be(r) = \u2329\u2211 k \u2211 k\u2032 F \u2217kFk\u2032 exp[i(k \u2212 k \u2032) \u00b7 x] exp[\u2212ik\u2032 \u00b7 r] \u232a .(3) Since the molecular surface is closed, the field is periodic within the size of the surface, which restricts the allowed wavenumbers to the harmonic boundary condition kn = (n2\u03c0/L)e\u0302k, where n \u2208 {0, 1, ...} is the order of the Fourier modes. As a consequence, all the cross terms with k\u2032 6= k average to zero and the remaining sum is \u03be(r) = ( L 2\u03c0 )3 \u222b d3k |Fk|2 exp[\u2212ik \u00b7 r]. (4) Hence the correlation function is the Fourier transform of the power spectrum P (k) = |Fk|2. This relationship is known as the Wiener-Khinchin theorem. The power spectrum measures amplitude correlations among the modes, discarding however information on the phase. We proceed to compute the Fourier transform Fk of the molecular surface inferred over a regular grid. The Fourier\u2013transformed surface measures the amplitude of the plane waves whose combination reproduces the information on the original surface. The frequencies of the plane waves range from the frequency corresponding to the extension of the surface (i.e. to n = 1), up to the Nyquist frequency corresponding to twice the bin size of the grid (i.e. to n = N/2, where N is the number of bins along a direction of the grid). The size of the molecular surfaces ranges between 5\u22127 nm (Table I). The smallest spatial scale of biological interest is the size of a typical cluster of aminoacids, which is of order xball \u223c 0.3 nm. We choose this spatial scale for the size of the grid, so that the largest frequency scale that can be probed is of order kball \u223c 10 nm\u22121. Furthermore, we assume that the field is isotropic, i.e. that it does not have a preferential direction, so that the power spectrum depends only on the distance between each pair of points. (See Fig. 2 left panel for an illustration.) By assuming isotropy, we are discarding information on the direction. We proceed to take the ensemble average of P (k) so that the power at the mode k is the sum of the power at all the points on a sphere of radius k from the zero\u2013mode, resulting in a one-dimensional function P (k). In this way, we collapse the information on the three-dimensional field over the molecular surface onto a one-dimensional power spectrum over the wavenumbers of the Fourier\u2013transformed molecular surface. PeerJ reviewing PDF | (v2013:06:567:0:1:NEW 7 Jun 2013) R ev ie w in g M an us cr ip t 4 k P P 1 2 21 1 P P P 1 2 3 21 2 k k k 3 3 Figure 2: Schematic representation of point configurations for correlations in harmonic space. Left panel: The configuration of the two-point correlation function contains one free parameter, k12, which is the distance in harmonic space between the two points P1 and P2. Right panel: The configuration of the three-point correlation function contains two free parameters, e.g. k12 and k23, describing the distances in harmonic space respectively between P1 and P2, and between P2 and P3. The third parameter k13 is related to the former two by the triangle condition k12 + k23 + k31 = 0. HCV helicaseStrA [72.4,64.8,55.1] HCV helicaseEM [72.8,65.1,55.5] HCV helicaseMD [72.3,65.5,56.3] HCV helicaseHM [71.5,65.7,55.9] HCV helicaseStrB [61.9,69.6,61.7] HCV polymerase [59.0,77.7,65.0] Mouse kinase [52.5,69.0,48.8] Table I: Sizes of the molecular surfaces along the [x,y,z]\u2013directions. Sizes are measured in A\u030a. For a given k, we are sampling a distribution, which we assume to be Gaussian with mean value \u3008Fk\u3009 and variance \u2329 |Fk|2 \u232a = P (k), from which the Fourier coefficients Fk are drawn. Hence there is a fundamental uncertainty about the underlying variance, which depends on the number of coefficients sampled at a given k. Since the number of k\u2019s on a sphere of radius k scales as k2 and for any real field it holds that F\u2212k = Fk\u2217, where the asterisk stands for the complex conjugate, then the uncertainty scales as \u2206P (k)/P (k) = \u221a 2/k2. method :",
+    "url": "https://peerj.com/articles/186/reviews/",
+    "review_1": "Jim Caunt \u00b7 Oct 2, 2013 \u00b7 Academic Editor\nACCEPT\nDear Prof. Niggli,\n\nMany thanks for your response and clarifications. Many congratulations on your interesting and informative paper: I enjoyed reading it. Have an excellent day!",
+    "review_2": "Jim Caunt \u00b7 Sep 16, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nI agree with the reviewers that this submission is largely clearly written and the experiments justify the conclusions, and that it represents an advance of current knowledge. However, there is a consensus view that a few minor experimental and manuscript points would need to be addressed prior to acceptance for publication, which fall under two main categories.\n\n1. Experimental\nThough the data look convincing, I agree with reviewer 2 that the inclusion of another negative control comprising the use of two primary antibodies to proteins that do not interact would increase confidence in the data.\n2. Manuscript\nI also agree with reviewer 1 that, although the manuscript is largely well-written, it could do with a proof-read to weed out minor mistakes. Examples here include the one highlighted by reviewer 2, I also noticed:\nLine 42: remove 'anymore'.\nLine 48: replace 'allows detection also of' with 'also allows detection of'.\nLine 115: replace 'We now investigated' with 'In the present study, we have investigated' or 'Here, we have investigated'.\nI also think some detailed clarification of the image analysis would also be helpful, especially given this is the focus of the paper. For example, in the materials and methods, it says 100 cells per condition were evaluated, but in the text some of the numbers corresponding to total cell number analysed don't tally with this figure (e.g. a total of 261 cells analysed for n=3 in Fig. 4B). Similarly, I think you should indicate if the data are derived from a confocal stack or from a representative optical section (I'm assuming it's the latter, but I'm not sure).",
+    "review_3": "Reviewer 1 \u00b7 Sep 13, 2013\nBasic reporting\nNo comments.\nExperimental design\nNo comments.\nValidity of the findings\nNo comments.\nAdditional comments\nBaumann et al. examine pair-wise proximity of proteins, such as flotillin-1 and 2, phosphorylated ERM, PSGL-1 and PIPKI\u03b390, that are important for structuring the uropod during T-cell polarization. Previously, FRET and biochemical approaches have suggested associations among these proteins. However, FRET is based on over-expression while biochemical assays lack spatial information and are prone to post-lysis artifacts. The authors use the Proximity Ligation Assay to investigate associations of the endogenous proteins before and after T-cell stimulation. They find that flotillin-1/flotillin-2, flotillin-2/p-ERM, flotillin-2/PSGL-1 and PIPKI\u03b390/p-ERM complexes pre-exist in un-stimulated cells and become confined in the uropod upon stimulation. The results nicely confirm and extend previous findings.\n\nThe experiments were performed and interpreted well and presented clearly. While studies on the dynamics of these complexes will be needed to reveal the connections between the preformed complexes and the ones confined in the uropod , as it stands the work presented is an important first step. In addition, whether these associations are part of the same complex or separately co-exist represent an important future direction. For the time being, the authors may simply choose to further discuss these points.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Analysis of close associations of uropod-associated proteins in human T-cells using the proximity ligation assay (v0.1)\". PeerJ https://doi.org/10.7287/peerj.186v0.1/reviews/1",
+    "review_4": "Reviewer 2 \u00b7 Sep 9, 2013\nBasic reporting\nThe manuscript is generally well presented and easy to follow. I would recommend careful proof-reading since one or two mistakes are still present eg line 152 Fig1C should be 2C.\nExperimental design\nThe use of the proximity ligation assay (PLA) adds information on the co-localisations seen between flotilins, PSGL-1, ERM and PIPKIgamma90 in the uropod of T-cells. The data presented is convincing as far as it goes; the minimum controls are shown for the the PLA assays (leaving out one of the two antibodies) but it's still difficult to really assess what's required for the PLA assay to give a positive signal eg only beta-actin and flotillin2 are shown as an 'irrelevent pair' and they do show significant interaction.\nValidity of the findings\nThe data are robust with the required number of repetitions and statistical analysis.\nAdditional comments\nOverall, I think this manuscript makes a modest advance over previous work published by this group and others but does not move us on much conceptually in terms of how close associations between proteins in the uropod are created.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"Analysis of close associations of uropod-associated proteins in human T-cells using the proximity ligation assay (v0.1)\". PeerJ https://doi.org/10.7287/peerj.186v0.1/reviews/2",
+    "pdf_1": "https://peerj.com/articles/186v0.2/submission",
+    "pdf_2": "https://peerj.com/articles/186v0.1/submission",
+    "all_reviews": "Review 1: Jim Caunt \u00b7 Oct 2, 2013 \u00b7 Academic Editor\nACCEPT\nDear Prof. Niggli,\n\nMany thanks for your response and clarifications. Many congratulations on your interesting and informative paper: I enjoyed reading it. Have an excellent day!\nReview 2: Jim Caunt \u00b7 Sep 16, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nI agree with the reviewers that this submission is largely clearly written and the experiments justify the conclusions, and that it represents an advance of current knowledge. However, there is a consensus view that a few minor experimental and manuscript points would need to be addressed prior to acceptance for publication, which fall under two main categories.\n\n1. Experimental\nThough the data look convincing, I agree with reviewer 2 that the inclusion of another negative control comprising the use of two primary antibodies to proteins that do not interact would increase confidence in the data.\n2. Manuscript\nI also agree with reviewer 1 that, although the manuscript is largely well-written, it could do with a proof-read to weed out minor mistakes. Examples here include the one highlighted by reviewer 2, I also noticed:\nLine 42: remove 'anymore'.\nLine 48: replace 'allows detection also of' with 'also allows detection of'.\nLine 115: replace 'We now investigated' with 'In the present study, we have investigated' or 'Here, we have investigated'.\nI also think some detailed clarification of the image analysis would also be helpful, especially given this is the focus of the paper. For example, in the materials and methods, it says 100 cells per condition were evaluated, but in the text some of the numbers corresponding to total cell number analysed don't tally with this figure (e.g. a total of 261 cells analysed for n=3 in Fig. 4B). Similarly, I think you should indicate if the data are derived from a confocal stack or from a representative optical section (I'm assuming it's the latter, but I'm not sure).\nReview 3: Reviewer 1 \u00b7 Sep 13, 2013\nBasic reporting\nNo comments.\nExperimental design\nNo comments.\nValidity of the findings\nNo comments.\nAdditional comments\nBaumann et al. examine pair-wise proximity of proteins, such as flotillin-1 and 2, phosphorylated ERM, PSGL-1 and PIPKI\u03b390, that are important for structuring the uropod during T-cell polarization. Previously, FRET and biochemical approaches have suggested associations among these proteins. However, FRET is based on over-expression while biochemical assays lack spatial information and are prone to post-lysis artifacts. The authors use the Proximity Ligation Assay to investigate associations of the endogenous proteins before and after T-cell stimulation. They find that flotillin-1/flotillin-2, flotillin-2/p-ERM, flotillin-2/PSGL-1 and PIPKI\u03b390/p-ERM complexes pre-exist in un-stimulated cells and become confined in the uropod upon stimulation. The results nicely confirm and extend previous findings.\n\nThe experiments were performed and interpreted well and presented clearly. While studies on the dynamics of these complexes will be needed to reveal the connections between the preformed complexes and the ones confined in the uropod , as it stands the work presented is an important first step. In addition, whether these associations are part of the same complex or separately co-exist represent an important future direction. For the time being, the authors may simply choose to further discuss these points.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Analysis of close associations of uropod-associated proteins in human T-cells using the proximity ligation assay (v0.1)\". PeerJ https://doi.org/10.7287/peerj.186v0.1/reviews/1\nReview 4: Reviewer 2 \u00b7 Sep 9, 2013\nBasic reporting\nThe manuscript is generally well presented and easy to follow. I would recommend careful proof-reading since one or two mistakes are still present eg line 152 Fig1C should be 2C.\nExperimental design\nThe use of the proximity ligation assay (PLA) adds information on the co-localisations seen between flotilins, PSGL-1, ERM and PIPKIgamma90 in the uropod of T-cells. The data presented is convincing as far as it goes; the minimum controls are shown for the the PLA assays (leaving out one of the two antibodies) but it's still difficult to really assess what's required for the PLA assay to give a positive signal eg only beta-actin and flotillin2 are shown as an 'irrelevent pair' and they do show significant interaction.\nValidity of the findings\nThe data are robust with the required number of repetitions and statistical analysis.\nAdditional comments\nOverall, I think this manuscript makes a modest advance over previous work published by this group and others but does not move us on much conceptually in terms of how close associations between proteins in the uropod are created.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"Analysis of close associations of uropod-associated proteins in human T-cells using the proximity ligation assay (v0.1)\". PeerJ https://doi.org/10.7287/peerj.186v0.1/reviews/2\nReview 5: nan\nReview 6: \n",
+    "LLM_extracted_review": "[\"1. Review 1: Many congratulations on your interesting and informative paper: I enjoyed reading it.\\n\\n2. Review 2: \\n   - Experimental: The inclusion of another negative control comprising the use of two primary antibodies to proteins that do not interact would increase confidence in the data.\\n   - Manuscript: The manuscript could do with a proof-read to weed out minor mistakes. Specific examples include:\\n     - Line 42: remove 'anymore'.\\n     - Line 48: replace 'allows detection also of' with 'also allows detection of'.\\n     - Line 115: replace 'We now investigated' with 'In the present study, we have investigated' or 'Here, we have investigated'.\\n   - Clarification of the image analysis is needed, especially regarding the total cell number analyzed and whether the data are derived from a confocal stack or a representative optical section.\\n\\n3. Review 3: The experiments were performed and interpreted well and presented clearly. While studies on the dynamics of these complexes will be needed, the work presented is an important first step. Future directions include discussing whether these associations are part of the same complex or separately co-exist.\\n\\n4. Review 4: \\n   - Basic reporting: The manuscript is generally well presented and easy to follow, but careful proof-reading is recommended as one or two mistakes are present (e.g., line 152 Fig1C should be 2C).\\n   - Experimental design: The minimum controls shown for the PLA assays make it difficult to assess what's required for the PLA assay to give a positive signal.\\n   - Validity of the findings: The data are robust with the required number of repetitions and statistical analysis.\\n   - Overall, the manuscript makes a modest advance over previous work but does not significantly enhance our conceptual understanding of how close associations between proteins in the uropod are created.\"]"
+}

peerj_json_files/PeerJ_Json_137.json ADDED Viewed

	@@ -0,0 +1,18 @@

+{
+    "v1_Abstract": "We have shown previously that the raft-associated proteins flotillin-1 and -2 are rapidly recruited to the uropods of chemoattractant-stimulated human neutrophils and T-cells and are involved in cell polarization. Other proteins such as the adhesion receptor PSGL-1, the actin-membrane linker proteins ezrin/radixin/moesin (ERM) and the signaling enzyme phosphatidylinositol-4-phosphate 5-kinase type I gamma 90 (PIPKIgamma 90) also accumulate in the T-cell uropod. Using the in situ proximity ligation assay (PLA) we now have investigated putative close associations of these proteins in human freshly isolated T-cells before and after chemokine addition. The PLA allows in situ subcellular localization of close proximity of endogenous proteins at single-molecule resolution in fixed cells. It allows detection also of weaker and transient complexes that would not be revealed with coimmunoprecipitation approaches. We previously provided evidence for heterodimer formation of tagged flotillin-1 and -2 in T-cells before and after chemokine addition using fluorescence resonance energy transfer (FRET). We now confirm these findings using PLA for the endogenous flotillins in fixed human T-cells. Moreover, in agreement with the literature, our PLA findings confirm a close association of endogenous PSGL-1 and ERM proteins both in resting and chemokine-activated human T-cells. In addition, we provide novel evidence using the PLA for close associations of endogenous activated ERM proteins with PIPKIgamma90 and of endogenous flotillins with PSGL-1 in human T-cells, before and after chemokine addition.Our findings suggest that preformed clusters of these proteins coalesce in the uropod upon cell stimulation.",
+    "v1_col_introduction": "introduction  : Chemotactic migration of the highly motile T-cells is indispensable for the fulfillment of their physiological functions. T-cell polarization is a prerequisite for directional migration. Polarization of leukocytes requires segregation and activation of specific signaling and cytoskeletal molecules in the retracting rear (uropod) and motile forward moving part (front) of the cells. Localized positive feedback loops and inhibitory effects of front signaling pathways on rear signaling and vice versa are thought to reinforce this structural and biochemical polarization (Friedl &\n2\n3\n4\n5\n6\n7\n8\n9\n10\n11\n12\n13\n14\n15\n16\n17\n18\n19\n20\n21\n22\nPeerJ reviewing PDF | (v2013:08:748:1:0:NEW 1 Oct 2013)\nR ev ie w in g M an\nus cr ip t\nWeigelin, 2008; Bagorda & Parent, 2008). Plasma membrane microdomains (\u201crafts\u201d) have been implicated to stabilize polarity of migrating leukocytes (Manes & Viola, 2006).\nThe uropod is a plasma membrane protrusion located in the rear of migrating leukocytes that contains specific organelles along with cytoskeletal, adhesion and signaling proteins such as activated phosphorylated ezrin/radixin/moesin proteins (P-ERM), the adhesion receptor Pselectin glycoprotein-1 (PSGL-1) and phosphatidylinositol-4-phosphate 5-kinase type I (PIPKI)\n\u03b390 (Lokuta et al., 2007; Sanchez-Madrid & Serrador, 2009; Mathis et al., 2013). Flotillins,\nmembrane microdomain scaffolding proteins, are also enriched in leukocyte uropods and are involved in uropod formation (Ludwig et al., 2010; Rossy et al., 2009; Affentranger et al., 2011; Baumann et al., 2012). The uropod may be especially required for T-cell migration through constricted spaces (Soriano et al., 2011). We currently study the mechanisms of targeting of specific proteins to the T-cell uropod. Transfection of human freshly isolated T-cells with a dominant-negative mutant of flotillin-2 impaired cell polarization and uropod capping of endogenous flotillin-1, PSGL-1 and GFP-\ntagged PIPKI\u03b390, indicating a functional role of flotillins in structuring the uropod (Affentranger\net al., 2011; Mathis et al., 2013). Moreover, expression of constitutively active ezrin in freshly isolated T-cells induced capping of flotillins and PSGL-1, and that of a dominant-negative ezrin mutant impaired flotillin and PSGL-1 capping, suggesting a scaffolding role also for P-ERM\n(Martinelli et al., 2013). Similarly overexpression of PIPKI\u03b387, a naturally occurring isoform\nlacking the last 26 amino acids, which does not locate to the uropod, suppresses T-cell uropod formation and impairs capping of uropod proteins such as flotillins (Mathis et al., 2013).\nWe have now studied in situ protein-protein interactions in human T-cells fixed before and after chemokine addition, using the proximity ligation assay (PLA), in order to obtain insight into the molecular processes involved in T-cell uropod formation. The PLA allows in situ subcellular localization of close proximity of proteins at single-molecule resolution (S\u00f6derberg et al., 2006). It also allows detection of weaker and transient complexes that would not be revealed with coimmunoprecipitation approaches. In contrast to fluorescence resonance energy transfer (FRET), which involves expression of tagged proteins, PLA allows analysis of complexes of unmodified endogenous proteins. We analyzed selected interactions of T-cell uropod-located proteins for which high quality antibodies working well in immunofluorescence are available. We focused on\n23\n24\n25\n26\n27\n28\n29\n30\n31\n32\n33\n34\n35\n36\n37\n38\n39\n40\n41\n42\n43\n44\n45\n46\n47\n48\n49\n50\n51\n52\nPeerJ reviewing PDF | (v2013:08:748:1:0:NEW 1 Oct 2013)\nR ev ie w in g M an\nus cr ip t\nflotillins, PSGL-1, activated ERM proteins and PIPKI\u03b390. We provide novel data indicating close\nproximity of P-ERM and PIPKI\u03b390 and of flotillins and PSGL-1 in T-cells before and after\nchemokine addition. As expected from previous data (Ivetic & Ridley, 2004), PSGL-1 also\nclosely associates with P-ERM. Associations of flotillins with P-ERM or with PIPKI\u03b390 appear\nto be less extensive.",
+    "v2_Abstract": "We have shown previously that the raft-associated proteins flotillin-1 and -2 are rapidly recruited to the uropods of chemoattractant-stimulated human neutrophils and T-cells and are involved in cell polarization. Other proteins such as the adhesion receptor PSGL-1, the actin-membrane linker proteins ezrin/radixin/moesin (ERM) and the signaling enzyme phosphatidylinositol-4-phosphate 5-kinase type I gamma 90 (PIPKIgamma 90) also accumulate in the T-cell uropod. Using the in situ proximity ligation assay (PLA) we now have investigated putative close associations of these proteins in human freshly isolated T-cells before and after chemokine addition. The PLA allows in situ subcellular localization of close proximity of endogenous proteins at single-molecule resolution in fixed cells. It allows detection also of weaker and transient complexes that would not be revealed with coimmunoprecipitation approaches. We previously provided evidence for heterodimer formation of tagged flotillin-1 and -2 in T-cells before and after chemokine addition using fluorescence resonance energy transfer (FRET). We now confirm these findings using PLA for the endogenous flotillins in fixed human T-cells. Moreover, in agreement with the literature, our PLA findings confirm a close association of endogenous PSGL-1 and ERM proteins both in resting and chemokine-activated human T-cells. In addition, we provide novel evidence using the PLA for close associations of endogenous activated ERM proteins with PIPKIgamma90 and of endogenous flotillins with PSGL-1 in human T-cells, before and after chemokine addition.Our findings suggest that preformed clusters of these proteins coalesce in the uropod upon cell stimulation.",
+    "v2_col_introduction": "introduction  : Chemotactic migration of the highly motile T-cells is indispensable for the fulfillment of their physiological functions. T-cell polarization is a prerequisite for directional migration. Polarization of leukocytes requires segregation and activation of specific signaling and cytoskeletal molecules in the retracting rear (uropod) and motile forward moving part (front) of the cells. Localized positive feedback loops and inhibitory effects of front signaling pathways on rear signaling and vice versa are thought to reinforce this structural and biochemical polarization (Friedl &\n2\n3\n4\n5\n6\n7\n8\n9\n10\n11\n12\n13\n14\n15\n16\n17\n18\n19\n20\n21\n22\nPeerJ reviewing PDF | (v2013:08:748:0:0:NEW 28 Aug 2013)\nR ev ie w in g M an\nus cr ip t\nWeigelin, 2008; Bagorda & Parent, 2008). Plasma membrane microdomains (\u201crafts\u201d) have been implicated to stabilize polarity of migrating leukocytes (Manes & Viola, 2006).\nThe uropod is a plasma membrane protrusion located in the rear of migrating leukocytes that contains specific organelles along with cytoskeletal, adhesion and signaling proteins such as activated phosphorylated ezrin/radixin/moesin proteins (P-ERM), the adhesion receptor Pselectin glycoprotein-1 (PSGL-1) and phosphatidylinositol-4-phosphate 5-kinase type I (PIPKI)\n\u03b390 (Lokuta et al., 2007; Sanchez-Madrid & Serrador, 2009; Mathis et al., 2013). Flotillins,\nmembrane microdomain scaffolding proteins, are also enriched in leukocyte uropods and are involved in uropod formation (Ludwig et al., 2010; Rossy et al., 2009; Affentranger et al., 2011; Baumann et al., 2012). The uropod may be especially required for T-cell migration through constricted spaces (Soriano et al., 2011). We currently study the mechanisms of targeting of specific proteins to the T-cell uropod. Transfection of human freshly isolated T-cells with a dominant-negative mutant of flotillin-2 impaired cell polarization and uropod capping of endogenous flotillin-1, PSGL-1 and GFP-\ntagged PIPKI\u03b390, indicating a functional role of flotillins in structuring the uropod (Affentranger\net al., 2011; Mathis et al., 2013). Moreover, expression of constitutively active ezrin in freshly isolated T-cells induced capping of flotillins and PSGL-1, and that of a dominant-negative ezrin mutant impaired flotillin and PSGL-1 capping, suggesting a scaffolding role also for P-ERM\n(Martinelli et al., 2013). Similarly overexpression of PIPKI\u03b387, a naturally occurring isoform\nlacking the last 26 amino acids, which does not anymore locate to the uropod, suppresses T-cell uropod formation and impairs capping of uropod proteins such as flotillins (Mathis et al., 2013).\nWe have now studied in situ protein-protein interactions in human T-cells fixed before and after chemokine addition, using the proximity ligation assay (PLA), in order to obtain insight into the molecular processes involved in T-cell uropod formation. The PLA allows in situ subcellular localization of close proximity of proteins at single-molecule resolution (S\u00f6derberg et al., 2006). It allows detection also of weaker and transient complexes that would not be revealed with coimmunoprecipitation approaches. In contrast to fluorescence resonance energy transfer (FRET), which involves expression of tagged proteins, PLA allows analysis of complexes of unmodified endogenous proteins. We analyzed selected interactions of T-cell uropod-located proteins for which high quality antibodies working well in immunofluorescence are available. We focused on\n23\n24\n25\n26\n27\n28\n29\n30\n31\n32\n33\n34\n35\n36\n37\n38\n39\n40\n41\n42\n43\n44\n45\n46\n47\n48\n49\n50\n51\n52\nPeerJ reviewing PDF | (v2013:08:748:0:0:NEW 28 Aug 2013)\nR ev ie w in g M an\nus cr ip t\nflotillins, PSGL-1, activated ERM proteins and PIPKI\u03b390. We provide novel data indicating close\nproximity of P-ERM and PIPKI\u03b390 and of flotillins and PSGL-1 in T-cells before and after\nchemokine addition. As expected from previous data (Ivetic & Ridley, 2004), PSGL-1 also\nclosely associates with P-ERM. Associations of flotillins with P-ERM or with PIPKI\u03b390 appear\nto be less extensive.",
+    "v1_text": "materials and methods : Stromal cell derived factor 1 (SDF-1): Peprotech, Paris, France. Bovine serum albumin (BSA): Serva, Germany. Gey\u2019s solution contained 138 mM NaCl, 6 mM KCl, 100 \u03bcM EGTA, 1 mM Na2HPO4, 5 mM NaHCO3, 5.5 mM glucose and 20 mM Hepes (pH 7.4). results : Interaction of flotillin-1 and -2 in freshly isolated human T-cells analyzed with the proximity ligation assay. Endogenous flotillin-1 and -2 showed marked colocalization in resting and chemokine-stimulated freshly isolated human T-cells (Fig. 1), as shown previously (Affentranger et al., 2011; Baumann et al., 2012). Resting cells were mainly spherical, with a punctate, membrane-associated location of both flotillin-1 and -2. Upon stimulation of cells with the chemokine SDF-1 for 15 minutes, the majority of the cells polarized, correlating with exclusive location of both flotillins at the plasma membrane at the tips of the uropods (Fig. 1A). Our previous results using FRET in cells transfected with tagged flotillin-1 and -2 suggest that preformed flotillin-1 and -2 heterooligomers coalesce in chemokine-stimulated T-cells (Baumann et al., 2012). In the present study, we have investigated interactions of endogenous flotillin-1 and -2, using the PLA, as a positive control. The in situ PLA is based on fixed cells incubated with primary murine and rabbit antibodies reacting for example with flotillin-1 and -2 respectively, followed by incubation of samples with modified secondary antibodies reacting with murine or rabbit IgG antibodies. These secondary antibodies are conjugated with oligonucleotides (PLA probe minus and PLA probe plus). Annealing of the probes occurs when the target proteins are in close proximity (less than 30-40 nm distance), which then initiates the amplification. The amplicons can be detected as red dots by fluorescence microscopy (S\u00f6derberg et al., 2006). We obtained a strong PLA signal in the majority of the T-cells using primary murine and rabbit antibodies to flotillin-2 and -1 respectively. This signal was located randomly along the plasma membrane in resting cells (range: 1 - 11 dots per cell; mean: 4 \u00b1 1 dots per cell; 58 cells analyzed in 3 experiments), and aligned linearly around the entire border of the uropod (at least 4 dots per uropod) in 65% (n=2; 86 cells analyzed) of the stimulated cells, corresponding to the location of endogenous flotillins (Fig. 1B; top panels: lower magnification; lower panels; higher magnification). These data are in agreement with our FRET studies indicating heterooligomerization of tagged flotillin-1 and -2 (Baumann et al., 2012). Very few cells with one red dot corresponding to a positive PLA reaction per cell were detected when the samples were only incubated with the flotillin-1 antibody (Fig. 1C). 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 PeerJ reviewing PDF | (v2013:08:748:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t Interactions of P-ERM with PSGL-1 and of flotillins with PSGL-1 and P-ERM in T-cells studied using PLA We studied in situ interactions of endogenous flotillins with the adhesion receptors PSGL-1 and activated phosphorylated ERM (P-ERM) proteins, and of PSGL-1 with P-ERM in fixed human Tcells. Immunofluorescence pictures indeed show partial or extensive colocalization of PSGL-1 with P-ERM (Fig. 2A) and of flotillins with PSGL-1 (Fig. 3A) and P-ERM (Fig. 4A) in resting Tcells and in the uropod of stimulated T-cells. We now analysed whether these colocalizations correlate with close interactions using PLA in human T-cells. As a positive control we studied the well established direct interaction between PSGL-1 and P-ERM using primary antibodies specifically recognizing PSGL-1 and P-ERM respectively which work well in immunofluorescence (Fig. 2A). As expected from previous findings (Ivetic & Ridley, 2004), we obtained positive PLA signals for PSGL-1 and P-ERM in 94 \u00b1 2% of resting and 87 \u00b1 3% (n=3) of chemokine-activated cells (Fig. 2B). In resting cells the dots indicating close proximity of the proteins were randomly located at the cell periphery (range: 4 \u2013 20 dots per cell; mean: 12 \u00b1 1 dot per cell, analysed in 60 cells derived from 3 experiments). In stimulated cells the dots lined the entire border of the uropod in 55 \u00b1 5% (n=3) of the polarized PLA-positive cells (a total of 248 cells analysed). The remainder of the polarized PLA-positive cells featured 1\u20132 dots /uropod. A negative control where the samples were only incubated with the P-ERM antibody is shown in Fig. 2C. A positive PLA reaction was also observed for PSGL-1 and flotillin-2 in resting and chemokineactivated T-cells, confirming and extending the data obtained in human neutrophils using coimmunoprecipitation of flotillin-2 and PSGL-1 (Rossy et al., 2009). Here we obtained positive PLA signals in 83 \u00b1 2 % (n=4) of the resting cells and 88 \u00b1 2% (n=4) of the chemokinestimulated T-cells (Fig. 3B), with fluorescent dots located at the plasma membrane of the resting cells (range: 1- 11 dots per cell; mean: 4 \u00b1 1 dots per cell analysed in 30 cells derived from 3 experiments), and along the entire uropod border in 67% (n=2; 198 cells analysed) of the polarized, PLA-positive stimulated cells. The remainder of the polarized PLA-positive cells featured 1\u20132 dots /uropod. Negative controls with only the anti-PSGL-1 antibody are shown in Figure 3C. The PLA of flotillin-2 and P-ERM was also positive in 88 \u00b1 1% of the resting T-cells (range: 1 - 6 dots per cell; mean: 3 \u00b1 1 dots per cell analysed in 59 cells derived from 2 experiments), and in 54 \u00b1 8% of the stimulated cells. Especially in the stimulated cells the number of dots per cell was clearly lower as compared to the result obtained for PSGL-1 and flotillin, indicating weaker, 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 PeerJ reviewing PDF | (v2013:08:748:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t possibly transient interactions (Fig. 4B). Only 15 \u00b1 7% (n=3) of polarized, PLA-positive cells featured dots lining the border of the entire uropod (a total of 261 cells analysed). The remainder of the polarized PLA-positive cells featured 1\u20132 dots /uropod. Negative controls using only the anti-P-ERM antibody are shown in Figure 4C. As a negative control we also applied the PLA assay to \u03b2-actin and flotillin-2 (supplementary Fig. 1). \u03b2-Actin is mainly located in protrusions at the front of polarized T-cells. Only small amounts of \u03b2-actin are detectable in the uropod (supplementary Fig. 1A). Indeed we observed only in 22 \u00b1 5% of the resting cells (n=3) and in 33 \u00b1 5% (n=3) of the stimulated cells weakly positive PLA signals (1 \u00b1 0 dots/cell; analysed in 68 cells with positive PLA derived from 3 independent experiments) for \u03b2-actin and flotillin-2. Both the % of cells with positive PLA and the number of dots per cell are thus clearly lower for this pair of antibodies as compared to the data shown in Figures 1 \u2013 5.In stimulated cells these dots were located outside of the uropod in 67 out of 68 inspected cells with positive PLA, derived from 3 independent experiments(supplementary Fig. 1B). Considering the uropod, this is thus a control showing zero response. The few dots detected outside of the uropod may be explained by an incidental close contact of occasional flotillin molecules located at the plasma membrane outside of the uropod with the abundant \u03b2-actin. The negative control with only the anti- \u03b2-actin antibody is shown in the supplementary Figure 1C. Interactions of PIPKI\u03b390 with P-ERM studied using PLA GFP-tagged PIPKI\u03b390 accumulates in the uropod of murine T-cells (Lokuta et al., 2007; Mathis et al., 2013). We expressed an N-terminally HA-tagged PIPKI\u03b390 construct in human T-cells and observed colocalization of this construct with P-ERM in the uropod (Fig. 5A). We studied possible interactions of PIPKI\u03b390 with P-ERM using the PLA. As we have no highly specific antibodies to PIPKI\u03b390 available which work well in immunofluorescence staining, we transfected cells with PIPKI\u03b390 tagged with HA. The PLA was carried out with murine anti-HA antibodies and rabbit antibodies reacting with P-ERM. To visualize the transfected cells, we incubated the fixed cells with a FITC-tagged anti-murine IgG antibody that detects the HA antibody, together with the PLA probes. The results are shown in Fig. 5B (top panels: lower magnification; bottom panels: high magnification). We observe an extensive close association of PIPKI\u03b390 with P-ERM at the plasma membrane in 98 \u00b1 1% of the transfected resting cells (range: 9 \u2013 24 dots per cell; mean: 15 \u00b1 1 dots per cell derived from 23 cells in 4 experiments). Similarly the PLA was positive in 89 \u00b1 7% (n=3) of the transfected polarized cells, mostly 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 PeerJ reviewing PDF | (v2013:08:748:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t restricted to the uropods and lining the uropods in 75 \u00b1 7% (n=3) of the cells (Fig. 5B). The remainder of the polarized PLA-positive cells featured 1\u20132 dots /uropod. Dots indicating a positive PLA occurred exclusively in the transfected cells as can be appreciated from the top panels, where both transfected and untransfected cells are shown in the same picture (Fig. 5B). Negative controls carried out only with the anti-HA antibody are shown in Fig. 5C. A less extensive interaction was detected for PIPKI\u03b390 and flotillin-2 (data not shown). discussion : In polarizing T-cells, specific signaling molecules, adhesion receptors, raft-associated proteins and cytoskeletal proteins segregate into the uropod, a site where protrusion formation is locally suppressed, possibly by a membrane cytoskeletal scaffold (Sanchez-Madrid & Serrador, 2009). The exact mechanism of uropod scaffolding is as yet unknown. We have recently shown that flotillins as well as activated phosphorylated ERM proteins cooperate in T-cell uropod formation and that they mutually enhance their uropod capping (Martinelli et al., 2013). This could be explained by direct or indirect interactions of these proteins. We have now used the PLA approach to study close associations of uropod-located proteins. This assay is very sensitive and specific and provides a high signal to noise ratio. In contrast to FRET it can be used to study close proximity of endogenous unmodified proteins in situ. Positive results obtained with the PLA indicate either very close proximity of proteins for example in the same microdomains or direct interactions. The question arises whether the intensity of the PLA signal is proportional to the extent of protein association. This has been addressed in a study analyzing ErbB2 homoassociation in cell lines comparing FRET and PLA (Mocanu et al., 2011). The authors used flow cytometry for quantification of PLA, correlated to the extent of labeling with fluorophoreconjugated antibodies. They observed a non-linear dependence of the PLA signal on the extent of protein expression and association, whereas the FRET data showed a linear correlation. Steric hindrance between densely packed proximity probes may prevent enzymes from taking part in the amplification process. The authors recommend the PLA approach as a semiquantitative tool for measuring in situ protein associations (Mocanu et al., 2011). We thus did not attempt quantification of the PLA data by flow cytometry. Evaluation of the number of dots per cell is made difficult by the fact that coalescence of dots occurs, especially in the uropod but also in resting cells. The number of dots detected obviously will also depend on the expression levels of the interaction partners. Specificity of the reaction is indicated by the loss of signal when only one of the antibodies is used for PLA. Moreover we could detect no PLA signals in the uropod 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 PeerJ reviewing PDF | (v2013:08:748:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t when we used antibodies to flotillin-2 and \u03b2-actin, the latter protein being located mainly outside of the uropod (supplementary Fig. 1). Using FRET of differently tagged flotillin-1 and -2 expressed in human T-cells we previously showed that these proteins form heterodimers in both resting and chemokine-stimulated cells, and we can now confirm these findings for the endogenous flotillins using PLA (Fig. 1B). Another positive control is the positive PLA observed for P-ERM and PSGL-1 in resting and activated Tcells (Fig. 2B), in agreement with the literature indicating direct interaction of these proteins obtained by in vitro methods (Ivetic & Ridley, 2004). We now provide novel data using in situ PLA in fixed human T-cells on close proximity of specific uropod components. We show that flotillins appear to closely associate with PSGL-1 in resting and stimulated T-cells, the reaction being restricted to the border of the uropod in polarized cells, confirming and extending the data obtained previously in human neutrophils using co-immunoprecipitation (Rossy et al., 2009). Flotillins and P-ERM appear to interact less extensively. Flotillins could thus indirectly recruit P-ERM to the uropod via interactions with PSGL-1. Another T-cell uropod component is the enzyme PIPKI\u03b390, which synthesizes phosphatidylinositol-4,5-bisphosphate (PIP-2) and is involved in regulating uropod retraction in leukocytes (Lokuta et al., 2007; Mathis et al., 2013). PIP-2 is required for ERM activation, inducing a conformational change and allowing subsequent C-terminal phosphorylation (Niggli & Rossy, 2008). Uropod-localized PIPKI\u03b390 could thus result in locally enhanced PIP-2 production and ERM activation. Here we provide novel data suggesting that PIPKI\u03b390 and P- ERM are in close proximity in resting and activated T-cells. Interestingly, the isoform PIPKI\u03b2 has also been reported to be targeted to the uropod of polarized neutrophils and to interact in vitro via a 83 amino acid C-terminal domain with EBP50 (ERM-binding phosphoprotein 50). These 83 C- terminal amino acids are not homologous in PIPKI\u03b390 isoforms (Lacalle et al., 2007). PIPKI\u03b2 and PIPKI\u03b390 may thus directly or indirectly interact with ERM proteins via differently structured binding sites. In summary our data suggest either direct interactions of flotillins with PSGL-1, and of PIPKI\u03b390 with activated ERM proteins in resting and chemokine-activated T-cells and/or the presence of these uropod-located proteins in the same membrane microdomains. Preformed complexes of these proteins present already in resting cells could thus coalesce upon cell stimulation into the uropod. Whether P-ERM and PIPKI\u03b390 or flotillins and PSGL-1 indeed interact directly will 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 PeerJ reviewing PDF | (v2013:08:748:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t have to be verified with other techniques. If so, the binding domains involved in interactions of flotillins with PSGL-1 and of PIPKI\u03b390 with P-ERM, and the possible regulation of these interactions by posttranslational modifications will have to be explored. transient transfections of t-lymphocytes : For transfections, 3-6x106 freshly isolated T-lymphocytes were resuspended in 100 \u00b5l human T cell nucleofector solution (Amaxa, K\u00f6ln, Germany) diluted 1:2 with PBS and 1 \u03bcg of plasmid DNA was added. Then, the cell suspension was transferred to a cuvette and nucleofection was carried out (Amaxa Nucleofector, program U-14). Immediately, 500 \u03bcl of medium with 20% FCS was added and the cells were transferred to a prewarmed 12-well plate containing 2.5 ml of medium with 20% FCS, followed by incubation at 37\u00b0C in a CO2 incubator for 4 h. Transfected cells were subsequently washed, resuspended in Gey\u2019s solution and used for experiments. acknowledgments : We thank Dr. Erwin Sigel for critical reading of the manuscript, Dr. Anna Huttenlocher for a generous gift of the PIP5KI\u03b3661 construct and Dr. C. Chaponnier for a kind gift of the anti-\u03b2actin antibody. This work was supported by the Swiss National Science Foundation (Grant No 3100A_129655/1, to VN). 265 266 267 268 269 270 271 272 PeerJ reviewing PDF | (v2013:08:748:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t antibodies : Monoclonal murine antibodies directed against flotillin-2 (Cat. No. E35820) and PSGL-1 (Cat. No. 556053) were obtained from Transduction Laboratories/BD Pharmingen, Germany. A polyclonal rabbit antibody directed against phospho ezrin (Thr567)/ radixin (Thr564)/ moesin (Thr558) (Cat. No. 3141) was from Cell Signaling Technology. Polyclonal rabbit antibodies directed against flotillin-2 (Cat. No. F1680) or flotillin-1(Cat. No. F1180), a monoclonal murine antibody recognizing the hemaglutinin tag (HA) (clone HA-7) and a FITC-conjugated antibody directed against murine IgG (Cat. No. F5387) were obtained from Sigma. A monoclonal murine antibody specifically recognizing \u03b2-cytoplasmic actin was kindly provided by C. Chaponnier (Dugina et al., 2009). isolation of human t-lymphocytes : Resting T-lymphocytes were isolated from buffy coats of healthy donor blood using the Pan T Cell Isolation Kit II (Miltenyi Biotec) and separation on LS columns (Miltenyi Biotec) as described previously (Martinelli et al., 2013). The buffy coats were obtained from the Central 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 PeerJ reviewing PDF | (v2013:08:748:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t Laboratory of the Swiss Red Cross, Bern, Switzerland. The resulting cell suspension contained >95% T-lymphocytes as assessed using anti-CD3 staining. The cells were used after over night incubation in RPMI with 10% FCS at 37\u00b0C and 5% CO2. construct : For preparation of N-terminally HA-tagged PIPKI\u03b390, a construct coding for wild-type PIP5KI\u03b3661 N-terminally tagged with EGFP in a pcDNA3.1 vector (Lokuta et al., 2007) was used as a PCR template. Then the PCR product was cloned into the phCMV2 Vector (Genlantis). immunofluorescence staining : T-cells were incubated as described in the figure or table legends, followed by fixation with TCA or PFA and staining for the indicated proteins as described (Affentranger et al., 2011). proximity ligation assay : The in situ PLA (kit obtained from Olink Bioscience; www.olink.com) was used to detect protein-protein interactions in fixed cells. Nontransfected or transfected cells were treated without or with SDF-1, as described in the Figure legends, followed by fixation with TCA or PFA as described (Affentranger et al., 2011), incubation with primary antibodies, incubation with the PLA probes (anti murine and anti-rabbit IgG antibodies conjugated with oligonucleotides), ligation and amplification according to the manufacturer\u2019s instructions. Imaging was performed on fixed samples with a confocal laser scanning microscope Olympus Fluoview FV1000-IX81, 60x oil immersion objective. For determination of the fraction of cells with at least one red fluorescent dot per cell 100 cells were evaluated per sample and experiment by microscopical analysis. For the evaluation of the specific number of dots per cell and the number of dots per uropod, pictures of representative optical sections of cells with positive PLA were used (the numbers of the cells analyzed per experiment and condition are indicated in the text). 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 PeerJ reviewing PDF | (v2013:08:748:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t 100 cells per sample and experiment (mean \u00b1 sem of 3 experiments). note that the majority : of the cells incubated with both flo1 and flo2 antibodies exhibited several dots per cell,whereas for controls only incubated with flo1 antibody, maximally 1 dot per cell occurred. Bars, 10 \u00b5m. PeerJ reviewing PDF | (v2013:08:748:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:08:748:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t Figure 2 Interaction of PSGL-1 and P-ERM in human T-cells studied with PLA (A,B) T-cells were preincubated for 30 min at 37\u00b0C, followed by a further incubation for 15 min without or with 40 ng/ml SDF-1, fixation with TCA and staining for endogenous PSGL-1 (monoclonal murine antibody) and P-ERM (polyclonal rabbit antibody), followed by (A) fluorescently labeled anti-murine and anti-rabbit IgG second antibodies (IF) or (B) PLA probes minus and plus, ligation and amplification. (C) For negative controls,T-cells were treated as described for (B), except that the anti-PSGL-1 antibody was omitted. For (B) and (C), the top panels are overviews at lower magnification whereas in the lower panels single cells are shown at higher magnification. The pictures are representative of 3 experiments. The percentage of cells with one or more red fluorescent dots per cell, indicating positivePLA, was determined for 100 cells per sample and experiment (mean \u00b1 sem of 3experiments). Bars, 10 \u00b5m. PeerJ reviewing PDF | (v2013:08:748:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:08:748:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t Figure 3 Interaction of PSGL-1 and flotillin-2 in human T-cells studied with PLA (A,B) T-cells were preincubated for 30 min at 37\u00b0C, followed by a further incubation for 15 min without or with 40 ng/ml SDF-1, fixation with TCA and staining for endogenous PSGL-1 (monoclonal murine antibody) and flotillin-2 (polyclonal rabbit antibody), followed by (A) fluorescently labeled anti-murine and anti-rabbit IgG second antibodies (IF) or (B) PLA probes minus and plus, ligation and amplification. (C) For negative controls, T-cells were treated as described for (B), except that the anti-flotillin-2 antibody was omitted. For (B) and (C),the top panels are overviews at lower magnification whereas in the lower panels single cells are shown at higher magnification. The pictures are representative of 3 experiments. The percentage of cells with one or more red fluorescent dots per cell was determined for PeerJ reviewing PDF | (v2013:08:748:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:08:748:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t Figure 4 Interaction of P-ERM and flotillin-2 in human T-cells studied with PLA (A,B) T-cells were preincubated for 30 min at 37\u00b0C, followed by a further incubation for 15 min without or with 40 ng/ml SDF-1, fixation with TCA and staining for endogenous flotillin-2 (monoclonal murine antibody) and P-ERM (polyclonal rabbit antibody), followed by (A) fluorescently labeled anti-murine and anti-rabbit IgG second antibodies (IF) or (B) PLA probes minus and plus, ligation and amplification. (C) For negative controls, T-cells were treated as described for (B), except that the anti-flotillin-2 antibody was omitted. For (B) and (C), the top panels are overviews at lower magnification whereas in the lower panels single cells are shown at higher magnification. The pictures are representative of 3 experiments. The percentage of cells with one or more red fluorescent dots per cell was determined for PeerJ reviewing PDF | (v2013:08:748:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:08:748:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t Figure 5 Interaction of PIPKIgamma90 and P-ERM in human T-cells studied with PLA. (A-C) T-cells were transfected with HA-PIPKIgamma90. Four h later the T-cells were preincubated for 30 min at 37\u00b0C, followed by a further incubation for 15 min without or with 40 ng/ml SDF-1, fixation with TCA and staining for HA (monoclonal murine antibody) and P- ERM (polyclonal rabbit antibody), followed by incubation with (A) fluorescently labeled second anti-murine (red fluorescence) and anti-rabbit IgG antibodies (green fluorescence) or (B) FITC-conjugated anti-mouse IgG, in order to visualize the transfected cells (green fluorescence), and PLA probes minus and plus, ligation and amplification (red fluorescence). (C) For negative controls, transfected cells were treated as described for (B) except that the anti-P-ERM antibody was omitted. The pictures are representative of 3 experiments. For (B) and (C), the top panels are overviews at lower magnification whereas in the lower panels single cells are shown at higher magnification. The percentage of transfected cells with one or more red fluorescent dots per cell was determined for 100 cells per sample and experiment (mean \u00b1 sem for 3 experiments). Bars, 10 \u00b5m. PeerJ reviewing PDF | (v2013:08:748:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:08:748:1:0:NEW 1 Oct 2013) R ev ie w in g M an us cr ip t",
+    "v2_text": "materials and methods : Stromal cell derived factor 1 (SDF-1): Peprotech, Paris, France. Bovine serum albumin (BSA): Serva, Germany. Gey\u2019s solution contained 138 mM NaCl, 6 mM KCl, 100 \u03bcM EGTA, 1 mM Na2HPO4, 5 mM NaHCO3, 5.5 mM glucose and 20 mM Hepes (pH 7.4). results : Interaction of flotillin-1 and -2 in freshly isolated human T-cells analyzed with the proximity ligation assay. 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 PeerJ reviewing PDF | (v2013:08:748:0:0:NEW 28 Aug 2013) R ev ie w in g M an us cr ip t Endogenous flotillin-1 and -2 showed marked colocalization in resting and chemokine-stimulated freshly isolated human T-cells (Fig. 1), as shown previously (Affentranger et al., 2011; Baumann et al., 2012). Resting cells were mainly spherical, with a punctate, membrane-associated location of both flotillin-1 and -2. Upon stimulation of cells with the chemokine SDF-1 for 15 minutes, the majority of the cells polarized, correlating with exclusive location of both flotillins at the plasma membrane at the tips of the uropods (Fig. 1A). Our previous results using FRET in cells transfected with tagged flotillin-1 and -2 suggest that preformed flotillin-1 and -2 heterooligomers coalesce in chemokine-stimulated T-cells (Baumann et al., 2012). We now investigated interactions of endogenous flotillin-1 and -2, using the PLA, as a positive control. The in situ PLA is based on fixed cells incubated with primary murine and rabbit antibodies reacting for example with flotillin-1 and -2 respectively, followed by incubation of samples with modified secondary antibodies reacting with murine or rabbit IgG antibodies. These secondary antibodies are conjugated with oligonucleotides (PLA probe minus and PLA probe plus). Annealing of the probes occurs when the target proteins are in close proximity (less than 30-40 nm distance), which then initiates the amplification. The amplicons can be detected as red dots by fluorescence microscopy (S\u00f6derberg et al., 2006). We obtained a strong PLA signal in the majority of the T-cells using primary murine and rabbit antibodies to flotillin-2 and -1 respectively. This signal was located randomly along the plasma membrane in resting cells (range: 1 - 11 dots per cell; mean: 4 \u00b1 1 dots per cell; 58 cells analyzed in 3 experiments), and aligned linearly around the entire border of the uropod (at least 4 dots per uropod) in 65% (n=2; 86 cells analyzed) of the stimulated cells, corresponding to the location of endogenous flotillins (Fig. 1B; top panels: lower magnification; lower panels; higher magnification). These data are in agreement with our FRET studies indicating heterooligomerization of tagged flotillin-1 and -2 (Baumann et al., 2012). Very few cells with one red dot corresponding to a positive PLA reaction per cell were detected when the samples were only incubated with the flotillin-1 antibody (Fig. 1C). Interactions of P-ERM with PSGL-1 and of flotillins with PSGL-1 and P-ERM in T-cells studied using PLA We studied in situ interactions of endogenous flotillins with the adhesion receptors PSGL-1 and activated phosphorylated ERM (P-ERM) proteins, and of PSGL-1 with P-ERM in fixed human T- 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 PeerJ reviewing PDF | (v2013:08:748:0:0:NEW 28 Aug 2013) R ev ie w in g M an us cr ip t cells. Immunofluorescence pictures indeed show partial or extensive colocalization of PSGL-1 with P-ERM (Fig. 2A) and of flotillins with PSGL-1 (Fig. 3A) and P-ERM (Fig. 4A) in resting Tcells and in the uropod of stimulated T-cells. We now analysed whether these colocalizations correlate with close interactions using PLA in human T-cells. As a positive control we studied the well established direct interaction between PSGL-1 and P-ERM using primary antibodies specifically recognizing PSGL-1 and P-ERM respectively which work well in immunofluorescence (Fig. 2A). As expected from previous findings (Ivetic & Ridley, 2004), we obtained positive PLA signals for PSGL-1 and P-ERM in 94 \u00b1 2% of resting and 87 \u00b1 3% (n=3) of chemokine-activated cells (Fig. 2B). In resting cells the dots indicating close proximity of the proteins were randomly located at the cell periphery (range: 4 \u2013 20 dots per cell; mean: 12 \u00b1 1 dot per cell, analysed in 60 cells derived from 3 experiments). In stimulated cells the dots lined the entire border of the uropod in 55 \u00b1 5% (n=3) of the polarized PLA-positive cells (a total of 248 cells analysed). The remainder of the polarized PLA-positive cells featured 1\u20132 dots /uropod. A negative control where the samples were only incubated with the P-ERM antibody is shown in Fig. 1C. A positive PLA reaction was also observed for PSGL-1 and flotillin-2 in resting and chemokineactivated T-cells, confirming and extending the data obtained in human neutrophils using coimmunoprecipitation of flotillin-2 and PSGL-1 (Rossy et al., 2009). Here we obtained positive PLA signals in 83 \u00b1 2 % (n=4) of the resting cells and 88 \u00b1 2% (n=4) of the chemokinestimulated T-cells (Fig. 3B), with fluorescent dots located at the plasma membrane of the resting cells (range: 1- 11 dots per cell; mean: 4 \u00b1 1 dots per cell analysed in 30 cells derived from 3 experiments), and along the entire uropod border in 67% (n=2; 198 cells analysed) of the polarized, PLA-positive stimulated cells. The remainder of the polarized PLA-positive cells featured 1\u20132 dots /uropod. Negative controls with only the anti-PSGL-1 antibody are shown in Figure 3C. The PLA of flotillin-2 and P-ERM was also positive in 88 \u00b1 1% of the resting T-cells (range: 1 - 6 dots per cell; mean: 3 \u00b1 1 dots per cell analysed in 59 cells derived from 2 experiments), and in 54 \u00b1 8% of the stimulated cells. Especially in the stimulated cells the number of dots per cell was clearly lower as compared to the result obtained for PSGL-1 and flotillin, indicating weaker, possibly transient interactions (Fig. 4B). Only 15 \u00b1 7% (n=3) of polarized, PLA-positive cells featured dots lining the border of the entire uropod (a total of 261 cells analysed). The remainder of the polarized PLA-positive cells featured 1\u20132 dots /uropod. Negative controls using only the anti-P-ERM antibody are shown in Figure 4C. 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 PeerJ reviewing PDF | (v2013:08:748:0:0:NEW 28 Aug 2013) R ev ie w in g M an us cr ip t As a negative control we also applied the PLA assay to \u03b2-actin and flotillin-2 (supplementary Fig. 1). \u03b2-Actin is mainly located in protrusions at the front of polarized T-cells. Only small amounts of \u03b2-actin are detectable in the uropod (supplementary Fig. 1A). Indeed we observed only in 22 \u00b1 5% of the resting cells (n=3) and in 33 \u00b1 5% (n=3) of the stimulated cells weakly positive PLA signals (approximately 1 small dot/cell) for \u03b2-actin and flotillin-2. In stimulated cells the dots were always located outside of the uropod (supplementary Fig. 1B). The negative control with only the anti- \u03b2-actin antibody is shown in the supplementary Figure 1C. Interactions of PIPKI\u03b390 with P-ERM studied using PLA GFP-tagged PIPKI\u03b390 accumulates in the uropod of murine T-cells (Lokuta et al., 2007; Mathis et al., 2013). We expressed an N-terminally HA-tagged PIPKI\u03b390 construct in human T-cells and observed colocalization of this construct with P-ERM in the uropod (Fig. 5A). We studied possible interactions of PIPKI\u03b390 with P-ERM using the PLA. As we have no highly specific antibodies to PIPKI\u03b390 available which work well in immunofluorescence staining, we transfected cells with PIPKI\u03b390 tagged with HA. The PLA was carried out with murine anti-HA antibodies and rabbit antibodies reacting with P-ERM. To visualize the transfected cells, we incubated the fixed cells with a FITC-tagged anti-murine IgG antibody that detects the HA antibody, together with the PLA probes. The results are shown in Fig. 5B (top panels: lower magnification; bottom panels: high magnification). We observe an extensive close association of PIPKI\u03b390 with P-ERM at the plasma membrane in 98 \u00b1 1% of the transfected resting cells (range: 9 \u2013 24 dots per cell; mean: 15 \u00b1 1 dots per cell derived from 23 cells in 4 experiments). Similarly the PLA was positive in 89 \u00b1 7% (n=3) of the transfected polarized cells, mostly restricted to the uropods and lining the uropods in 75 \u00b1 7% (n=3) of the cells (Fig. 5B). The remainder of the polarized PLA-positive cells featured 1\u20132 dots /uropod. Dots indicating a positive PLA occurred exclusively in the transfected cells as can be appreciated from the top panels, where both transfected and untransfected cells are shown in the same picture (Fig. 5B). Negative controls carried out only with the anti-HA antibody are shown in Fig. 5C. A less extensive interaction was detected for PIPKI\u03b390 and flotillin-2 (data not shown). discussion : In polarizing T-cells, specific signaling molecules, adhesion receptors, raft-associated proteins and cytoskeletal proteins segregate into the uropod, a site where protrusion formation is locally suppressed, possibly by a membrane cytoskeletal scaffold (Sanchez-Madrid & Serrador, 2009). 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 PeerJ reviewing PDF | (v2013:08:748:0:0:NEW 28 Aug 2013) R ev ie w in g M an us cr ip t The exact mechanism of uropod scaffolding is as yet unknown. We have recently shown that flotillins as well as activated phosphorylated ERM proteins cooperate in T-cell uropod formation and that they mutually enhance their uropod capping (Martinelli et al., 2013). This could be explained by direct or indirect interactions of these proteins. We have now used the PLA approach to study close associations of uropod-located proteins. This assay is very sensitive and specific and provides a high signal to noise ratio. In contrast to FRET it can be used to study close proximity of endogenous unmodified proteins in situ. Positive results obtained with the PLA indicate either very close proximity of proteins for example in the same microdomains or direct interactions. The question arises whether the intensity of the PLA signal is proportional to the extent of protein association. This has been addressed in a study analyzing ErbB2 homoassociation in cell lines comparing FRET and PLA (Mocanu et al., 2011). The authors used flow cytometry for quantification of PLA, correlated to the extent of labeling with fluorophoreconjugated antibodies. They observed a non-linear dependence of the PLA signal on the extent of protein expression and association, whereas the FRET data showed a linear correlation. Steric hindrance between densely packed proximity probes may prevent enzymes from taking part in the amplification process. The authors recommend the PLA approach as a semiquantitative tool for measuring in situ protein associations (Mocanu et al., 2011). We thus did not attempt quantification of the PLA data by flow cytometry. Evaluation of the number of dots per cell is made difficult by the fact that coalescence of dots occurs, especially in the uropod but also in resting cells. The number of dots detected obviously will also depend on the expression levels of the interaction partners. Specificity of the reaction is indicated by the loss of signal when only one of the antibodies is used for PLA. Moreover we could detect no PLA signals in the uropod when we used antibodies to flotillin-2 and \u03b2-actin, the latter protein being located mainly outside of the uropod (supplementary Fig. 1). Using FRET of differently tagged flotillin-1 and -2 expressed in human T-cells we previously showed that these proteins form heterodimers in both resting and chemokine-stimulated cells, and we can now confirm these findings for the endogenous flotillins using PLA (Fig. 1B). Another positive control is the positive PLA observed for P-ERM and PSGL-1 in resting and activated Tcells (Fig. 2B), in agreement with the literature indicating direct interaction of these proteins obtained by in vitro methods (Ivetic & Ridley, 2004). We now provide novel data using in situ PLA in fixed human T-cells on close proximity of specific uropod components. We show that flotillins appear to closely associate with PSGL-1 in resting and stimulated T-cells, the reaction being restricted to the border of the uropod in 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 PeerJ reviewing PDF | (v2013:08:748:0:0:NEW 28 Aug 2013) R ev ie w in g M an us cr ip t polarized cells, confirming and extending the data obtained previously in human neutrophils using co-immunoprecipitation (Rossy et al., 2009). Flotillins and P-ERM appear to interact less extensively. Flotillins could thus indirectly recruit P-ERM to the uropod via interactions with PSGL-1. Another T-cell uropod component is the enzyme PIPKI\u03b390, which synthesizes phosphatidylinositol-4,5-bisphosphate (PIP-2) and is involved in regulating uropod retraction in leukocytes (Lokuta et al., 2007; Mathis et al., 2013). PIP-2 is required for ERM activation, inducing a conformational change and allowing subsequent C-terminal phosphorylation (Niggli & Rossy, 2008). Uropod-localized PIPKI\u03b390 could thus result in locally enhanced PIP-2 production and ERM activation. Here we provide novel data suggesting that PIPKI\u03b390 and P- ERM are in close proximity in resting and activated T-cells. Interestingly, the isoform PIPKI\u03b2 has also been reported to be targeted to the uropod of polarized neutrophils and to interact in vitro via a 83 amino acid C-terminal domain with EBP50 (ERM-binding phosphoprotein 50). These 83 C- terminal amino acids are not homologous in PIPKI\u03b390 isoforms (Lacalle et al., 2007). PIPKI\u03b2 and PIPKI\u03b390 may thus directly or indirectly interact with ERM proteins via differently structured binding sites. In summary our data suggest either direct interactions of flotillins with PSGL-1, and of PIPKI\u03b390 with activated ERM proteins in resting and chemokine-activated T-cells and/or the presence of these uropod-located proteins in the same membrane microdomains. Preformed complexes of these proteins present already in resting cells could thus coalesce upon cell stimulation into the uropod. Whether P-ERM and PIPKI\u03b390 or flotillins and PSGL-1 indeed interact directly will have to be verified with other techniques. If so, the binding domains involved in interactions of flotillins with PSGL-1 and of PIPKI\u03b390 with P-ERM, and the possible regulation of these interactions by posttranslational modifications will have to be explored. transient transfections of t-lymphocytes : For transfections, 3-6x106 freshly isolated T-lymphocytes were resuspended in 100 \u00b5l human T cell nucleofector solution (Amaxa, K\u00f6ln, Germany) diluted 1:2 with PBS and 1 \u03bcg of plasmid DNA was added. Then, the cell suspension was transferred to a cuvette and nucleofection was carried out (Amaxa Nucleofector, program U-14). Immediately, 500 \u03bcl of medium with 20% FCS was added and the cells were transferred to a prewarmed 12-well plate containing 2.5 ml of medium with 20% FCS, followed by incubation at 37\u00b0C in a CO2 incubator for 4 h. Transfected cells were subsequently washed, resuspended in Gey\u2019s solution and used for experiments. acknowledgments : We thank Dr. Erwin Sigel for critical reading of the manuscript, Dr. Anna Huttenlocher for a generous gift of the PIP5KI\u03b3661 construct and Dr. C. Chaponnier for a kind gift of the anti-\u03b2actin antibody. This work was supported by the Swiss National Science Foundation (Grant No 3100A_129655/1, to VN). 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 PeerJ reviewing PDF | (v2013:08:748:0:0:NEW 28 Aug 2013) R ev ie w in g M an us cr ip t antibodies : Monoclonal murine antibodies directed against flotillin-2 (Cat. No. E35820) and PSGL-1 (Cat. No. 556053) were obtained from Transduction Laboratories/BD Pharmingen, Germany. A polyclonal rabbit antibody directed against phospho ezrin (Thr567)/ radixin (Thr564)/ moesin (Thr558) (Cat. No. 3141) was from Cell Signaling Technology. Polyclonal rabbit antibodies directed against flotillin-2 (Cat. No. F1680) or flotillin-1(Cat. No. F1180), a monoclonal murine antibody recognizing the hemaglutinin tag (HA) (clone HA-7) and a FITC-conjugated antibody directed against murine IgG (Cat. No. F5387) were obtained from Sigma. A monoclonal murine antibody specifically recognizing \u03b2-cytoplasmic actin was kindly provided by C. Chaponnier (Dugina et al., 2009). isolation of human t-lymphocytes : Resting T-lymphocytes were isolated from buffy coats of healthy donor blood using the Pan T Cell Isolation Kit II (Miltenyi Biotec) and separation on LS columns (Miltenyi Biotec) as described previously (Martinelli et al., 2013). The buffy coats were obtained from the Central 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 PeerJ reviewing PDF | (v2013:08:748:0:0:NEW 28 Aug 2013) R ev ie w in g M an us cr ip t Laboratory of the Swiss Red Cross, Bern, Switzerland. The resulting cell suspension contained >95% T-lymphocytes as assessed using anti-CD3 staining. The cells were used after over night incubation in RPMI with 10% FCS at 37\u00b0C and 5% CO2. construct : For preparation of N-terminally HA-tagged PIPKI\u03b390, a construct coding for wild-type PIP5KI\u03b3661 N-terminally tagged with EGFP in a pcDNA3.1 vector (Lokuta et al., 2007) was used as a PCR template. Then the PCR product was cloned into the phCMV2 Vector (Genlantis). immunofluorescence staining : T-cells were incubated as described in the figure or table legends, followed by fixation with TCA or PFA and staining for the indicated proteins as described (Affentranger et al., 2011). proximity ligation assay : The in situ PLA (kit obtained from Olink Bioscience; www.olink.com) was used to detect protein-protein interactions in fixed cells. Nontransfected or transfected cells were treated without or with SDF-1, as described in the Figure legends, followed by fixation with TCA or PFA as described (Affentranger et al., 2011), incubation with primary antibodies, incubation with the PLA probes (anti murine and anti-rabbit IgG antibodies conjugated with oligonucleotides), ligation and amplification according to the manufacturer\u2019s instructions. Imaging was performed on fixed samples with a confocal laser scanning microscope Olympus Fluoview FV1000-IX81, 60x oil immersion objective. The amount of cells with at least one red fluorescent dot per cell was evaluated for 100 cells per sample and experiment. 100 cells per sample and experiment (mean \u00b1 sem of 3 experiments). note that the majority : of the cells incubated with both flo1 and flo2 antibodies exhibited several dots per cell,whereas for controls only incubated with flo1 antibody, maximally 1 dot per cell occurred. Bars, 10 \u00b5m. PeerJ reviewing PDF | (v2013:08:748:0:0:NEW 28 Aug 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:08:748:0:0:NEW 28 Aug 2013) R ev ie w in g M an us cr ip t Figure 2 Interaction of PSGL-1 and P-ERM in human T-cells studied with PLA (A,B) T-cells were preincubated for 30 min at 37\u00b0C, followed by a further incubation for 15 min without or with 40 ng/ml SDF-1, fixation with TCA and staining for endogenous PSGL-1 (monoclonal murine antibody) and P-ERM (polyclonal rabbit antibody), followed by (A) fluorescently labeled anti-murine and anti-rabbit IgG second antibodies (IF) or (B) PLA probes minus and plus, ligation and amplification. (C) For negative controls,T-cells were treated as described for (B), except that the anti-PSGL-1 antibody was omitted. For (B) and (C), the top panels are overviews at lower magnification whereas in the lower panels single cells are shown at higher magnification. The pictures are representative of 3 experiments. The percentage of cells with one or more red fluorescent dots per cell, indicating positivePLA, was determined for 100 cells per sample and experiment (mean \u00b1 sem of 3experiments). Bars, 10 \u00b5m. PeerJ reviewing PDF | (v2013:08:748:0:0:NEW 28 Aug 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:08:748:0:0:NEW 28 Aug 2013) R ev ie w in g M an us cr ip t Figure 3 Interaction of PSGL-1 and flotillin-2 in human T-cells studied with PLA (A,B) T-cells were preincubated for 30 min at 37\u00b0C, followed by a further incubation for 15 min without or with 40 ng/ml SDF-1, fixation with TCA and staining for endogenous PSGL-1 (monoclonal murine antibody) and flotillin-2 (polyclonal rabbit antibody), followed by (A) fluorescently labeled anti-murine and anti-rabbit IgG second antibodies (IF) or (B) PLA probes minus and plus, ligation and amplification. (C) For negative controls, T-cells were treated as described for (B), except that the anti-flotillin-2 antibody was omitted. For (B) and (C),the top panels are overviews at lower magnification whereas in the lower panels single cells are shown at higher magnification. The pictures are representative of 3 experiments. The percentage of cells with one or more red fluorescent dots per cell was determined for PeerJ reviewing PDF | (v2013:08:748:0:0:NEW 28 Aug 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:08:748:0:0:NEW 28 Aug 2013) R ev ie w in g M an us cr ip t Figure 4 Interaction of P-ERM and flotillin-2 in human T-cells studied with PLA (A,B) T-cells were preincubated for 30 min at 37\u00b0C, followed by a further incubation for 15 min without or with 40 ng/ml SDF-1, fixation with TCA and staining for endogenous flotillin-2 (monoclonal murine antibody) and P-ERM (polyclonal rabbit antibody), followed by (A) fluorescently labeled anti-murine and anti-rabbit IgG second antibodies (IF) or (B) PLA probes minus and plus, ligation and amplification. (C) For negative controls, T-cells were treated as described for (B), except that the anti-flotillin-2 antibody was omitted. For (B) and (C), the top panels are overviews at lower magnification whereas in the lower panels single cells are shown at higher magnification. The pictures are representative of 3 experiments. The percentage of cells with one or more red fluorescent dots per cell was determined for PeerJ reviewing PDF | (v2013:08:748:0:0:NEW 28 Aug 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:08:748:0:0:NEW 28 Aug 2013) R ev ie w in g M an us cr ip t Figure 5 Interaction of PIPKIgamma90 and P-ERM in human T-cells studied with PLA. (A-C) T-cells were transfected with HA-PIPKIgamma90. Four h later the T-cells were preincubated for 30 min at 37\u00b0C, followed by a further incubation for 15 min without or with 40 ng/ml SDF-1, fixation with TCA and staining for HA (monoclonal murine antibody) and P- ERM (polyclonal rabbit antibody), followed by incubation with (A) fluorescently labeled second anti-murine (red fluorescence) and anti-rabbit IgG antibodies (green fluorescence) or (B) FITC-conjugated anti-mouse IgG, in order to visualize the transfected cells (green fluorescence), and PLA probes minus and plus, ligation and amplification (red fluorescence). (C) For negative controls, transfected cells were treated as described for (B) except that the anti-P-ERM antibody was omitted. The pictures are representative of 3 experiments. For (B) and (C), the top panels are overviews at lower magnification whereas in the lower panels single cells are shown at higher magnification. The percentage of transfected cells with one or more red fluorescent dots per cell was determined for 100 cells per sample and experiment (mean \u00b1 sem for 3 experiments). Bars, 10 \u00b5m. PeerJ reviewing PDF | (v2013:08:748:0:0:NEW 28 Aug 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:08:748:0:0:NEW 28 Aug 2013) R ev ie w in g M an us cr ip t",
+    "url": "https://peerj.com/articles/187/reviews/",
+    "review_1": "Jonathan Eisen \u00b7 Oct 4, 2013 \u00b7 Academic Editor\nACCEPT\nI have reviewed your revisions and also sent the paper out to the prior reviewers. One of the reviewers responded and believes your revisions are acceptable. The other reviewer has not responded. As I and one of the reviews believe your revisions make the paper acceptable AND as I believe it is unnecessary to get the other reviewers comments, I am going to recommend accepting it.",
+    "review_2": "Jonathan Badger \u00b7 Oct 4, 2013\nBasic reporting\nI have compiled and run the new version of phybin successfully on my test set. As my primary concerns in the previous version were technical, I believe they have now been addressed.\nExperimental design\nNo further comments in this section.\nValidity of the findings\nThe additional supplementary table on which compares nearest, average, and furthest neighbor on the Wolbachia dataset is a good addition to the manuscript\nCite this review as\nBadger JH (2013) Peer Review #1 of \"PhyBin: binning trees by topology (v0.2)\". PeerJ https://doi.org/10.7287/peerj.187v0.2/reviews/1",
+    "review_3": "Jonathan Eisen \u00b7 Aug 22, 2013 \u00b7 Academic Editor\nMAJOR REVISIONS\nThe reviewers comments all seem reasonable to me and should be addressed. In particular I think anything you can add on comparing PhyBin to other tree clustering methods would be very useful.",
+    "review_4": "Toni Gabald\u00f3n \u00b7 Aug 22, 2013\nBasic reporting\nThe paper presents a new utility to bin trees according to their topologies. They showcase the methods with an example of an analysis of Wolbachia gene tree sets. The method is publicly available and worked properly in my hands.\n\nIt is unclear to me whay the method includes a step of tree rooting. In fact RF distance can be computed on unrooted trees. Is this just a matter of computing optimization? is it possible that two topologies are rendered different solely because of a different root is used?. These issues should be clarified.\n\nSimilarly, Does the method enable trees including multifurcations as an input?\n\nThe paper reports the number of bins in their Wolbachia example, and mention that there is one majoritary topology, but what fraction of trees fall within the majoritary topology is not mentioned. It would be useful to provide the data of how many trees fall in each bin.\nExperimental design\nRather than a primary research article, this is a methodological article. Although the method does not enable any fundamentally new type of analyses (such type of analyses could be done by concatenating a tree comparison tool, building a matrix and clustering), I think that the tool will be of interest to many users that want a simple and fast solution to the problem of tree-binning.\n\nThe tool enables collapsing clades based on branch lenghts. This is definitely useful, but I wonder why the method does not enable collapsing trees based on clade support values (e.g. bootstrap). That I guess would be a general need among potential users.\n\nOne clear limitation of the method is that it is only meant to compare trees that share identical number of labels, and contain one-to-one orthologs. As exemplified by the ~500 trees in their Wolbachia dataset this is a minimal fraction of the entire topological diversity (trees with paralogs, or missing genes are not included). Perhaps the authors could comment that this would be solvable by decomposing trees with duplication nodes with solutions such as the treeKO algorithm (Marcet-Houben et. al. http://www.ncbi.nlm.nih.gov/pubmed/21335609).\nValidity of the findings\nThe paper includes an example of using the program on real data. The biological implications of the observed binning (\"interaction with the host drives alternative evolution pattern) are not really highly supported and would need further work, but I think the example works fine to show the potential use of the utility.\n\nIt is somehow a pity that the comparison with other methodologies is restricted to the speed of RF computation, rather than with the bining itself. An idea is perhaps to compare how bins derived from PhyBin compare to other tree clusters (e,.g. Puigbo et. al. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3123530/) perhaps in other, already published datasets).\nCite this review as\nGabald\u00f3n T (2013) Peer Review #1 of \"PhyBin: binning trees by topology (v0.1)\". PeerJ https://doi.org/10.7287/peerj.187v0.1/reviews/1",
+    "pdf_1": "https://peerj.com/articles/187v0.2/submission",
+    "pdf_2": "https://peerj.com/articles/187v0.1/submission",
+    "review_5": "Jonathan Badger \u00b7 Aug 6, 2013\nBasic reporting\nThis is a short and straightforward manuscript about a potentially useful tree binning/clustering tool. If the technical issues I encountered (described below) are dealt with, I would support the publication of this manuscript.\n\nThe authors acknowledge that they aren't the only people to have tried to deal with tree clustering, and they adequately explain the motivation for why their tool was created.\n\nIn terms of language, the only problem I encountered was the ungrammatical \"can utilized Robinson-Fould's\" in the abstract.\nExperimental design\nAs this manuscript describes a bioinformatic tool, there isn't an experimental design as such, but they do give a test case showing an application of tree binning in an Wolbachia orthologous gene set.\nValidity of the findings\nWhile I have not tested the speed of the algorithm myself, it did seem to run quickly (on the data where I ran it successfully; see below), so the findings that their tool runs faster than most others seems plausible.\nAdditional comments\nI initially tried to run the binary version of phybin 0.2.11 on Mac OSX 10.8.4. I had a set of 397 trees of orthologous genes from 7 deltaproteobacterial species that I thought would be a good test case. However, I encountered the error \"openFile: resource exhausted (Too many open files)\". I created a subset of 50 tree files and ran it again -- it managed to go further but finally died with \"Waiting for asynchronous tasks to finish... phybin: Error running utility program: Error messages from neato:\" (there were no messages after that line).\n\nI did manage to successfully run the program on these trees on Linux, but the OSX version should be fixed. I also tried compiling the code myself with cabal but got the error: \"Bio/Phylogeny/PhyBin.hs:45:18: Could not find module `Bio.Phylogeny.PhyBin.PreProcessor'\", but I confess I know little about Haskell and this might have an obvious solution. Some tips for compilation on the homepage might help.\nCite this review as\nBadger JH (2013) Peer Review #2 of \"PhyBin: binning trees by topology (v0.1)\". PeerJ https://doi.org/10.7287/peerj.187v0.1/reviews/2",
+    "all_reviews": "Review 1: Jonathan Eisen \u00b7 Oct 4, 2013 \u00b7 Academic Editor\nACCEPT\nI have reviewed your revisions and also sent the paper out to the prior reviewers. One of the reviewers responded and believes your revisions are acceptable. The other reviewer has not responded. As I and one of the reviews believe your revisions make the paper acceptable AND as I believe it is unnecessary to get the other reviewers comments, I am going to recommend accepting it.\nReview 2: Jonathan Badger \u00b7 Oct 4, 2013\nBasic reporting\nI have compiled and run the new version of phybin successfully on my test set. As my primary concerns in the previous version were technical, I believe they have now been addressed.\nExperimental design\nNo further comments in this section.\nValidity of the findings\nThe additional supplementary table on which compares nearest, average, and furthest neighbor on the Wolbachia dataset is a good addition to the manuscript\nCite this review as\nBadger JH (2013) Peer Review #1 of \"PhyBin: binning trees by topology (v0.2)\". PeerJ https://doi.org/10.7287/peerj.187v0.2/reviews/1\nReview 3: Jonathan Eisen \u00b7 Aug 22, 2013 \u00b7 Academic Editor\nMAJOR REVISIONS\nThe reviewers comments all seem reasonable to me and should be addressed. In particular I think anything you can add on comparing PhyBin to other tree clustering methods would be very useful.\nReview 4: Toni Gabald\u00f3n \u00b7 Aug 22, 2013\nBasic reporting\nThe paper presents a new utility to bin trees according to their topologies. They showcase the methods with an example of an analysis of Wolbachia gene tree sets. The method is publicly available and worked properly in my hands.\n\nIt is unclear to me whay the method includes a step of tree rooting. In fact RF distance can be computed on unrooted trees. Is this just a matter of computing optimization? is it possible that two topologies are rendered different solely because of a different root is used?. These issues should be clarified.\n\nSimilarly, Does the method enable trees including multifurcations as an input?\n\nThe paper reports the number of bins in their Wolbachia example, and mention that there is one majoritary topology, but what fraction of trees fall within the majoritary topology is not mentioned. It would be useful to provide the data of how many trees fall in each bin.\nExperimental design\nRather than a primary research article, this is a methodological article. Although the method does not enable any fundamentally new type of analyses (such type of analyses could be done by concatenating a tree comparison tool, building a matrix and clustering), I think that the tool will be of interest to many users that want a simple and fast solution to the problem of tree-binning.\n\nThe tool enables collapsing clades based on branch lenghts. This is definitely useful, but I wonder why the method does not enable collapsing trees based on clade support values (e.g. bootstrap). That I guess would be a general need among potential users.\n\nOne clear limitation of the method is that it is only meant to compare trees that share identical number of labels, and contain one-to-one orthologs. As exemplified by the ~500 trees in their Wolbachia dataset this is a minimal fraction of the entire topological diversity (trees with paralogs, or missing genes are not included). Perhaps the authors could comment that this would be solvable by decomposing trees with duplication nodes with solutions such as the treeKO algorithm (Marcet-Houben et. al. http://www.ncbi.nlm.nih.gov/pubmed/21335609).\nValidity of the findings\nThe paper includes an example of using the program on real data. The biological implications of the observed binning (\"interaction with the host drives alternative evolution pattern) are not really highly supported and would need further work, but I think the example works fine to show the potential use of the utility.\n\nIt is somehow a pity that the comparison with other methodologies is restricted to the speed of RF computation, rather than with the bining itself. An idea is perhaps to compare how bins derived from PhyBin compare to other tree clusters (e,.g. Puigbo et. al. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3123530/) perhaps in other, already published datasets).\nCite this review as\nGabald\u00f3n T (2013) Peer Review #1 of \"PhyBin: binning trees by topology (v0.1)\". PeerJ https://doi.org/10.7287/peerj.187v0.1/reviews/1\nReview 5: Jonathan Badger \u00b7 Aug 6, 2013\nBasic reporting\nThis is a short and straightforward manuscript about a potentially useful tree binning/clustering tool. If the technical issues I encountered (described below) are dealt with, I would support the publication of this manuscript.\n\nThe authors acknowledge that they aren't the only people to have tried to deal with tree clustering, and they adequately explain the motivation for why their tool was created.\n\nIn terms of language, the only problem I encountered was the ungrammatical \"can utilized Robinson-Fould's\" in the abstract.\nExperimental design\nAs this manuscript describes a bioinformatic tool, there isn't an experimental design as such, but they do give a test case showing an application of tree binning in an Wolbachia orthologous gene set.\nValidity of the findings\nWhile I have not tested the speed of the algorithm myself, it did seem to run quickly (on the data where I ran it successfully; see below), so the findings that their tool runs faster than most others seems plausible.\nAdditional comments\nI initially tried to run the binary version of phybin 0.2.11 on Mac OSX 10.8.4. I had a set of 397 trees of orthologous genes from 7 deltaproteobacterial species that I thought would be a good test case. However, I encountered the error \"openFile: resource exhausted (Too many open files)\". I created a subset of 50 tree files and ran it again -- it managed to go further but finally died with \"Waiting for asynchronous tasks to finish... phybin: Error running utility program: Error messages from neato:\" (there were no messages after that line).\n\nI did manage to successfully run the program on these trees on Linux, but the OSX version should be fixed. I also tried compiling the code myself with cabal but got the error: \"Bio/Phylogeny/PhyBin.hs:45:18: Could not find module `Bio.Phylogeny.PhyBin.PreProcessor'\", but I confess I know little about Haskell and this might have an obvious solution. Some tips for compilation on the homepage might help.\nCite this review as\nBadger JH (2013) Peer Review #2 of \"PhyBin: binning trees by topology (v0.1)\". PeerJ https://doi.org/10.7287/peerj.187v0.1/reviews/2\nReview 6: \n",
+    "LLM_extracted_review": "[\"Review 1: Jonathan Eisen \u00b7 Oct 4, 2013 \u00b7 Academic Editor  \\n- One reviewer believes revisions are acceptable.  \\n- Other reviewer has not responded.  \\n- Recommend accepting the paper.\\n\\nReview 2: Jonathan Badger \u00b7 Oct 4, 2013  \\n- Technical concerns from previous version have been addressed.  \\n- Additional supplementary table comparing neighbors on the Wolbachia dataset is a good addition.\\n\\nReview 3: Jonathan Eisen \u00b7 Aug 22, 2013 \u00b7 Academic Editor  \\n- Reviewers' comments should be addressed.  \\n- Suggest adding comparisons of PhyBin to other tree clustering methods.\\n\\nReview 4: Toni Gabald\u00f3n \u00b7 Aug 22, 2013  \\n- Method includes a step of tree rooting; clarification needed on its necessity.  \\n- Uncertainty about handling multifurcations as input.  \\n- Lack of data on the fraction of trees within the majority topology.  \\n- Method does not enable collapsing trees based on clade support values.  \\n- Limitation: only compares trees with identical labels and one-to-one orthologs.  \\n- Suggests potential solutions for trees with duplication nodes.  \\n- Biological implications of observed binning need further support.  \\n- Comparison with other methodologies is limited to speed rather than binning.\\n\\nReview 5: Jonathan Badger \u00b7 Aug 6, 2013  \\n- Manuscript is straightforward but has technical issues to address.  \\n- Language issue noted in the abstract.  \\n- No experimental design as it describes a bioinformatic tool.  \\n- Tool seems to run quickly based on observations.  \\n- Encountered errors while running on Mac OSX; successful on Linux.  \\n- Compilation issues noted; suggestions for improvement on the homepage would be helpful.\"]"
+}

peerj_json_files/PeerJ_Json_138.json ADDED Viewed

	@@ -0,0 +1,16 @@

+{
+    "v1_Abstract": "A major goal of many evolutionary analyses is to determine the true evolutionary history of an organism.  Molecular methods that rely on the phylogenetic signal generated by a few to a handful of loci can be used to approximate the evolution of the entire organism but fall short of providing a global, genome\u00adwide, perspective on evolutionary processes.  Indeed, individual genes in a genome may have different evolutionary histories.  Therefore, it is informative to analyze the number and kind of phylogenetic topologies found within an orthologous set of genes across a genome.  Here we present PhyBin: a flexible program for clustering gene trees based on topological structure.  PhyBin can generate bins of topologies corresponding to exactly identical trees or can utilize Robinson\u00adFould\u2019s distance matrices to generate clusters of similar trees, using a user\u00addefined threshold.  Additionally, PhyBin allows the user to adjust for potential noise in the dataset (as may be produced when comparing very closely related organisms) by pre\u00adprocessing trees to collapse very short branches or those nodes not meeting a defined bootstrap threshold. As a test case, we generated individual trees based on an orthologous gene set from 10 Wolbachia species across four different supergroups (A\u00adD) and utilized PhyBin to categorize the complete set of topologies produced from this dataset.  Using this approach, we were able to show that although a single topology generally dominated the analysis, confirming the separation of the supergroups, many genes supported alternative evolutionary histories.  Because PhyBin\u2019s output provides the user with lists of gene trees in each topological cluster, it can be used to explore potential reasons for discrepancies between phylogenies including homoplasies, long\u00adbranch attraction, or horizontal gene transfer events. Availability: PhyBin is a standalone open\u00adsource program available: http://hackage.haskell.org/package/phybin Introduction: The advent of genomic sequencing has produced a large amount of data available for phylogenetic analysis and many researchers have attempted to utilize the phylogenetic signal found across the bacterial genome to develop species trees (Daubin, Gouy et al. 2001; Sicheritz\u00adPonten and Andersson 2001; Daubin, Moran et al. 2003; Bapteste, Boucher et al. 2004; Zhaxybayeva, Gogarten et al. 2006; Ellegaard, Klasson et al. 2013).  What has become clear from these analyses is that significant fractions of bacterial genomes do not follow the evolutionary history of their resident genome (Bapteste, Boucher et al. 2004).  These rogue genes are potentially undergoing evolutionary processes distinct from those felt by the rest of the resident genome or have arrived there via horizontal gene transfer events.  In order, then, to understand the evolution of the genome, it would be useful to achieve an understanding of the evolution of each gene in the genome.  Previous work by Sicheritz\u00adPonten and Andersson presented scripts combined the existing utilities BLAST, Clustalw, Paup 4.0* to provide a complete pipeline from genome to tree\u00ad binning analysis (Sicheritz\u00adPonten and Andersson 2001). These kinds of complete solutions are 1",
+    "v2_Abstract": "A major goal of many evolutionary analyses is to determine the true evolutionary history of an organism. Molecular methods that rely on the phylogenetic signal generated by a few to a handful of loci can be used to approximate the evolution of the entire organism but fall short of providing a global, genome-wide, perspective on evolutionary processes. Indeed, individual genes in a genome may have different evolutionary histories. Therefore, it is informative to analyze the number and kind of phylogenetic topologies found within an orthologous set of genes across a genome. Here we present PhyBin: a flexible program for clustering gene trees based on topological structure. PhyBin can generate bins of topologies corresponding to exactly identical trees or can utilize Robinson-Fould\u2019s distance matrices to generate clusters of similar trees, using a user-defined threshold. Additionally, PhyBin allows the user to adjust for potential noise in the dataset (as may be produced when comparing very closely related organisms) by pre-processing trees to collapse very short branches. As a test case, we generated individual trees based on an orthologous gene set from 10 Wolbachia species across four different supergroups (A-D) and utilized PhyBin to categorize the complete set of topologies produced from this dataset. Using this approach, we were able to show that although a single topology generally dominated the analysis, confirming the separation of the supergroups, many genes supported alternative evolutionary histories. Because PhyBin\u2019s output provides the user with lists of gene trees in each topological cluster, it can be used to explore potential reasons for discrepancies between phylogenies including homoplasies, long-branch attraction, or horizontal gene transfer events. Availability: PhyBin is a standalone open-source program available: http://hackage.haskell.org/package/phybin Introduction: The advent of genomic sequencing has produced a large amount of data available for phylogenetic analysis and many researchers have attempted to utilize the phylogenetic signal found across the bacterial genome to develop species trees (Daubin, Gouy et al. 2001; Sicheritz-Ponten and Andersson 2001; Daubin, Moran et al. 2003; Bapteste, Boucher et al. 2004; Zhaxybayeva, Gogarten et al. 2006; Ellegaard, Klasson et al. 2013). What has become clear from these analyses is that significant fractions of bacterial genomes do not follow the evolutionary history of their resident genome (Bapteste, Boucher et al. 2004). These rogue genes are potentially undergoing evolutionary processes distinct from those felt by the rest of the 1",
+    "v1_text": "introduction: : The advent of genomic sequencing has produced a large amount of data available for phylogenetic analysis and many researchers have attempted to utilize the phylogenetic signal found across the bacterial genome to develop species trees (Daubin, Gouy et al. 2001; SicheritzPonten and Andersson 2001; Daubin, Moran et al. 2003; Bapteste, Boucher et al. 2004; Zhaxybayeva, Gogarten et al. 2006; Ellegaard, Klasson et al. 2013). What has become clear from these analyses is that significant fractions of bacterial genomes do not follow the evolutionary history of their resident genome (Bapteste, Boucher et al. 2004). These rogue genes are potentially undergoing evolutionary processes distinct from those felt by the rest of the resident genome or have arrived there via horizontal gene transfer events. In order, then, to understand the evolution of the genome, it would be useful to achieve an understanding of the evolution of each gene in the genome. Previous work by SicheritzPonten and Andersson presented scripts combined the existing utilities BLAST, Clustalw, Paup 4.0* to provide a complete pipeline from genome to tree binning analysis (SicheritzPonten and Andersson 2001). These kinds of complete solutions are 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 PeerJ reviewing PDF | (v2013:07:679:1:1:NEW 23 Sep 2013) R ev ie w in g M an us cr ip t convenient but constrain the user to the specific utilities chosen by the authors for alignment and phylogeny generation. Here we present PhyBin, a computer program aimed at binning precomputed sets of non reticulated trees in Newick format, a file format produced by the majority of tree building software. PhyBin is a utility rather than a complete solution; it can serve as a component in many genomics pipelines, and provides a useful addition to the landscape of tools for dissecting and visualizing large numbers of trees. After the user applies their chosen ortholog prediction and treebuilding algorithms, PhyBin offers a quick way to visualize and browse the different evolutionary histories, either binned by topology and sorted by bin size, or in the form of a full hierarchical clustering based on RobinsonFoulds distance: i.e. a tree of trees. Method and Implementation: Generating orthologous sets and input trees Genomic sequences were downloaded from NCBI Microbial Genome Projects. The Wobachia species complex is made up of several major clades, called supergroups, designated by alphabetical letters (Baldo and Werren 2007). Accession numbers for the genomes analyzed here include: wUni and wVitA (submissions pending to genbank\u2019s ncbi), wBm (NC_006833.1), wPipPel (NC_010981.1), wHa (NC_021089.1), wRi (NC_012416.1), wMel (NC_002978.6), wNo (NC_021084.1), wAlbB (CAGB00000000.1), wBm (NC_006833.1), wOo (NC_018267.1). Orthologous gene sets were determined by Reciprocal Smallest Distance (RSD) algorithm (Wall, Fraser et al. 2003) with a 103 cutoff for significance threshold and alignment length threshold of 80%. Orthologs were then aligned using ClustalW (Larkin, Blackshields et al. 2007) and ML trees were generated using RAxML (Stamatakis 2006). The Newick format trees that resulted were used as input to PhyBin. The number of orthologous genes identified in this manner across all 10 taxa was 503. Description of the Program: PhyBin is a standalone commandline program, portable across all major operating systems. It runs in batchmode and is easily usable from scripts. PhyBin has two major modes: it can run very quickly and classify identical tree topologies into bins, or it can compute the distance (Robinson and Foulds 1981) between all pairs of trees and use that distance matrix to produce a configurable clustering of trees. Fast Binning Mode The key algorithm PhyBin performs in this mode is tree normalization, computing a rooted, ordered normal form for all inputs (which are labeled, unrooted, unordered tree topologies). Previous work in this area has described a number of viable normal forms (Chi, Yang et al. 2005). Conversion to a normal form ensures that all equivalent unrooted trees are converted into the same rooted tree, with a canonical root chosen. After conversion, the rooted trees are much faster to compare for equality than the unrooted trees would be, which enables fast binning. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 PeerJ reviewing PDF | (v2013:07:679:1:1:NEW 23 Sep 2013) R ev ie w in g M an us cr ip t PhyBin chooses the following strategy: it attempts to order subtrees by weight (number of tree nodes) and select the root node which is most balanced by weight (not depth)that is, which minimizes the maximum weight of any child of the root. Node labels are used only to ``break ties'' between equally weighted subtrees, or equally balanced roots. Because input trees in Newick format are typically labeled only on the leaves (taxa), PhyBin generates labels for intermediate nodes in the tree by creating a set of all the leaves contained in that subtree, given a root to determine up/down direction. This set can be represented as a bitvector and is also a key ingredient of computing RobinsonFoulds distance, which relies on identifying all such subsets (i.e. bipartitions induced by the tree). With labels for all nodes, equally weighted subtrees are ordered by label, and ties between potential roots are broken by comparing the labels of their children. Once input trees are normalized, testing for equality of two trees is as simple as comparing their representation in memory (a single, linear traversal). Normalization itself appears expensive due to the cost of labeling interior nodes with all leaves under them (O(N * I) for N taxa and I interior nodes), compounded by the fact that each intermediate node may have to consider each of its neighbors as a possible root and relabel itself b times in a tree of maximum branching factor b, yielding an O(N*I*b) asymptotic cost. However, in binning mode PhyBin runs much faster in the average case. One feature that enables PhyBin's efficiency is that it computes tree metadata interior labels and ``balanced'' ratingslazily, that is, on demand. Only when ``tie breaking'' is necessary between equallyweighted subtrees is an interior label computed at all. Likewise, only nodes ``near the center'' of the unrooted tree need to be considered for root status, those near the leaves need never be scored for balance. After normalization, PhyBin performs binning, which amounts to inserting all normalized trees into a data structure indexed by tree topology. We define a total order over normalized trees (made possible by labels), and thereby represent the table of bins as a sizebalanced binary tree supporting O(log(n)) insertion times. A hashtable would be an alternative, but the tree representation allows us to insert trees into the table without evaluating (forcing) unnecessary interior labels in the normal forms, whereas hashing requires traversing the entirety of each normalized tree to compute its hash. When execution completes, the contents of each bin are written out to disk, in addition to a visualization of a representative average tree for that topology, computed by averaging branch lengths of the bin members. PreProcessing Data PhyBin helps users extract a clean dataset and detect problems with the data, such as trees with mismatching numbers of taxa. In order to facilitate comparisons across trees with different taxon names (i.e. gene names), PhyBin can extract portions of designations or use a separate table of rules for mapping genes to taxa. In addition, PhyBin can restrict its analyses to a subset of taxon, ignoring others (prune). A problem with the simple binning approach is that it is fragile to minor differences in trees caused by noise (e.g. short length branches with high variability). This becomes increasingly problematic with large numbers of taxa, especially when closely related taxa (different strains) are compared. Fortunately, a simple preprocessing step that addresses this problem: PhyBin provides an option to collapse branches under two different conditions, a length threshold (for 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 PeerJ reviewing PDF | (v2013:07:679:1:1:NEW 23 Sep 2013) R ev ie w in g M an us cr ip t example, a length threshold of 0.01 would collapse all branches less than 0.01, in their place inserting a star topology) or a bootstrap support threshold (such that nodes with less than that threshold would be collapsed and the branch lengths from the taxa to the parent node would be added). Full Clustering Mode using RobinsonFoulds Distance Matrix: PhyBin reimplements the HashRF algorithm for full alltoall Robinson Foulds distance (Sul and Williams 2007), which is significantly faster than computing the distance matrix with repeated comparison of individual trees (e.g. PAUP (Swofford and Sullivan 2009)). The HashRF algorithm is fast for today\u2019s data sizes (e.g. hundreds of taxa and thousands of trees), but is scales much more poorly than the basic binning algorithm at significantly larger sizes. Because ortholog sets across different genomic comparisons will produce trees with different taxon memberships (as a result of paralogs or gene losses), a user may consider decomposing their trees with other software solutions (such as treeKO, (MarcetHouben and Gabaldon 2011)). Further, PhyBin is also capable of directly comparing these trees with different numbers of taxa using the leaf pruning method implemented in STRAW (Shaw, Ruan et al. 2013). Specifically, in comparing trees with different taxa (tolerant mode), the program first removes taxa that are not contained within each tree. If the taxon removed is in a polytomy, the parent and sister taxon are unchanged. However, in a binary node, taxon pruning would remove the intermediate node, retaining the branch lengths from the ancestor to the unpruned taxon. The \u2013tolerant mode comes with a cost, however, as the more efficient HashRF algorithm cannot be used; instead Phybin falls back to the earlier PAUPstyle algorithm. A distance matrix alone is not directly useful for exploring the direct relationships between different gene trees. Thus, PhyBin uses the RobinsonFoulds distance matrix to compute a clustering of tree topologies, similar to the output of the simple binning mode, but able to identify trees that are merely similar, although not identical. A hierarchical clustering method is used. (If the user desires a different clustering method, they may use the distance matrix produced by PhyBin as input to a different processing pipeline.) With the hierarchical clustering method, there remain several clustering options to configure. The choice of clustering options can dramatically alter bin membership (Supplementary Table 1), and running with several different options is a good way to get a sense for the range of possible outcomes. Specifically, the user may define the edit distance tolerated within clusters by providing a threshold, and may choose single, complete, or UPGMA linkage for clustering. Also if desired, rather than viewing a flat clustering of trees, the user may directly view a hierarchical clustering of the trees as a dendrogram. We believe PhyBin is the first program to date to provide this treeoftrees output. Output Formats: PhyBin is meant to be used in scripts and by other programs. Every output produced by PhyBin goes into a separate, simple text filefor example, the consensus tree for each cluster and the 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 PeerJ reviewing PDF | (v2013:07:679:1:1:NEW 23 Sep 2013) R ev ie w in g M an us cr ip t RobinsonFoulds distance matrix. Visualizations are produced separately and automatically in PDF files. performance: : There are very large differences in performance between existing programs for computing RobinsonFould\u2019s distance matrices. The fundamental datastructures in this problem domain are sets and finite maps, for which there are many alternate representations (bit vectors, hash tables, balanced trees, etc), providing a large space of possible implementations to explore. The sharpest contrast is between those programs that directly compare individual pairs of trees (PAUP, DendroPy), vs. those that insert all tree\u2019s bipartitions into a global structure and summarize it as a separate phase (e.g. HashRF). The later approach achieves much better cache locality. PhyBin is written in a very high level language, Haskell, which supports radical forms of optimization, including safe semiautomatic parallelism. PhyBin uses purely functional (immutable) datastructures for representing trees and their bipartitons; in particular it relies heavily on the balancedtree implementations Data.Map and Data.Set from the standard library. Nevertheless, when computing a matrix for a 150taxa, 100tree test (Table 1), PhyBin is 82 times faster than Philip (ANSI C) and 47.5 times faster than DendroPy (Python). However, PhyBin is still slower than HashRF by a factor of 2.8X4.8X. HashRF was the first implementation that introduced highperformance techniques for RF matrices, and it introduced the algorithm on which PhyBin\u2019s implementation is based. Unfortunately, the more widely used software (PAUP, DendroPy, Philip, etc), remains slow. HashRF, the currently available fast alternative, is delicate and must be used carefully (for example, an extra character of whitespace in the input file results in a segmentation fault with no error message in version 6.0.1). Additionally, because HashRF provides only the core RF distance computation, other tools are required for a biologist to be able to derive any conclusions from the output. As a final note on performance, PhyBin was straightforward to parallelize (using our \u201cLVar\u201d parallelism library) and achieves a 2.54X parallel speedup at four cores, and peaks at a 3.11X speedup at eight cores, making it a bit faster than HashRF on our target platform (Table 1). Future work will focus on reducing contention on shared data structures to improve scaling. Results and Discussion: We used PhyBin to identify how many phylogenies within the Wolbachia orthologous gene set support the supergroup divisions proposed by multilocus sequence typing (Baldo and Werren 2007). For comparative purposes in this analysis, a phylogeny for these 10 taxa was created using the concatenated, orthologous gene set (Figure 1A). In actuality, PhyBin does not require an expectation for tree topology and searches through tree space for distinct topological categories. As an illustration of PhyBin\u2019s ability to reduce the noise in a dataset produced by small branch lengths (i.e., closely related taxa), we used the program in binning mode on the set of Wolbachia orthologs under increasing branch length thresholds (Table 2). We chose a threshold of 0.01 for our dataset as the average branch length over the entire set of validated trees was 0.04 with minimum and maximum branch lengths of 0 and 2.31, respectively. Using this threshold, in 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 PeerJ reviewing PDF | (v2013:07:679:1:1:NEW 23 Sep 2013) R ev ie w in g M an us cr ip t binning mode, the largest bin contains a topology that agrees with that of the published supergroups (133 members in largest bin, 175 total bins, Table 2, Figure 1B). However, 174 other potential topologies exist in the dataset with 129 alternative topologies supported by only a single ortholog tree (Table 2). In order to better explore this tree set, we took advantage of PhyBin\u2019s ability to generate a distance matrix for all trees. By calculating the RobinsonFoulds (RF) distance between all trees, we can better assess the differences between clusters in the tree dataset. For example, by increasing the RFdistance threshold to 2 and using the averageneighbor clustering algorithm to group our trees, the number of clusters drops dramatically to only 77 with the largest cluster containing a majority (72%) of genes. Again, this topology agrees with the published supergroup data and our result from the binning approach (Figure 1C). Increasing the RFdistance threshold further provides increasing stringency in the detection of aberrant phylogenies \u2013 topologies not falling into the largest cluster at larger distance thresholds are likely to represent genes of interest in comparing evolutionary trajectories of these supergroups. To test this hypothesis, we identified those Wolbachia genes that continue to display alternative evolutionary histories (that is, falling outside of the majority) even when clustering trees using increasingly large RF distances (Figure 2B, Table 3). As expected, a large number of distinct topologies are not inconsistent with the supergroup clades (65 distinct tree clusters do not support the major topology, using an RFdistance threshold of 1 and a branch length cutoff of 0.02, Table 3, Figure 2B). We further investigated the ortholog set supporting the dissolution of supergroup A (Table 4). Interestingly, a majority of these orthologs are predicted to be secreted (using the Effective database predictions of sec signal or eukaryotic domains (Jehl, Arnold et al. 2011), suggesting that perhaps interaction with the host would drive some of these orthologs in a different evolutionary direction compared to their resident genome. Another test of PhyBin\u2019s ability to detect orthologs under different evolutionary pressures would focus on the Wolbachia prophage, a mobile genetic element known to undergo horizontal transmission between strains (Bordenstein and Wernegreen 2004; Chafee, Funk et al. 2010; Kent and Bordenstein 2010; Kent, Salichos et al. 2011). However, these phage orthologs do not occur across all of our 10 taxa included here and are therefore not suitable for testing support for the supergroups. In conclusion, we PhyBin is a new software program that efficiently and quickly groups phylogenies either by strict topological congruence or by clustering using RF distance. We believe that this tool, due to its ease of use, its speed, and informative output, will be of interest to evolutionary biologists and bioinformaticians alike. Figure Legends: Figure 1. In each of two modes (full clustering and binning) PhyBin is able to correctly recover the expected topology for the Wolbachia pipientis orthologs used herein. (A) Concatenated phylogeny based on 508 genes (using RAxML GTRGAMMA, bootstrap support based on 10,000 replicates). The four major supergroups are highlighted and denoted. (B) These same groups are recovered when PhyBin is run in either binning mode or (C) full clustering mode. Figure 2. RobinsonFoulds distance matricies produced by PhyBin are also visualized as a dendrogram by the software. (A) A tree of trees for the Wolbachia ortholog set (508 trees), clustered using an edit distance of 0, where identical topologies (nodes \u2013 grey ovals) are shown connected by a red line. Length of the branches connecting each node is proportional to the RF distance. (B) This dendogram is simplified by increasing the RF distance at which the 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 PeerJ reviewing PDF | (v2013:07:679:1:1:NEW 23 Sep 2013) R ev ie w in g M an us cr ip t trees are clustered (shown RF = 3). The top 10 clusters and their support different topologies are colored as indicated in the legend (with largest bin size for each cluster cluster in parentheses). 243 244 PeerJ reviewing PDF | (v2013:07:679:1:1:NEW 23 Sep 2013) R ev ie w in g M an us cr ip t phybin : Table 4. List of Wolbachia orthologous gene sets not conforming to the dominant topology when PhyBin is run using full clustering mode (--UPGMA, --editdist=3). Protein products predicted to be secreted (based on screening using the Effective database ( Jehl, Arnold et al. 2011 ) are italicized. PeerJ reviewing PDF | (v2013:07:679:1:1:NEW 23 Sep 2013) R ev ie w in g M an us cr ip t Table 4. List of Wolbachia orthologous gene sets not conforming to the dominant topology when PhyBin is run using full clustering mode (UPGMA, editdist=3). Protein products predicted to be secreted (based on screening using the Effective database (Jehl, Arnold et al. 2011)) are italicized. Topology group Orthologs (using wMel designations) Support for splitting group A Major facilitator family transporter (WD0470) Diaminopimelate epimerase (WD1208) GTP cyclohydrolase (WD0003) Metalopeptidase (WD0059) Periplasmic divalent cation tolerance (WD0828) RodA (WD1108) 1 2 3 PeerJ reviewing PDF | (v2013:07:679:1:1:NEW 23 Sep 2013) R ev ie w in g M an us cr ip t",
+    "v2_text": "results and discussion: : We used PhyBin to identify how many phylogenies within the Wolbachia orthologous gene set support the supergroup divisions proposed by multi-locus sequence typing (Baldo and Werren 2007). For comparative purposes in this analysis, a phylogeny for these 10 taxa was created using the concatenated, orthologous gene set (Figure 1A). In actuality, PhyBin does not require an expectation for tree topology and searches through tree space for distinct topological categories. As an illustration of PhyBin\u2019s ability to reduce the noise in a dataset produced by small branch lengths (i.e., closely related taxa), we used the program in binning mode on the set of Wolbachia orthologs under increasing branch length thresholds (Table 2). We chose a threshold of 0.01 for our dataset as the average branch length over the entire set of validated trees was 0.04 with minimum and maximum branch lengths 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 PeerJ reviewing PDF | (v2013:07:679:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t of 0 and 2.31, respectively. Using this threshold, in binning mode, the largest bin contains a topology that agrees with that of the published supergroups (Figure 1B). However, 174 other potential topologies exist in the dataset with 129 alternative topologies supported by only a single ortholog tree. In order to better explore this tree set, we took advantage of PhyBin\u2019s ability to generate a distance matrix for all trees. By calculating the Robinson-Foulds (RF) distance between all trees, we can better assess the differences between clusters in the tree dataset. For example, by increasing the RF-distance threshold to 2 and using the average-neighbor clustering algorithm to group our trees, the number of clusters drops dramatically to only 77 with the largest cluster containing a majority (72%) of genes. Again, this topology agrees with the published supergroup data and our result from the binning approach (Figure 1C). Increasing the RF-distance threshold further provides increasing stringency in the detection of aberrant phylogenies \u2013 topologies not falling into the largest cluster at larger distance thresholds are likely to represent genes of interest in comparing evolutionary trajectories of these supergroups. To test this hypothesis, we identified those Wolbachia genes that continue to display alternative evolutionary histories (that is, falling outside of the majority) even when clustering trees using increasingly large RF distances (Figure 2B, Table 3). As expected, a large number of distinct topologies are not inconsistent with the supergroup clades (Figure 2B). We further investigated the ortholog set supporting the dissolution of supergroup A (Table 4). Interestingly, a majority of these orthologs are predicted to be secreted (using the Effective database predictions of sec signal or eukaryotic domains (Jehl, Arnold et al. 2011), suggesting that perhaps interaction with the host would drive some of these orthologs in a different evolutionary direction compared to their resident genome. Another test of PhyBin\u2019s ability to detect orthologs under different evolutionary pressures would focus on the Wolbachia prophage, a mobile genetic element known to undergo horizontal transmission between strains (Bordenstein and Wernegreen 2004; Chafee, Funk et al. 2010; Kent and Bordenstein 2010)(Kent, Salichos et al. 2011). However, these phage orthologs do not occur across all of our 10 taxa included here and are therefore not suitable for testing support for the supergroups. In conclusion, we PhyBin is a new software program that efficiently and quickly groups phylogenies either by strict topological congruence or by clustering using RF distance. We believe that this tool, due to its ease of use, its speed, and informative output, will be of interest to evolutionary biologists and bioinformaticians alike. figure legends: : Figure 1. In each of two modes (full clustering and binning) PhyBin is able to correctly recover the expected topology for the Wolbachia pipientis orthologs used herein. (A) 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 PeerJ reviewing PDF | (v2013:07:679:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t Concatenated phylogeny based on 508 genes (using RAxML GTRGAMMA, bootstrap support based on 10,000 replicates). The four major supergroups are highlighted and denoted. (B) These same groups are recovered when PhyBin is run in either binning mode or (C) full clustering mode. Figure 2. Robinson-Foulds distance matricies produced by PhyBin are also visualized as a dendrogram by the software. (A) A tree of trees for the Wolbachia ortholog set (508 trees), clustered using an edit distance of 0, where identical topologies (nodes \u2013 grey ovals) are shown connected by a red line. Length of the branches connecting each node is proportional to the RF distance. (B) This dendogram is simplified by increasing the RF distance at which the trees are clustered (shown RF = 3). The top 10 clusters are colored and support different topologies, indicated in the legend. 256 257 258 259 260 261 262 263 264 265 266 PeerJ reviewing PDF | (v2013:07:679:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t introduction: : The advent of genomic sequencing has produced a large amount of data available for phylogenetic analysis and many researchers have attempted to utilize the phylogenetic signal found across the bacterial genome to develop species trees (Daubin, Gouy et al. 2001; Sicheritz-Ponten and Andersson 2001; Daubin, Moran et al. 2003; Bapteste, Boucher et al. 2004; Zhaxybayeva, Gogarten et al. 2006; Ellegaard, Klasson et al. 2013). What has become clear from these analyses is that significant fractions of bacterial genomes do not follow the evolutionary history of their resident genome (Bapteste, Boucher et al. 2004). These rogue genes are potentially undergoing evolutionary processes distinct from those felt by the rest of the 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 PeerJ reviewing PDF | (v2013:07:679:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t resident genome or have arrived there via horizontal gene transfer events. In order, then, to understand the evolution of the genome, it would be useful to achieve an understanding of the evolution of each gene in the genome. Previous work by Sicheritz-Ponten and Andersson presented scripts combined the existing utilities BLAST, Clustalw, Paup 4.0* to provide a complete pipeline from genome to tree-binning analysis (Sicheritz-Ponten and Andersson 2001). These kinds of complete solutions are convenient but constrain the user to the specific utilities chosen by the authors for alignment and phylogeny generation. Here we present PhyBin, a computer program aimed at binning precomputed sets of trees in Newick format, a file format produced by the majority of tree building software. PhyBin is a utility rather than a complete solution; it can serve as a component in many genomics pipelines, and provides a useful addition to the landscape of tools for dissecting and visualizing large numbers of trees. After the user applies their chosen ortholog prediction and tree-building algorithms, PhyBin offers a quick way to visualize and browse the different evolutionary histories, either binned by topology and sorted by bin size, or in the form of a full hierarchical clustering based on Robinson-Foulds distance: i.e. a tree of trees. method and implementation: : Generating orthologous sets and input trees Genomic sequences were downloaded from NCBI Microbial Genome Projects. The Wobachia species complex is made up of several major clades, called supergroups, designated by alphabetical letters (Baldo and Werren 2007). Accession numbers for the genomes analyzed here include: wUni and wVitA (submissions pending to genbank\u2019s ncbi), wBm (NC_006833.1), wPip-Pel (NC_010981.1), wHa (NC_021089.1), wRi (NC_012416.1), wMel (NC_002978.6), wNo (NC_021084.1), wAlbB (CAGB00000000.1), wBm (NC_006833.1), wOo (NC_018267.1). Orthologous gene sets were determined by Reciprocal Smallest Distance (RSD) algorithm (Wall, Fraser et al. 2003) with a 103 cutoff for significance threshold and alignment length threshold of 80%. Orthologs were then aligned using ClustalW (Larkin, Blackshields et al. 2007) and ML trees were generated using RAxML (Stamatakis 2006). The Newick format trees that resulted were used as input to PhyBin. The number of orthologous genes identified in this manner across all 10 taxa was 503. description of the program: : PhyBin is a standalone command-line program, portable across all major operating systems. It runs in batch-mode and is easily usable from scripts. PhyBin has two major modes: it can run very quickly and classify identical tree topologies into bins, or it can compute the distance (Robinson and Foulds 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 PeerJ reviewing PDF | (v2013:07:679:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t 1981) between all pairs of trees and use that distance matrix to produce a configurable clustering of trees. fast binning mode : The key algorithm PhyBin performs in this mode is tree normalization, computing a rooted, ordered normal form for all inputs (which are labeled, unrooted, unordered tree topologies). Previous work in this area has described a number of viable normal forms (Chi, Yang et al. 2005). PhyBin chooses the following strategy: it attempts to order subtrees by weight (number of tree nodes) and select the root node which is most balanced by weight (not depth)---that is, which minimizes the maximum weight of any child of the root. Node labels are used only to ``break ties'' between equally weighted subtrees, or equally balanced roots. Because input trees in Newick format are typically labeled only on the leaves (taxa), PhyBin generates labels for intermediate nodes in the tree by creating a set of all the leaves contained in that subtree, given a root to determine up/down direction. This set can be represented as a bit-vector and is also a key ingredient of computing Robinson-Foulds distance, which relies on identifying all such subsets (i.e. bipartitions induced by the tree). With labels for all nodes, equally weighted subtrees are ordered by label, and ties between potential roots are broken by comparing the labels of their children. Once input trees are normalized, comparing their topologies is as simple as comparing their representation in memory (a single, linear traversal). Normalization itself appears expensive due to the cost of labeling interior nodes with all leaves under them (O(N * I) for N taxa and I interior nodes), compounded by the fact that each intermediate node may have to consider each of its neighbors as a possible root and relabel itself b times in a tree of maximum branching factor b, yielding an O(N*I*b) asymptotic cost. However, in binning mode PhyBin runs much faster in the average case. One feature that enables PhyBin's efficiency is that it computes tree metadata---interior labels and ``balanced'' ratings---lazily, that is, on demand. Only when ``tie breaking'' is necessary between equally-weighted subtrees is an interior label computed at all. Likewise, only nodes ``near the center'' of the unrooted tree need to be considered for root status, those near the leaves need never be scored for balance. Next PhyBin does the binning, which amounts to inserting all normalized trees into a data structure indexed by tree topology. We define a total order over normalized trees (made possible by labels), and thereby represent the table of bins as a size-balanced binary tree supporting O(log(n)) insertion times. A hash-table would be an alternative, but the tree representation allows us to insert trees into the table without evaluating (forcing) unnecessary interior labels in the normal forms, whereas hashing requires traversing the entirety of each normalized tree to compute its hash. When execution completes, the contents of each bin are written out to disk, in 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 PeerJ reviewing PDF | (v2013:07:679:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t addition to a visualization of a representative average tree for that topology, computed by averaging branch lengths of the bin members. pre-processing data : PhyBin helps users extract a clean dataset and detect problems with the data, such as trees with mismatching numbers of taxa. In order to facilitate comparisons across trees with different taxon names (i.e. gene names), PhyBin can extract portions of designations or use a separate table of rules for mapping genes to taxa. A problem with the simple binning approach is that it is fragile to minor differences in trees caused by noise (e.g. short length branches with high variability). This becomes increasingly problematic with large numbers of taxa, especially when closely related taxa (different strains) are compared. Fortunately, a simple pre-processing step that addresses this problem: PhyBin provides an option to collapse branches under a length threshold (for example, a length threshold of 0.01 would collapse all branches less than 0.01, in their place inserting a star topology). full clustering mode using robinson-foulds distance matrix: : PhyBin reimplements the HashRF algorithm for full all-to-all Robinson Foulds distance (Sul and Williams 2007), which is significantly faster than computing the distance matrix with repeated comparison of individual trees (e.g. PAUP (Swofford and Sullivan 2009)). The HashRF algorithm is fast for today\u2019s data sizes (e.g. hundreds of taxa and thousands of trees), but is scales much more poorly than the basic binning algorithm at significantly larger sizes. A distance matrix alone is not directly useful for exploring the direct relationships between different gene trees. Thus, PhyBin uses the Robinson-Foulds distance matrix to compute a clustering of tree topologies, similar to the output of the simple binning mode, but able to identify trees that are merely similar, although not identical. The user may define the edit distance tolerated within clusters by providing a threshold, and may choose single, complete, or UPGMA linkage for clustering. Also if desired, rather than viewing a flat clustering of trees, the user may directly view a hierarchical clustering of the trees as a dendrogram. We believe PhyBin is the first program to date to provide this tree-of-trees output. output formats: : PhyBin is meant to be used in scripts and by other programs. Every output produced by PhyBin goes into a separate, simple text file---for example, the consensus tree for each cluster and the Robinson-Foulds distance matrix. Visualizations are produced separately and automatically in PDF files. performance: : There are very large differences in performance between existing programs for computing Robinson-Fould\u2019s distance matrices. The fundamental 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 PeerJ reviewing PDF | (v2013:07:679:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t data-structures in this problem domain are sets and finite maps, for which there are many alternate representations (bit vectors, hash tables, balanced trees, etc), providing a large space of possible implementations to explore. The sharpest contrast is between those programs that directly compare individual pairs of trees, vs. those that insert all tree\u2019s bipartitions into a global structure and summarize it as a separate phase (e.g. HashRF). The later approach achieves much better cache locality. PhyBin is written in a very high level language, Haskell, which supports radical forms of optimization, including safe semi-automatic parallelism. PhyBin uses purely functional (immutable) data-structures for representing trees and their bipartitons; in particular it relies heavily on the balanced-tree implementations Data.Map and Data.Set from the standard library. Nevertheless, when computing a matrix for a 150-taxa, 100-tree test (Table 1), PhyBin is 82 times faster than Philip (ANSI C) and 47.5 times faster than DendroPy (Python). However, PhyBin is still slower than HashRF by a factor of 2.8X-4.8X. HashRF is the implementation that introduced high-performance techniques for RF matrices, and developed the algorithm on which PhyBin\u2019s is based. Unfortunately, the widely used software (PAUP, DendroPy, Philip, etc), remains slow HashRF, the currently available fast alternative, is delicate and must be used carefully (for example, an extra character of whitespace in the input file results in a segmentation fault with no error message in version 6.0.1). Additionally, because HashRF provides only the core RF-distance computation, other tools are required for a biologist to be able to derive any conclusions from the output. As a final note on performance, PhyBin was straightforward to parallelize (using our \u201cLVar\u201d parallelism library) and achieves a 2.54X parallel speedup at four cores, and peaks at a 3.11X speedup at eight cores, making it a bit faster than HashRF on our target platform (Table 1). Future work will focus on reducing contention on shared data structures to improve scaling. trees phybin hashr : Branch length threshold Number of bins Number of singletons Size of largest bin RF-distance threshold Branch Length cutoff Number of clusters Number of singletons Size of largest cluster 0 n/a 222 149 16 1 n/a 140 67 34 2 n/a 77 29 56 0 0.01 175 129 133 0 0.02 95 68 201 1 0.02 66 35 246 2011)) are italicized. Topology group Orthologs (using wMel designations) Support for splitting group A Major facilitator family transporter (WD0470) Diaminopimelate epimerase (WD1208) GTP cyclohydrolase (WD0003) Metalopeptidase (WD0059) Periplasmic divalent cation tolerance (WD0828) RodA (WD1108) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 PeerJ reviewing PDF | (v2013:07:679:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t Figure 1 Wolbachia supergroup trees produced by concatenation of a dataset of 508 orthologs or by PhyBin's binning and clustering algorithm. PeerJ reviewing PDF | (v2013:07:679:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t Figure 2 Comparison of PhyBin's tree of trees output using RF distance of 0 and 3 PeerJ reviewing PDF | (v2013:07:679:0:1:NEW 30 Jul 2013) R ev ie w in g M an us cr ip t",
+    "url": "https://peerj.com/articles/188/reviews/",
+    "review_1": "Gavin Stewart \u00b7 Oct 8, 2013 \u00b7 Academic Editor\nACCEPT\nYour response to the reviewers questions have addressed substantive concerns and improved the manuscript. Thank you for engaging positively with the peer review process.",
+    "review_2": "Gavin Stewart \u00b7 Sep 5, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nPlease respond to the reviewers comments by amending the manuscript. Respond positively where you are able and include discussion of the issue where you disagree with views expressed by the reviewers. I draw particular attention to the issue of scale identified by reviewer two and exchangeability (fungability?) by reviewer three. Please ensure that there is appropriate discussion of the potential impacts of both on your conclusions.",
+    "review_3": "Nicola Randall \u00b7 Aug 20, 2013\nBasic reporting\nNo comment\nExperimental design\nPlease could the authors clarify the following?\n\nHow is the optimal yield for each crop/land area calculated & what assumptions are made/what factors are taken into account?\n\nWhat countries/regions are included in Figure 1? Is it all regions (including western Europe, Canada, US etc. as Figures 3&4 appear to show, or only the 'food insecure' regions specified in Figure 2?\n\nAlso, the data set used for 'current' land use appears to be 13 years old. Are the authors aware of how land-use/farming systems may have changed in any regions during that period & of any impacts said changes may have on their results/conclusions?\nValidity of the findings\nThis is an interesting concept, the implications of which are perhaps oversimplified in the conclusions.\n\nAlthough the authors recognise that there may be limitations to optimisation, I am concerned that as it stands at the moment some of the most important practical limitations (and potential impacts) are not acknowledged, and need to be discussed.\nThese include limitations caused by water scarcity (and any potential impacts of the proposed model on water security), any requirement for rotations to manage productivity, trade limitations/impacts, environmental impacts & any impact on nutritional deficiencies.\n\nI am unsure whether the statement in the final sentence is directly relevant to the findings and would be more comfortable with conclusions that encouraged elements of the optimisation approach to be adopted where feasible.\nCite this review as\nRandall NP (2013) Peer Review #1 of \"Transformative optimisation of agricultural land use to meet future food demands (v0.1)\". PeerJ https://doi.org/10.7287/peerj.188v0.1/reviews/1",
+    "review_4": "Stuart Pimm \u00b7 Jul 27, 2013\nBasic reporting\nI must stress that I am not an expert on agricultural ecosystems or crop yields and what constrains them. That said, I'm interested in global environmental issues and understand both estimates of production and diversity that form the core of this paper. I accepted the chance to review this manuscript because I have followed the senior author's work closely and admire his creative approaches to many problems.\n\nThis manuscript addresses a disarmingly simple question: can we grow more crops if we adjusted the mix to maximise productivity. The answers answer an emphatic \"yes\" for cereals while improvements for oilseeds are much smaller. All that said, the most interesting aspects are why countries have not optimised productivity. The authors suggest a variety of possibilities. At the risk of asking them to expand a paper that is short and to the point, it seems that a more complete examination of what limits optimal production is warranted.\n\n1. I consider the issue of spatial scale in the next section.\n\n2. Would an optimal production lead to greater profits for the farmers? And to what extent are allocations driven by national subsidies?\n\n3. Crop diversity is important. I found figures 3 and 4 to be most informative. Clearly most production is either close to optimal diversity or exceeds it considerably. The USA and China, for example, would need to move towards much less diverse croplands if they were to improve cereal production. Spain would need to become a cereal crop monoculture, for example.\n\n4. Large food producers are unlikely to wish to modify current allocations to feed other countries. I would like to see the improvements of production ranked by the net balance of food exports and imports. Could food importers avoid such dependency? And at what cost in terms of crop specialization?\nExperimental design\n1. My first worry is about scale. 10 x 10km is fine scale, certainly, but I'd like to be reassured that the following possibility is excluded. For such a pixel, quite possibly irrigated rice may attain the highest productivity within a small piece of that. Extrapolating such productivity across the pixel may be impossible. Just think of what happens along the Nile, for example, where one can stand with one foot in very productive crops and the other in desert. Yes, irrigated rice is more productive than rainfed wheat, but that doesn't mean one can grow rice everywhere within the pixel.\n\n2. The largest changes proposed would be to replace maize with wheat in Central Africa \u2014 for a huge increase in production \u2014 and to reduce wheat in China, but grow more rice. The authors mention these changes (page 5), but do not further investigate why the changes haven't been made. Water may well prevent rice from replacing wheat in China, and soil nutrients (and water) may well prevent wheat from replacing maize in Africa, especially one considers the scale issues I have already mentioned.\n\nThe way to investigate these possibilities is to examine a sample of pixels that seem particularly suboptimal \u2014where, for example, rice production is high per unit area within the pixel, but only a small fraction of the pixel grows rice. If that's an irrigation issue, then the authors need to assess how large an error this causes.\nValidity of the findings\nSee concerns expressed above.\nAdditional comments\nI view this as being most interesting as a way of documenting what the limitations are to increased production. The bottom line \u2014 substantial improvements \u2014 are subject to many caveats. The value of this manuscript is to list what some of them are.\nCite this review as\nPimm SL (2013) Peer Review #2 of \"Transformative optimisation of agricultural land use to meet future food demands (v0.1)\". PeerJ https://doi.org/10.7287/peerj.188v0.1/reviews/2",
+    "pdf_1": "https://peerj.com/articles/188v0.2/submission",
+    "pdf_2": "https://peerj.com/articles/188v0.1/submission",
+    "review_5": "William Laurance \u00b7 Jul 25, 2013\nBasic reporting\nThe article is clearly written, interesting, and well prepared. The figures support the reported conclusions.\n\nA few points of clarification are needed:\n\n1) Introduction, line 1: According to the UN Population Division, the human population exceeded 7 billion in October 2012 (on Halloween, notably, although this is obviously just an approximation).\n\n2) Introduction, on the assumption of the fungibility of crops: Obviously, this is quite a large assumption in the context of the present analysis. One could imagine lots of reasons for farmers electing to have multiple crops, ranging from balancing their dietary requirements to bet-hedging against crop-specific pathogens, weather, and crop-price fluctuations. Some brief discussion of this later in the paper would be warranted.\n\n3) Results section: One point on which I was not clear was crop transport. Some crops might be produced near to where their demand is concentrated, even if that locale is suboptimal. Is this factored into the analysis? I presume not. Again, this might be mentioned briefly in the Discussion.\nExperimental design\nThe design of the analysis is effective and well considered, and falls within the scope of the journal. The paper contains a great deal of interesting analysis and interpretation.\nValidity of the findings\nMy sense is that the analyses are reasonably robust and effectively interpreted, using the best available information and data sets at hand. The conclusions seem broadly justified by the analyses, notwithstanding the need for some minor points of clarification as indicated above.\nAdditional comments\nI found much of interest in this paper. It is appropriately framed as a sort of thought experiment, and addresses some very big and important questions.\nCite this review as\nLaurance W (2013) Peer Review #3 of \"Transformative optimisation of agricultural land use to meet future food demands (v0.1)\". PeerJ https://doi.org/10.7287/peerj.188v0.1/reviews/3",
+    "all_reviews": "Review 1: Gavin Stewart \u00b7 Oct 8, 2013 \u00b7 Academic Editor\nACCEPT\nYour response to the reviewers questions have addressed substantive concerns and improved the manuscript. Thank you for engaging positively with the peer review process.\nReview 2: Gavin Stewart \u00b7 Sep 5, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nPlease respond to the reviewers comments by amending the manuscript. Respond positively where you are able and include discussion of the issue where you disagree with views expressed by the reviewers. I draw particular attention to the issue of scale identified by reviewer two and exchangeability (fungability?) by reviewer three. Please ensure that there is appropriate discussion of the potential impacts of both on your conclusions.\nReview 3: Nicola Randall \u00b7 Aug 20, 2013\nBasic reporting\nNo comment\nExperimental design\nPlease could the authors clarify the following?\n\nHow is the optimal yield for each crop/land area calculated & what assumptions are made/what factors are taken into account?\n\nWhat countries/regions are included in Figure 1? Is it all regions (including western Europe, Canada, US etc. as Figures 3&4 appear to show, or only the 'food insecure' regions specified in Figure 2?\n\nAlso, the data set used for 'current' land use appears to be 13 years old. Are the authors aware of how land-use/farming systems may have changed in any regions during that period & of any impacts said changes may have on their results/conclusions?\nValidity of the findings\nThis is an interesting concept, the implications of which are perhaps oversimplified in the conclusions.\n\nAlthough the authors recognise that there may be limitations to optimisation, I am concerned that as it stands at the moment some of the most important practical limitations (and potential impacts) are not acknowledged, and need to be discussed.\nThese include limitations caused by water scarcity (and any potential impacts of the proposed model on water security), any requirement for rotations to manage productivity, trade limitations/impacts, environmental impacts & any impact on nutritional deficiencies.\n\nI am unsure whether the statement in the final sentence is directly relevant to the findings and would be more comfortable with conclusions that encouraged elements of the optimisation approach to be adopted where feasible.\nCite this review as\nRandall NP (2013) Peer Review #1 of \"Transformative optimisation of agricultural land use to meet future food demands (v0.1)\". PeerJ https://doi.org/10.7287/peerj.188v0.1/reviews/1\nReview 4: Stuart Pimm \u00b7 Jul 27, 2013\nBasic reporting\nI must stress that I am not an expert on agricultural ecosystems or crop yields and what constrains them. That said, I'm interested in global environmental issues and understand both estimates of production and diversity that form the core of this paper. I accepted the chance to review this manuscript because I have followed the senior author's work closely and admire his creative approaches to many problems.\n\nThis manuscript addresses a disarmingly simple question: can we grow more crops if we adjusted the mix to maximise productivity. The answers answer an emphatic \"yes\" for cereals while improvements for oilseeds are much smaller. All that said, the most interesting aspects are why countries have not optimised productivity. The authors suggest a variety of possibilities. At the risk of asking them to expand a paper that is short and to the point, it seems that a more complete examination of what limits optimal production is warranted.\n\n1. I consider the issue of spatial scale in the next section.\n\n2. Would an optimal production lead to greater profits for the farmers? And to what extent are allocations driven by national subsidies?\n\n3. Crop diversity is important. I found figures 3 and 4 to be most informative. Clearly most production is either close to optimal diversity or exceeds it considerably. The USA and China, for example, would need to move towards much less diverse croplands if they were to improve cereal production. Spain would need to become a cereal crop monoculture, for example.\n\n4. Large food producers are unlikely to wish to modify current allocations to feed other countries. I would like to see the improvements of production ranked by the net balance of food exports and imports. Could food importers avoid such dependency? And at what cost in terms of crop specialization?\nExperimental design\n1. My first worry is about scale. 10 x 10km is fine scale, certainly, but I'd like to be reassured that the following possibility is excluded. For such a pixel, quite possibly irrigated rice may attain the highest productivity within a small piece of that. Extrapolating such productivity across the pixel may be impossible. Just think of what happens along the Nile, for example, where one can stand with one foot in very productive crops and the other in desert. Yes, irrigated rice is more productive than rainfed wheat, but that doesn't mean one can grow rice everywhere within the pixel.\n\n2. The largest changes proposed would be to replace maize with wheat in Central Africa \u2014 for a huge increase in production \u2014 and to reduce wheat in China, but grow more rice. The authors mention these changes (page 5), but do not further investigate why the changes haven't been made. Water may well prevent rice from replacing wheat in China, and soil nutrients (and water) may well prevent wheat from replacing maize in Africa, especially one considers the scale issues I have already mentioned.\n\nThe way to investigate these possibilities is to examine a sample of pixels that seem particularly suboptimal \u2014where, for example, rice production is high per unit area within the pixel, but only a small fraction of the pixel grows rice. If that's an irrigation issue, then the authors need to assess how large an error this causes.\nValidity of the findings\nSee concerns expressed above.\nAdditional comments\nI view this as being most interesting as a way of documenting what the limitations are to increased production. The bottom line \u2014 substantial improvements \u2014 are subject to many caveats. The value of this manuscript is to list what some of them are.\nCite this review as\nPimm SL (2013) Peer Review #2 of \"Transformative optimisation of agricultural land use to meet future food demands (v0.1)\". PeerJ https://doi.org/10.7287/peerj.188v0.1/reviews/2\nReview 5: William Laurance \u00b7 Jul 25, 2013\nBasic reporting\nThe article is clearly written, interesting, and well prepared. The figures support the reported conclusions.\n\nA few points of clarification are needed:\n\n1) Introduction, line 1: According to the UN Population Division, the human population exceeded 7 billion in October 2012 (on Halloween, notably, although this is obviously just an approximation).\n\n2) Introduction, on the assumption of the fungibility of crops: Obviously, this is quite a large assumption in the context of the present analysis. One could imagine lots of reasons for farmers electing to have multiple crops, ranging from balancing their dietary requirements to bet-hedging against crop-specific pathogens, weather, and crop-price fluctuations. Some brief discussion of this later in the paper would be warranted.\n\n3) Results section: One point on which I was not clear was crop transport. Some crops might be produced near to where their demand is concentrated, even if that locale is suboptimal. Is this factored into the analysis? I presume not. Again, this might be mentioned briefly in the Discussion.\nExperimental design\nThe design of the analysis is effective and well considered, and falls within the scope of the journal. The paper contains a great deal of interesting analysis and interpretation.\nValidity of the findings\nMy sense is that the analyses are reasonably robust and effectively interpreted, using the best available information and data sets at hand. The conclusions seem broadly justified by the analyses, notwithstanding the need for some minor points of clarification as indicated above.\nAdditional comments\nI found much of interest in this paper. It is appropriately framed as a sort of thought experiment, and addresses some very big and important questions.\nCite this review as\nLaurance W (2013) Peer Review #3 of \"Transformative optimisation of agricultural land use to meet future food demands (v0.1)\". PeerJ https://doi.org/10.7287/peerj.188v0.1/reviews/3\nReview 6: \n",
+    "LLM_extracted_review": "[\"Review 1: Your response to the reviewers questions have addressed substantive concerns and improved the manuscript. Thank you for engaging positively with the peer review process.\\n\\nReview 2: Please respond to the reviewers comments by amending the manuscript. Respond positively where you are able and include discussion of the issue where you disagree with views expressed by the reviewers. I draw particular attention to the issue of scale identified by reviewer two and exchangeability (fungibility?) by reviewer three. Please ensure that there is appropriate discussion of the potential impacts of both on your conclusions.\\n\\nReview 3: Please could the authors clarify how the optimal yield for each crop/land area is calculated and what assumptions are made/what factors are taken into account? What countries/regions are included in Figure 1? Is it all regions or only the 'food insecure' regions specified in Figure 2? The data set used for 'current' land use appears to be 13 years old. Are the authors aware of how land-use/farming systems may have changed in any regions during that period? Some of the most important practical limitations (and potential impacts) are not acknowledged, including limitations caused by water scarcity, any requirement for rotations to manage productivity, trade limitations/impacts, environmental impacts, and any impact on nutritional deficiencies. I would be more comfortable with conclusions that encouraged elements of the optimisation approach to be adopted where feasible.\\n\\nReview 4: I would like to see a more complete examination of what limits optimal production. Would an optimal production lead to greater profits for the farmers? And to what extent are allocations driven by national subsidies? The largest changes proposed would be to replace maize with wheat in Central Africa and to reduce wheat in China, but grow more rice. The authors need to assess how large an error this causes regarding irrigation issues.\\n\\nReview 5: A few points of clarification are needed regarding the assumption of the fungibility of crops and crop transport. Some crops might be produced near to where their demand is concentrated, even if that locale is suboptimal. Is this factored into the analysis?\"]"
+}

peerj_json_files/PeerJ_Json_139.json ADDED Viewed

	@@ -0,0 +1,16 @@

+{
+    "v1_Abstract": "Human population is expected to reach ~9 billion by 2050. The ensuing demands for water, food and energy would intensify land-use conflicts and exacerbate environmental impacts. Therefore we urgently need to reconcile our growing consumptive needs with environmental protection. Here, we explore the potential of a land-use optimisation strategy to increase global agricultural production on two major groups of crops: cereals and oilseeds. We implemented a spatially-explicit computer simulation model across 173 countries based on the following algorithm: on any cropland, always produce the most productive crop given all other crops currently being produced locally and the site-specific biophysical, economic and technological constraints to production. Globally, this strategy resulted in net increases in annual production of cereal and oilseed crops from 1.9 billion to 2.9 billion tons (46%), and from 427 million to 481 million tons (13%), respectively, without any change in total land area harvested for cereals or oilseeds. This thought experiment demonstrates that, in theory, more optimal use of existing farmlands could help meet future crop demands. In practice there might be cultural, social and institutional barriers that limit the full realisation of this theoretical potential. Nevertheless, these constraints have to be weighed against the consequences of not producing enough food, particularly in regions already facing food shortages.",
+    "v2_Abstract": "Human population is expected to reach 9.1 billion by 2050. The ensuing demands for water, food and energy would intensify land-use conflicts and exacerbate environmental impacts. Therefore we urgently need to reconcile our growing consumptive needs with environmental protection. Here, we explore the potential of a land-use optimisation strategy to increase global agricultural production on two major groups of crops: cereals and oilseeds. We implemented a spatially-explicit computer simulation model across 173 countries based on the following algorithm: on any cropland, always produce the most productive crop given all other crops currently being produced locally and the site-specific biophysical, economic and technological constraints to production. Globally, this strategy resulted in net increases in annual production of cereal and oilseed crops from 1.9 billion to 2.9 billion tons (46%), and from 427 million to 481 million tons (13%), respectively, without any change in total land area harvested for cereals or oilseeds. This thought experiment demonstrates that, in theory, more optimal use of existing farmlands could help meet future crop demands. In practice there might be cultural, social and institutional barriers that limit the full realisation of this theoretical potential. Nevertheless, these constraints have to be weighed against the consequences of not producing enough food, particularly in regions already facing food shortages.",
+    "v1_text": "results : Globally, our strategy resulted in net increases in annual production of cereal and oilseed crops from 1.9 billion to 2.9 billion tons (46%), and from 427 million to 481 million tons (13%), respectively, without any change in total land area harvested for cereals (651 million ha) or oilseeds (184 million ha) (Tables S1-S4). Accordingly, annual production of vegetable oil and protein meal (the primary products of oilseeds) increased from 86 million to 94 million tons (10%), and from 176 million to 228 million tons (29%), respectively. Global demand for cereals is projected to increase to 2.7 billion tons by 2030, and to 3 billion tons by 2050 (including its use as animal feed) (FAO 2006). As such, land-use optimisation could contribute substantially to meeting future demands for cereals (at least until 2030). In contrast, the modest benefits of optimisation for vegetable oil production would not be sufficient to meet expected demands in 2030 (216 million tons) or 2050 (293 million tons) (FAO 2006). 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 PeerJ reviewing PDF | (v2013:07:656:1:1:NEW 19 Sep 2013) R ev ie w in g M an us cr ip t Among cereal crops, maize and rice underwent the largest expansions in harvested area, accompanied by increases in annual crop production by 746 million tons and 560 million tons, respectively (Fig. 2a). All other cereal crops, with the exception of sorghum, declined in both area (by at least 50%) and production (by 44-53%; Figure 1a). Although the harvested area for sorghum declined by 16 million ha (40%), annual production increased by 22 million tons (38%). This is due to an increase in sorghum\u2019s average annual yield from 1.7 to 3.4 tons/ha, as a result of land-use optimisation (Fig. S1). In the case of oilseeds, soy was the only crop that expanded in area (by 60 million ha or 81%; Fig. 2b). Soy was also the only oilseed crop to experience a decrease in average annual yield (from 2.4 to 2.2 tons/ha), as most of its expansion occurred on lands that were sub-optimal for soy but still more productive under soy than under any other crop (Fig. S2). Even so, annual production of soy increased by 98 million tons (60%). Oil palm production also increased by 23 million tons (20%; Fig. 2b). We next assessed whether the benefits of land-use optimisation would be manifested where most required, by exploring its implications for cereal production in five regions of the world that face the most severe food shortages, and would likely continue to do so in the future. These regions, which include South Asia, China, Southeast Asia, East Africa and Central Africa, contain 75% (657 million) of the world\u2019s malnourished people (FAO 2012; Lobell et al. 2008). We found that in South Asia, China and Southeast Asia, rice would remain the dominant cereal crop (Fig. 3). In fact, rice-cultivated area would increase from 129 million to 176 million ha at the expense of wheat, millet and sorghum which, incidentally, are thought to be the most vulnerable to climate change impacts in South Asia (Lobell et al. 2008) (Fig. 3). In East Africa and Central Africa, maize would no longer be the dominant cereal crop. Instead, East Africa would grow mainly rice (3.1 million ha), maize (3 million ha) and barley (2.7 million ha), while Central Africa would specialise in wheat (2.2 million ha) and rice (1.1 million ha) (Fig. 3). Following land-use 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 PeerJ reviewing PDF | (v2013:07:656:1:1:NEW 19 Sep 2013) R ev ie w in g M an us cr ip t optimisation, Central Africa would almost double its annual production of cereals (Fig. 3). The other four regions would also experience increases in cereal production (11-68%) (Fig. 3). Discussion We recognise that besides productivity, other cultural and socio-political considerations also determine actual land use and production systems. For example, land-use optimisation entails reducing annual rice production in Thailand from 24.4 million to 5.9 million tons (Table S3). Rice farmers in Thailand might be hesitant to switch to planting other cereal crops, as rice has a long history of cultivation and consumption in the region, in the same way that maize is intimately associated with the cultures and history of the Americas. To explore the effects of such cultural constraints, we re-ran the model with a modified algorithm, which excluded rice-cultivated areas from the optimisation process. In this case optimised global annual production for cereals was 2.7 billion tons, slightly less than the 2.9 billion tons projected based on the optimisation of all six cereal crops. Thus the exclusion of rice from optimisation has little overall impact on the production of cereals. The specialisation of production systems implies homogenisation of farms and agricultural landscapes. Yet some farmers might prefer to maintain multiple crops for various reasons, including balancing dietary requirements, and bet-hedging against outbreaks of pests and diseases, adverse weather conditions and price fluctuations that a monoculture might be more sensitive to. While land use optimisation might indeed drive homogenisation within individual farmlands, it is not necessarily so at national and regional scales: there is considerable variation in crop diversity following optimisation, with diversity actually increasing in many countries and regions (Figs. 3-5). We do not necessarily advocate that nations should pursue, solely, a production maximization strategy, but rather our results indicate the potential for substantial increases in crop production with such an approach. In practice, other considerations, including the benefits of maintaining diverse cropping systems, will necessarily affect the agricultural decisions taken. Neither do we imply that land-use optimisation is the only solution. On the contrary, a move towards optimisation should be implemented alongside other solutions, such as closing yield gaps, which are especially high for maize in Sub-Saharan Africa (World Bank 2008). In fact, land-use optimisation needs to be complemented by improvements in farming technologies and institutional structures, such as education, and market and financial risk management systems, all of which farmers need to make best use of the land and technologies available to them. Furthermore, given that smallholder farming often is the most common form of agricultural organisation, especially (but not only) in the tropics, smallholders will need to be 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 PeerJ reviewing PDF | (v2013:07:656:1:1:NEW 19 Sep 2013) R ev ie w in g M an us cr ip t integrated in any land-use optimisation approach through the provision of education, technology, and market and finance opportunities. In conclusion, our assessment demonstrates that in theory future crop demands, at least for cereals, can be substantially met on existing agricultural land area through the pursuit of more optimal use of farmlands. In practice there might be cultural, social and institutional barriers that limit the full realisation of this theoretical potential. Nevertheless, these constraints have to be weighed against the consequences of not producing enough food, particularly in regions already facing food shortages. 150 151 152 153 154 155 156 157 PeerJ reviewing PDF | (v2013:07:656:1:1:NEW 19 Sep 2013) R ev ie w in g M an us cr ip t material and methods : We assessed geospatial information on current land-use and crop-yield for these crops at the farmland scale across 173 countries (Monfreda et al. 2008). We based our analyses on a published global geospatial dataset at 5 arc-minute resolution (~10\u00d710 km grid cell) that depicts, for the year 2000, the proportion of harvested area and actual yield reported for each crop in each grid cell (Monfreda et al. 2008). We overlaid these data to produce a new data layer of intersected polygons (i.e. land areas sharing unique geospatial information on observed yield for each crop; referred to as farmlands in text). For cereals, these data encompass a total area of 651 million ha (~42% of Earth\u2019s total arable and permanent croplands) (FAO 2012) and comprise 788,557 data polygons (polygon mean area=826\u00b14.1 ha [\u00b1 standard error]); for oilseeds, these data encompass a total area of 184 million ha and comprise 426,000 data polygons (mean area=433\u00b12.9 ha). We carried out a three-step procedure to estimate optimised crop production within each farmland (Fig. 1). First, we established a baseline of current total production of cereal or oilseed crops within each farmland. We did so by multiplying harvested area with observed yield of each cereal or oilseed crop (Monfreda et al. 2008). Unlike cereals, whereby cereal grain is the prime economically-important product, oilseeds are produced for both oil and meal. We calculated vegetable oil and protein meal production amounts by multiplying crop production with an oil- or meal-conversion factor (derived from 2008/09 data on global crop, oil and meal production) (USDA-FAS 2011). Second, we identified an optimal cereal or oilseed crop within each farmland. The optimal cereal crop was the one with the highest observed yield within each farmland. In identifying an optimal oilseed crop, we assessed relative productivity based on the combined quantity of oil and meal produced. Given that global demand for protein meal is higher than that for vegetable oil, in optimising for oilseed production, we ascribed meal a relative weightage of 1.77 tons for every ton of oil produced (derived from 2008/09 data on global crop, oil and meal consumption) (USDA-FAS 2011). 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 PeerJ reviewing PDF | (v2013:07:656:1:1:NEW 19 Sep 2013) R ev ie w in g M an us cr ip t Third, we simulated land use optimisation by converting each farmland to a monoculture of the identified optimal cereal or oilseed crop. We calculated the resultant cereal or oilseed production for each optimised farmland by multiplying total harvested area with observed yield of the optimal crop. We calculated the benefits of land use optimisation by comparing this new production volume with the baseline production prior to optimisation. In our analyses we made the following assumptions. First, we assumed that the reported yield of each crop is uniform within each farmland (mean area=826\u00b14.1 ha for cereals and 433\u00b12.9 ha for oilseeds). If parts of a farmland had substantially lower (or higher) than the reported yield for the optimal crop, we would have overestimated (or underestimated) the benefits of land use optimisation. Second, we assumed that site-specific biophysical, economic and technological constraints to production are also uniform at the scale of each farmland, such that optimisation of a farmland for any optimal crop would not be limited by, for example, variations in water scarcity or soil nutrient levels across different parts of a farmland. transformative optimisation of agricultural land use to meet future food demands : Human population is expected to reach ~9 billion by 2050. The ensuing demands for water, food and energy would intensify land-use conflicts and exacerbate environmental impacts. Therefore we urgently need to reconcile our growing consumptive needs with environmental protection. Here, we explore the potential of a land-use optimisation strategy to increase global agricultural production on two major groups of crops: cereals and oilseeds. We implemented a spatially-explicit computer simulation model across 173 countries based on the following algorithm: on any cropland, always produce the most productive crop given all other crops currently being produced locally and the site-specific biophysical, economic and technological constraints to production. Globally, this strategy resulted in net increases in annual production of cereal and oilseed crops from 1.9 billion to 2.9 billion tons (46%), and from 427 million to 481 million tons (13%), respectively, without any change in total land area harvested for cereals or oilseeds. This thought experiment demonstrates that, in theory, more optimal use of existing farmlands could help meet future crop demands. In practice there might be cultural, social and institutional barriers that limit the full realisation of this theoretical potential. Nevertheless, these constraints have to be weighed against the consequences of not producing enough food, particularly in regions already facing food shortages. PeerJ reviewing PDF | (v2013:07:656:1:1:NEW 19 Sep 2013) R ev ie w in g M an us cr ip t Lian Pin Koh1,2*, Thomas Koellner3, and Jaboury Ghazoul1 1Department of Environmental Systems Science, ETH Zurich, CHN G 73.1, Universit\u00e4tstrasse 16, Zurich 8092, Switzerland 2Current address: Woodrow Wilson School of Public and International Affairs, Princeton University, Robertson Hall, Princeton, New Jersey 08544-1013, USA 3Faculty of Biology, Chemistry and Geosciences, University of Bayreuth, Universitaetstrasse 30, 95440 Bayreuth, Germany *Corresponding author: Lian Pin Koh, Woodrow Wilson School of Public and International Affairs, Princeton University, Robertson Hall, Princeton, New Jersey 08544-1013, USA , phone: +16097590952, email: lianpinkoh@gmail.com 1 2 3 4 5 6 7 8 9 10 PeerJ reviewing PDF | (v2013:07:656:1:1:NEW 19 Sep 2013) R ev ie w in g M an us cr ip t Introduction By 2050, global human population will have grown from the current ~7 billion to ~9 billion people (United Nations 2008). These people will require more food (Evans 2009; Godfray et al. 2010). They are also likely to demand a higher proportion of meat and dairy products that require more land, water and energy to produce (Royal Society of London 2009; Tilman et al. 2001). Meeting this demand is daunting by virtue of the need to reduce greenhouse-gas emissions (Meinshausen et al. 2009), minimise fertiliser and pesticide inputs (Moss 2007), and avoid further impacts on natural ecosystems and wildlife (Ehrlich & Pringle 2008). Additionally, we might have to cope with the yet unclear implications of climate change on food security (Brown & Funk 2008; Lobell et al. 2008; Parry et al. 2004). These challenges might be met by closing yield gaps (i.e. difference between potential and actual yields) or raising yield ceilings, reducing food lost to waste, and switching to less protein-rich or more aquaculture-based diets (Foley et al. 2011; Godfray et al. 2010). Additionally, we propose that a complementary approach is to maximise agricultural returns by planting crops that are best suited to site-specific conditions. While this strategy might seem obvious, the degree to which agricultural land use is optimised and the benefits of optimisation have not been evaluated at a global scale by which benefits might be maximally realised. To test the efficacy of this land-use optimisation approach, we developed a spatiallyexplicit computer simulation model based on the following algorithm: on any cropland, always produce the most productive crop given all other crops currently being produced locally and the site-specific biophysical, economic and technological constraints to production. By evaluating crops based on their realised yields, the algorithm captures both the local biophysical limitations to production (e.g. the need for irrigation), and the behaviour of farmers in response to these constraints (e.g. the decision to irrigate or not). Therefore, for a farmer who is currently growing barley, maize, wheat and irrigated rice on his land, and if irrigated rice has the highest per-hectare realised yield given local conditions, then land-use optimisation would entail devoting the entire farmland to irrigated rice production. An implicit requirement of this approach is that goods being considered are fungible, such that individual units of different crops within a commodity group (e.g. cereals or vegetable oil) are mutually substitutable. Therefore, we illustrate our approach by optimising land use within each of two groups of essential and fungible food crops: cereals (barley, maize, millet, rice, sorghum and wheat) and oilseeds (soy, cottonseed, rapeseed, sunflower seed, groundnut and oil palm). We optimised land use by replacing all currently harvested area, for cereals or oilseeds, with the most productive crop in the set of currently harvested crops within each farmland (Fig. 1) (Monfreda et al. 2008). 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 PeerJ reviewing PDF | (v2013:07:656:1:1:NEW 19 Sep 2013) R ev ie w in g M an us cr ip t",
+    "v2_text": "results : Globally, our strategy resulted in net increases in annual production of cereal and oilseed crops from 1.9 billion to 2.9 billion tons (46%), and from 427 million to 481 million tons (13%), respectively, without any change in total land area harvested for cereals (651 million ha) or oilseeds (184 million ha) (Tables S1-S4). Accordingly, annual production of vegetable oil and protein meal (the primary products of oilseeds) increased from 86 million to 94 million tons (10%), and from 176 million to 228 million tons (29%), respectively. Global demand for cereals is projected to increase to 2.7 billion tons by 2030, and to 3 billion tons by 2050 (including its use as animal feed) (FAO, 2006). As such, land-use optimisation could contribute substantially to meeting future demands for cereals (at least until 2030). In contrast, the modest benefits of optimisation for vegetable oil production would not be sufficient to meet expected demands in 2030 (216 million tons) or 2050 (293 million tons) (FAO, 2006). Among cereal crops, maize and rice underwent the largest expansions in harvested area, accompanied by increases in annual crop production by 746 million tons and 560 million tons, respectively (Fig. 1a). All other cereal crops, with the exception of sorghum, declined in both area (by at least 50%) and production (by 44-53%; Figure 1a). Although the harvested area for sorghum declined by 16 million ha (40%), annual production increased by 22 million tons (38%). This is due to an increase in sorghum\u2019s average annual yield from 1.7 to 3.4 tons/ha, as a result of land-use optimisation (Fig. S1). PeerJ reviewing PDF | (v2013:07:656:0:0:NEW 17 Jul 2013) R ev ie w in g M an us cr ip t 5 5 In the case of oilseeds, soy was the only crop that expanded in area (by 60 million ha or 81%; Fig. 1b). Soy was also the only oilseed crop to experience a decrease in average annual yield (from 2.4 to 2.2 tons/ha), as most of its expansion occurred on lands that were sub-optimal for soy but still more productive under soy than under any other crop (Fig. S2). Even so, annual production of soy increased by 98 million tons (60%). Oil palm production also increased by 23 million tons (20%; Fig. 1b). We next assessed whether the benefits of land-use optimisation would be manifested where most required, by exploring its implications for cereal production in five regions of the world that face the most severe food shortages, and would likely continue to do so in the future. These regions, which include South Asia, China, Southeast Asia, East Africa and Central Africa, contain 75% (657 million) of the world\u2019s malnourished people (FAO, 2012, Lobell et al. , 2008 ). We found that in South Asia, China and Southeast Asia, rice would remain the dominant cereal crop (Fig. 2). In fact, rice-cultivated area would increase from 129 million to 176 million ha at the expense of wheat, millet and sorghum which, incidentally, are thought to be the most vulnerable to climate change impacts in South Asia (Lobell et al. , 2008 ) (Fig. 2). In East Africa and Central Africa, maize would no longer be the dominant cereal crop. Instead, East Africa would grow mainly rice (3.1 million ha), maize (3 million ha) and barley (2.7 million ha), while Central Africa would specialise in wheat (2.2 million ha) and rice (1.1 million ha) (Fig. 2). Following land-use optimisation, Central Africa would almost double its annual production of cereals (Fig. 2). The other four regions would also experience increases in cereal production (11-68%) (Fig. 2). Discussion We recognise that besides productivity, other cultural and socio-political considerations also determine actual land use and production systems. For example, land-use optimisation entails reducing annual rice production in Thailand from 24.4 million to 5.9 million tons (Table S3). Rice farmers in Thailand might be hesitant to switch to planting other cereal crops, as rice has a long history of cultivation and consumption in the region, in the same way that maize is intimately associated with the cultures and history of the Americas. To explore the effects of such cultural constraints, we re-ran the model with a modified algorithm, which excluded rice-cultivated areas from the optimisation process. In this case PeerJ reviewing PDF | (v2013:07:656:0:0:NEW 17 Jul 2013) R ev ie w in g M an us cr ip t 6 6 optimised global annual production for cereals was 2.7 billion tons, slightly less than the 2.9 billion tons projected based on the optimisation of all six cereal crops. Thus the exclusion of rice from optimisation has little overall impact on the production of cereals. Some might also argue against the specialisation of production systems because it implies homogenisation of agricultural landscapes. While this is true within individual farmlands, it is not necessarily so at national and regional scales: there is considerable variation in crop diversity following optimisation, with diversity actually increasing in many countries and regions (Figs. 2-4). Neither do we imply that land-use optimisation is the only solution. On the contrary, a move towards optimisation should be implemented alongside other solutions, such as closing yield gaps, which are especially high for maize in Sub-Saharan Africa (World Bank, 2008). In fact, land-use optimisation needs to be complemented by improvements in farming technologies and institutional structures, such as education, and market and financial risk management systems, all of which farmers need to make best use of the land and technologies available to them. Furthermore, given that smallholder farming often is the most common form of agricultural organisation, especially (but not only) in the tropics, smallholders will need to be integrated in any land-use optimisation approach through the provision of education, technology, and market and finance opportunities. In conclusion, our assessment demonstrates that in theory future crop demands, at least for cereals, can be substantially met on existing agricultural land area through the pursuit of more optimal use of farmlands. In practice there might be cultural, social and institutional barriers that limit the full realisation of this theoretical potential. Nevertheless, these constraints have to be weighed against the consequences of not producing enough food, particularly in regions already facing food shortages. Our results suggest that a combined scientific and socioeconomic approach that targets smallholder farmers, especially those in developing countries, might contribute to ensuring future food security and alleviating rural poverty. PeerJ reviewing PDF | (v2013:07:656:0:0:NEW 17 Jul 2013) R ev ie w in g M an us cr ip t 7 7 figure legends : Figure 2. Benefits of land-use optimisation in food-insecure regions. Left: changes in relative proportions of harvested area of cereal crops; right: changes in total crop production (numbers in parentheses indicate percentage change). Figure 3. Cereal crop diversity. Crop diversity was calculated based on the reciprocal Simpson\u2019s diversity index, \u2211\u2212 21 ip , whereby p represents the proportional area of the ith crop type. This index reflects the probability that two randomly chosen cropland areas are not cultivated for the same cereal crop. Circle areas reflect relative total harvested area for cereal crops. Dashed line indicates no change in crop diversity between current and optimal land uses. For clarity of presentation, not all country labels are shown. Figure 4. Oilseed crop diversity. Crop diversity was calculated based on the reciprocal Simpson\u2019s diversity index, \u2211\u2212 21 ip , whereby p represents the proportional area of the ith crop type. This index reflects the probability that two randomly chosen cropland areas are not cultivated for the same oilseed crop. Circle areas reflect relative total harvested area for oilseed crops. Dashed line indicates no change in crop diversity between current and optimal land uses. For clarity of presentation, not all country labels are shown. PeerJ reviewing PDF | (v2013:07:656:0:0:NEW 17 Jul 2013) R ev ie w in g M an us cr ip t Figure 1 Harvested area and crop production. Changes in cultivated area and production amounts of cereal (a) and oilseed crops (b) under current and optimal land-use allocations. Filled bars represent current land use; open bars represent optimal land use; and numbers in parentheses indicate percentage change. PeerJ reviewing PDF | (v2013:07:656:0:0:NEW 17 Jul 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:07:656:0:0:NEW 17 Jul 2013) R ev ie w in g M an us cr ip t Figure 2 Benefits of land-use optimisation in food-insecure regions. Left: changes in relative proportions of harvested area of cereal crops; right: changes in total crop production (numbers in parentheses indicate percentage change). PeerJ reviewing PDF | (v2013:07:656:0:0:NEW 17 Jul 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:07:656:0:0:NEW 17 Jul 2013) R ev ie w in g M an us cr ip t Figure 3 material and methods : We assessed geospatial information on current land-use and crop-yield for these crops at the farmland scale across 173 countries (Monfreda et al. , 2008 ). We based our analyses on a published global geospatial dataset at 5 arc-minute resolution (~10\u00d710 km grid cell) that depicts, for the year 2000, the proportion of harvested area and actual yield reported for each crop in each grid cell (Monfreda et al. , 2008 ). We overlaid these data to produce a new data layer of intersected polygons (i.e. land areas sharing the same geospatial information; referred to as farmlands in text). For cereals, these data encompass a total area of 651 million ha (~42% of Earth\u2019s total arable and permanent croplands) (FAO, 2012) and comprise 788,557 data polygons (polygon mean area=826\u00b14.1 ha [\u00b1 standard error]); for oilseeds, these data encompass a total area of 184 million ha and comprise 426,000 data polygons (mean area=433\u00b12.9 ha). To calculate current crop production within each polygon, we multiplied harvested area with observed yield (Monfreda et al. , 2008 ). In the case of oilseeds, vegetable oil and protein meal production amounts were calculated by multiplying crop production with an oil- or meal-conversion factor (derived from 2008/09 data on global crop, oil and meal production) (USDA-FAS, 2011). Unlike cereals, whereby cereal grain is the prime economically-important product, oilseeds are produced for both oil and meal. Therefore, in identifying an optimal oilseed crop, we assessed relative productivity based on the combined quantity of oil and meal produced. Given that global demand for protein meal is higher than that for vegetable oil, in optimising for oilseed production, we ascribed PeerJ reviewing PDF | (v2013:07:656:0:0:NEW 17 Jul 2013) R ev ie w in g M an us cr ip t 4 4 meal a relative weightage of 1.77 tons for every ton of oil produced (derived from 2008/09 data on global crop, oil and meal consumption) (USDA-FAS, 2011). transformative optimisation of agricultural land use to meet future food demands : Human population is expected to reach 9.1 billion by 2050. The ensuing demands for water, food and energy would intensify land-use conflicts and exacerbate environmental impacts. Therefore we urgently need to reconcile our growing consumptive needs with environmental protection. Here, we explore the potential of a land-use optimisation strategy to increase global agricultural production on two major groups of crops: cereals and oilseeds. We implemented a spatially-explicit computer simulation model across 173 countries based on the following algorithm: on any cropland, always produce the most productive crop given all other crops currently being produced locally and the site-specific biophysical, economic and technological constraints to production. Globally, this strategy resulted in net increases in annual production of cereal and oilseed crops from 1.9 billion to 2.9 billion tons (46%), and from 427 million to 481 million tons (13%), respectively, without any change in total land area harvested for cereals or oilseeds. This thought experiment demonstrates that, in theory, more optimal use of existing farmlands could help meet future crop demands. In practice there might be cultural, social and institutional barriers that limit the full realisation of this theoretical potential. Nevertheless, these constraints have to be weighed against the consequences of not producing enough food, particularly in regions already facing food shortages. PeerJ reviewing PDF | (v2013:07:656:0:0:NEW 17 Jul 2013) R ev ie w in g M an us cr ip t 1 1 Lian Pin Koh1,2*, Thomas Koellner3, and Jaboury Ghazoul1 1Department of Environmental Systems Science, ETH Zurich, CHN G 73.1, Universit\u00e4tstrasse 16, Zurich 8092, Switzerland 2Current address: Woodrow Wilson School of Publica and International Affairs, Princeton University, Robertson Hall, Princeton, New Jersey 08544-1013, USA 3Faculty of Biology, Chemistry and Geosciences, University of Bayreuth, Universitaetstrasse 30, 95440 Bayreuth, Germany *Corresponding author: Lian Pin Koh, Woodrow Wilson School of Publica and International Affairs, Princeton University, Robertson Hall, Princeton, New Jersey 08544-1013, USA , phone: +16097590952, email: lianpinkoh@gmail.com PeerJ reviewing PDF | (v2013:07:656:0:0:NEW 17 Jul 2013) R ev ie w in g M an us cr ip t 2 2 Introduction By 2050, global human population will have grown from the current 6.9 billion to 9.1 billion people (United Nations, 2008). These people will require more food (Evans, 2009, Godfray et al. , 2010 ). They are also likely to demand a higher proportion of meat and dairy products that require more land, water and energy to produce (Royal Society of London, 2009, Tilman et al. , 2001 ). Meeting this demand is daunting by virtue of the need to reduce greenhouse-gas emissions (Meinshausen et al. , 2009 ), minimise fertiliser and pesticide inputs (Moss, 2007), and avoid further impacts on natural ecosystems and wildlife (Ehrlich & Pringle, 2008). Additionally, we might have to cope with the yet unclear implications of climate change on food security (Brown & Funk, 2008, Lobell et al. , 2008 , Parry et al. , 2004 ). These challenges might be met by closing yield gaps (i.e. difference between potential and actual yields) or raising yield ceilings, reducing food lost to waste, and switching to less protein-rich or more aquaculture-based diets (Foley et al. , 2011 , Godfray et al. , 2010 ). Additionally, we propose that a complementary approach is to maximise agricultural returns by planting crops that are best suited to site-specific conditions. While this strategy might seem obvious, the degree to which agricultural land use is optimised and the benefits of optimisation have not been evaluated at a global scale by which benefits might be maximally realised. To test the efficacy of this land-use optimisation approach, we developed a spatially-explicit computer simulation model based on the following algorithm: on any cropland, always produce the most productive crop given all other crops currently being produced locally and the site-specific biophysical, economic and technological constraints to production. By evaluating crops based on their realised yields, the algorithm captures both the local biophysical limitations to production (e.g. the need for irrigation), and the behaviour of farmers in response to these constraints (e.g. the decision to irrigate or not). Therefore, for a farmer who is currently growing barley, maize, wheat and irrigated rice on his land, and if irrigated rice has the highest per-hectare realised yield given local conditions, then land-use optimisation would entail devoting the entire farmland to irrigated rice production. An implicit requirement of this approach is that goods being considered are fungible, such that individual units of different crops within a commodity group (e.g. cereals or vegetable oil) are mutually substitutable. Therefore, we illustrate our approach by optimising land use within each of two groups of essential and fungible food crops: cereals (barley, maize, millet, rice, sorghum and wheat) and oilseeds (soy, PeerJ reviewing PDF | (v2013:07:656:0:0:NEW 17 Jul 2013) R ev ie w in g M an us cr ip t 3 3 cottonseed, rapeseed, sunflower seed, groundnut and oil palm). We optimised land use by replacing all currently harvested area, for cereals or oilseeds, with the most productive crop in the set of currently harvested crops within each farmland (Monfreda et al. , 2008 ). cereal crop diversity. : Crop diversity was calculated based on the reciprocal Simpson\u2019s diversity index. This index reflects the probability that two randomly chosen cropland areas are not cultivated for the same cereal crop. Circle areas reflect relative total harvested area for cereal crops. Dashed line indicates no change in crop diversity between current and optimal land uses. For clarity of presentation, not all country labels are shown. PeerJ reviewing PDF | (v2013:07:656:0:0:NEW 17 Jul 2013) R ev ie w in g M an us cr ip t Figure 4 oilseed crop diversity. : Crop diversity was calculated based on the reciprocal Simpson\u2019s diversity index. This index reflects the probability that two randomly chosen cropland areas are not cultivated for the same oilseed crop. Circle areas reflect relative total harvested area for oilseed crops. Dashed line indicates no change in crop diversity between current and optimal land uses. For clarity of presentation, not all country labels are shown. PeerJ reviewing PDF | (v2013:07:656:0:0:NEW 17 Jul 2013) R ev ie w in g M an us cr ip t",
+    "url": "https://peerj.com/articles/189/reviews/",
+    "review_1": "Gavin Stewart \u00b7 Oct 8, 2013 \u00b7 Academic Editor\nACCEPT\nThank for you for clarifying the points raised by myself and the reviewers, and for constructive engagement with peer review. I think this paper is a great addition to the literature.",
+    "review_2": "Gavin Stewart \u00b7 Sep 5, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nThe work is a useful contribution, but responding to the reviewers comments will result in improvements. i would encourage you to amend the manuscript to address the points they make where possible particularly in the case of reviewer one. Reviewer three is critical of your methods for model selection. I would suggest that you clarify the approach you adopted in the methods and results and include some discussion of the range of sensitivities. If you wish to add more discussion of the uncertainties surrounding model choice, it might be worth discussing reversible jump MCMC as an alternative to model averaging, perhaps as a suggestion for further work. Highlighting the dangers inherent in using p values or stepwise approaches would also be useful. I do not think that you need to adopt a new strategy, simply explain the potential problems a little more. Finally I would also encourage you to run a web of science search, to make sure that no meta-analyses are available on this topic. yours truly\n\nGavin Stewart",
+    "review_3": "Reviewer 1 \u00b7 Aug 26, 2013\nBasic reporting\nThe article meets all basic reporting standards. It is well written, clear, and appropriate for publication.\nExperimental design\nI am uncertain on the methodology. I am not familiar with animal tracking or movement ecology to any real extent. However, use of rangefinder cameras or some other sampling methodology to confirm that the method used herein, tracks in show, was effective. What if animals move a lot less when it has snowed? or move very differently? This could have been easily validated as a method and would have made this paper far more useful.\nValidity of the findings\nValid findings provided the methodology effectively estimates animal movements.\nAdditional comments\nGood paper. Useful. Please consider linking to theory a bit more with respect to habitat frag, connectivity theory, circuit theory for conservation, or whatever is appropriate. Also, are there any other additional stats to be added to explore impermeability of roads or natural mitigation opportunities that may be available to us to promote movement, ie any covariates in landscape etc.\n\nPlease also add #total number of tracks per species etc. I think you just present the crossings per km. Finally, can you apply rarefaction curves or some other measure to explore accumulation of species or tracks?\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Using multi-scale distribution and movement effects along a montane highway to identify optimal crossing locations for a large-bodied mammal community (v0.1)\". PeerJ https://doi.org/10.7287/peerj.189v0.1/reviews/1",
+    "review_4": "Paul Beier \u00b7 Aug 23, 2013\nBasic reporting\nThis paper describes and illustrates methods for identifying optimal locations for highway-crossing structures for a diverse mammal community. It demonstrates the superiority of free, remotely-sensed data to costly hand-digitized data, the superiority of multi-scale models to single scale models, and presents one approach to integrating across single-species models. The methods are appropriate (with one possible exception) and clearly described. Inferences are well-grounded in the data and statistical analyses.\nThe intro states a major objective that is not mentioned in the Abstract. The abstract should state: \u201cFreely-available remotely-sensed habitat landscape data were better than more costly, manually-digitized microhabitat maps in supporting models that identified good crossing sites; however models using both types of data were better yet.\u201d The Abstract should also mention that in 6 of 8 cases the multi-scale models performed better than models at any single scale.\nThe paper is generally well-written and clear. I think you should feel free to be a bit more punchy and colloquial. For instance, instead of dryly referring to the \u201chighway and off-highway transect models,\u201d you could refer to the latter as a model of the probability that an animal approaches the highway, and the former as a model that an animal, having reached the highway, crosses it. The title should highlight the idea of identifying optimal locations for crossing structures. And the table captions should go much further in telling the reader \u201cwhat\u2019s the point.\u201d Right now every caption is \u201cHere are some numbers. Guess why we put them there. Guess what they mean.\u201d See Kroodsma (2000. Auk 117:1081-1083).\nTable 1 can be improved. State that the unit of measurement is \u201c% of area\u201d unless otherwise stated. What is the unit of measurement for road/path, and railroad? Replace the acronym column with a \u201cvariable name\u201d column that uses clear descriptions (\u201c% unvegetated\u201d instead of \u201cno_veg\u201d and \u201c% dense conifer forest\u201d instead of \u201ccon_d\u201d). Reverse the column orders (1Variable name, 2Variable description, 3Source). In the caption or in a table footnote, state that the PAP variables were hand-digitized from 1-m photos (just to remind the reader).\nTable 2. Spell out words. If \u201cInd crossings\u201d means \u201csuccessful crossings\u201d use that term (spelled out).\nTable 3, 4, 5. Captions too terse.\nTable 3 Caption should explain the point of the table.\nIn Table 5, write out the variable names; transpose rows and columns so the names will fit.\nYou seem to use \u201ccovariate\u201d and \u201cparameter\u201d and \u201cpredictor\u201d and occasionally \u201ccharacteristic\u201d interchangeably, for the same things that are more commonly referred to as \u201cexplanatory variables\u201d or \u201cvariables.\u201d Not a big deal, but I think the terms \u201cvariables\u201d and \u201cpredictors\u201d are less-offputting and more accessible to most readers. Line 167: \u201clandscape parameter\u201d should probably be \u201cland cover or condition\u201d?\nExperimental design\nMy one real concern is with averaging the maps predicting the best crossing sites for individual species to produce a map intended to serve all species. If a site A is excellent (10 on a scale of 10) for 9 species and 0/10 for the 10th species, the average is 9. If site B scores 8 for all species, its average is 8. But I\u2019d argue that B is better, no? On lines 338-9, you state that landscape traits associated with preferred crossing sites differed among species, suggesting that some species could be poorly served by a site that serves the majority. Please discuss limitations and advantages of averaging as a way of combining predictions across species, and alternative (existing or potential) ways of combining across species. Perhaps a post-hoc procedure to identify \u201closers\u201d and accommodate their needs.\nOther than this one issue, this is well-designed and well-executed. Thank you for acknowledging me for feedback on the GIS analyses, even though I do not recall providing any such feedback. At this point I feel you have more to offer me on this topic than I can offer you.\nValidity of the findings\nThe Discussion might mention that the road permeability calculation almost certainly overestimates road permeability because the denominator (# of crossings of transects 10-900m from highway, mean 175 m from highway) does not reflect the numbers of animals who avoided the road at an average distance > 175 m.\nIn the abstract, I suggest deleting \u201clikely resulting in population fragmentation\u201d unless you present evidence this is true. Puma data are too sparse to support inference about fragmentation. For other species, perhaps 10% passage rates are enough to avert fragmentation.\nAdditional comments\nThank you for the opportunity to review this fine contribution.\nCite this review as\nBeier P (2013) Peer Review #2 of \"Using multi-scale distribution and movement effects along a montane highway to identify optimal crossing locations for a large-bodied mammal community (v0.1)\". PeerJ https://doi.org/10.7287/peerj.189v0.1/reviews/2",
+    "pdf_1": "https://peerj.com/articles/189v0.2/submission",
+    "pdf_2": "https://peerj.com/articles/189v0.1/submission",
+    "review_5": "Reviewer 3 \u00b7 Aug 14, 2013\nBasic reporting\nThe subject, methods, and results for this paper are presented clearly. The paper is very well written with only a few typographical errors. Some components of the methods, in particular sample design, could be more clearly explained. I make specific comments below for the authors.\nExperimental design\nI had no concerns with the substantive components of the paper. The data and methods fit the objectives very nicely. Results were followed by a lucid discussion of the findings. My primary concern is the choice of method for model selection and validation.\n\nThe authors conduct an exhaustive analysis of the 4 types of count distribution. Interesting, but often these details are subsumed in the unreported elements of an analysis.\n\nThe hybrid information theoretic(IT)/p-value approach for identifying the most \u2018parsimonious\u2019 model was overly complex, obtuse and technically flawed. The authors selected/combined models using a combination of best subsets, screening based on the p-values of the covariates, and finally model averaging. The best subsets method can result in models that are particular to a set of data \u2013 this is useful for exploring data, but not developing models that generalise to other study areas or time periods. Choosing individual variables based on their p-values assumes that the coefficient and SE are independent of other variables in the model. This is not the case \u2013 each covariate is dependent on the contribution of other covariates. Furthermore, this mixes two model selection philosophies: IT and hypothesis testing. I don\u2019t know if a simpler model selection process (i.e., identify 15-20 models based on existing literature and theory and select best model using AICc delta) would change the results, but the current approach is certainly awkward if not statistically incorrect.\n\nAIC provides a relative not absolute measure of model fit. Thus, we have some idea of what is the best model of the set, but not if that model has any ecological or predictive validity (i.e., even the best model could be a very poor predictor). The best model should be tested against a set of withheld data; this might involve testing the observed versus predicted probability for each count or by looking at the simple residuals (observed minus predicted counts for the withheld data). This is especially important when considering that these models will be used to predict counts on the highway and guide the location of mitigation strategies.\nValidity of the findings\nAlthough the model selection process was odd and perhaps incorrect (see previous section), the sampling and data are valid and the general statistical treatment was correct. The results support the objectives and conclusions of this work.\nAdditional comments\nGeneral Comments\nThe subject, methods, and results for this paper are presented clearly. The paper is very well written with only a few typographical errors. Some components of the methods, in particular sample design, could be more clearly explained. I make specific comments below for the authors.\n\nI had no concerns with the substantive components of the paper. The data and methods fit the objectives very nicely. Results were followed by a lucid discussion of the findings. My primary concern is the choice of method for model selection and validation.\n\nThe authors conduct an exhaustive analysis of the 4 types of count distribution. Interesting, but often these details are subsumed in the unreported elements of an analysis.\n\nThe hybrid information theoretic/p-value approach for identifying the most \u2018parsimonious\u2019 model was overly complex, obtuse and technically flawed. The authors selected/combined models using a combination of best subsets, screening based on the p-values of the covariates, and finally model averaging. The best subsets method can result in models that are particular to a set of data \u2013 this is useful for exploring data, but not developing models that generalise to other study areas or time periods. Choosing individual variables based on their p-values assumes that the coefficient and SE are independent of other variables in the model. This is not the case \u2013 each covariate is dependent on the contribution of other covariates. Furthermore, this mixes two model selection philosophies: IT and hypothesis testing. I don\u2019t know if a simpler model selection process (i.e., identify 15-20 models based on existing literature and theory and select best model using AICc delta) would change the results, but the current approach is certainly awkward if not statistically incorrect.\n\nAIC provides a relative not absolute measure of model fit. Thus, we have some idea of what is the best model of the set, but not if that model has any ecological or predictive validity (i.e., even the best model could be a very poor predictor). The best model should be tested against a set of withheld data; this might involve testing the observed versus predicted probability for each count or by looking at the simple residuals (observed minus predicted counts for the withheld data). This is especially important when considering that these models will be used to predict counts on the highway and guide the location of mitigation strategies.\n\nDetailed Comments\nAbstract\nThe abstract is well written and provides a balanced summary of the key objectives and results of the paper. However, some additional description of results (1 additional sentence) would be useful.\nL6: comma should follow \u201cCanada\u201d\n\nIntroduction\nL57-59: This sentence is somewhat confusing and should be simplified \u2013 doesn\u2019t modelling animal movement result in predictive models?\nL60: Unclear why there is an emphasis on migratory birds and reptiles; there is much published evidence of multi-scale habitat selection and movement by mammals as well.\nL77-84: The authors provide a very nice statement of the study objectives.\n\nMethods\nL104: The number and type of sampling locations should be a simple idea, but it is not clearly explained. In particular, why were transects surveyed versus the edge of a road and how do these 10 transects relate to the 9 transects on L128? Why is there a reference to a right-of-way \u2013 is this the highway or another type of linear feature? Did the authors survey highways and off-road/gravel roads?\nL108-109: The methods for sampling tracks/transects is key to understanding the results and should be reported not referenced.\nL202-208: Not necessary to derive the Vuong test; a citation will suffice.\nL217-231: This is an overly complex approach for model selection that some would classify as data dredging. Using a classic IT approach, each model should serve as a hypothesis. The approach used by the authors might result in a very predictive model, but the best subsets approach will likely capitalise on unique correlations that limit the generalisation of the \u2018best model\u2019. Furthermore, the method of selecting variables based on individual p-values is unusual and fails to consider the covariation of the covariates (i.e., the magnitude and variance of any one covariate is a function of other covariates in the model, thus, they cannot be considered as unique entities).\nL237: AIC provides a relative not absolute measure of model fit. See comment above.\nL239: What is meant by \u201cexhaustive model tests\u201d?\nL248: The identity/source of the polygon is unclear; is a polygon a pixel? If so, then just refer to pixels.\nL253: Are the \u201cmammal group estimates\u201d, estimates of the counts of crossings?\n\nResults\nL265-266: I recommend presenting tables in parentheses only; no reason to waste text to introduce a table.\nL279: Did the authors screen for excessively high multicollinearity using VIF/tolerance scores? The inclusion of the same variable at the three scales suggests that this might be a problem.\n\nDiscussion\nThe discussion was well formulated. However, there is much reference to mitigation strategies, but no mention of what those strategies might be. Two or three sentences describing mitigation would be of value to those readers not familiar with highway planning.\n\nTables and Figures\nTable 5: Not clear what is meant by \u201cselection frequencies\u201d; are these the averaged coefficients from the best models?\nThe figure captions are too brief \u2013 they should state where and what, relative to the objectives of the study.\n\nFigure 2: how does one differentiate between the predictive scores for the transect and the highway?\nCite this review as\nAnonymous Reviewer (2013) Peer Review #3 of \"Using multi-scale distribution and movement effects along a montane highway to identify optimal crossing locations for a large-bodied mammal community (v0.1)\". PeerJ https://doi.org/10.7287/peerj.189v0.1/reviews/3",
+    "all_reviews": "Review 1: Gavin Stewart \u00b7 Oct 8, 2013 \u00b7 Academic Editor\nACCEPT\nThank for you for clarifying the points raised by myself and the reviewers, and for constructive engagement with peer review. I think this paper is a great addition to the literature.\nReview 2: Gavin Stewart \u00b7 Sep 5, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nThe work is a useful contribution, but responding to the reviewers comments will result in improvements. i would encourage you to amend the manuscript to address the points they make where possible particularly in the case of reviewer one. Reviewer three is critical of your methods for model selection. I would suggest that you clarify the approach you adopted in the methods and results and include some discussion of the range of sensitivities. If you wish to add more discussion of the uncertainties surrounding model choice, it might be worth discussing reversible jump MCMC as an alternative to model averaging, perhaps as a suggestion for further work. Highlighting the dangers inherent in using p values or stepwise approaches would also be useful. I do not think that you need to adopt a new strategy, simply explain the potential problems a little more. Finally I would also encourage you to run a web of science search, to make sure that no meta-analyses are available on this topic. yours truly\n\nGavin Stewart\nReview 3: Reviewer 1 \u00b7 Aug 26, 2013\nBasic reporting\nThe article meets all basic reporting standards. It is well written, clear, and appropriate for publication.\nExperimental design\nI am uncertain on the methodology. I am not familiar with animal tracking or movement ecology to any real extent. However, use of rangefinder cameras or some other sampling methodology to confirm that the method used herein, tracks in show, was effective. What if animals move a lot less when it has snowed? or move very differently? This could have been easily validated as a method and would have made this paper far more useful.\nValidity of the findings\nValid findings provided the methodology effectively estimates animal movements.\nAdditional comments\nGood paper. Useful. Please consider linking to theory a bit more with respect to habitat frag, connectivity theory, circuit theory for conservation, or whatever is appropriate. Also, are there any other additional stats to be added to explore impermeability of roads or natural mitigation opportunities that may be available to us to promote movement, ie any covariates in landscape etc.\n\nPlease also add #total number of tracks per species etc. I think you just present the crossings per km. Finally, can you apply rarefaction curves or some other measure to explore accumulation of species or tracks?\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Using multi-scale distribution and movement effects along a montane highway to identify optimal crossing locations for a large-bodied mammal community (v0.1)\". PeerJ https://doi.org/10.7287/peerj.189v0.1/reviews/1\nReview 4: Paul Beier \u00b7 Aug 23, 2013\nBasic reporting\nThis paper describes and illustrates methods for identifying optimal locations for highway-crossing structures for a diverse mammal community. It demonstrates the superiority of free, remotely-sensed data to costly hand-digitized data, the superiority of multi-scale models to single scale models, and presents one approach to integrating across single-species models. The methods are appropriate (with one possible exception) and clearly described. Inferences are well-grounded in the data and statistical analyses.\nThe intro states a major objective that is not mentioned in the Abstract. The abstract should state: \u201cFreely-available remotely-sensed habitat landscape data were better than more costly, manually-digitized microhabitat maps in supporting models that identified good crossing sites; however models using both types of data were better yet.\u201d The Abstract should also mention that in 6 of 8 cases the multi-scale models performed better than models at any single scale.\nThe paper is generally well-written and clear. I think you should feel free to be a bit more punchy and colloquial. For instance, instead of dryly referring to the \u201chighway and off-highway transect models,\u201d you could refer to the latter as a model of the probability that an animal approaches the highway, and the former as a model that an animal, having reached the highway, crosses it. The title should highlight the idea of identifying optimal locations for crossing structures. And the table captions should go much further in telling the reader \u201cwhat\u2019s the point.\u201d Right now every caption is \u201cHere are some numbers. Guess why we put them there. Guess what they mean.\u201d See Kroodsma (2000. Auk 117:1081-1083).\nTable 1 can be improved. State that the unit of measurement is \u201c% of area\u201d unless otherwise stated. What is the unit of measurement for road/path, and railroad? Replace the acronym column with a \u201cvariable name\u201d column that uses clear descriptions (\u201c% unvegetated\u201d instead of \u201cno_veg\u201d and \u201c% dense conifer forest\u201d instead of \u201ccon_d\u201d). Reverse the column orders (1Variable name, 2Variable description, 3Source). In the caption or in a table footnote, state that the PAP variables were hand-digitized from 1-m photos (just to remind the reader).\nTable 2. Spell out words. If \u201cInd crossings\u201d means \u201csuccessful crossings\u201d use that term (spelled out).\nTable 3, 4, 5. Captions too terse.\nTable 3 Caption should explain the point of the table.\nIn Table 5, write out the variable names; transpose rows and columns so the names will fit.\nYou seem to use \u201ccovariate\u201d and \u201cparameter\u201d and \u201cpredictor\u201d and occasionally \u201ccharacteristic\u201d interchangeably, for the same things that are more commonly referred to as \u201cexplanatory variables\u201d or \u201cvariables.\u201d Not a big deal, but I think the terms \u201cvariables\u201d and \u201cpredictors\u201d are less-offputting and more accessible to most readers. Line 167: \u201clandscape parameter\u201d should probably be \u201cland cover or condition\u201d?\nExperimental design\nMy one real concern is with averaging the maps predicting the best crossing sites for individual species to produce a map intended to serve all species. If a site A is excellent (10 on a scale of 10) for 9 species and 0/10 for the 10th species, the average is 9. If site B scores 8 for all species, its average is 8. But I\u2019d argue that B is better, no? On lines 338-9, you state that landscape traits associated with preferred crossing sites differed among species, suggesting that some species could be poorly served by a site that serves the majority. Please discuss limitations and advantages of averaging as a way of combining predictions across species, and alternative (existing or potential) ways of combining across species. Perhaps a post-hoc procedure to identify \u201closers\u201d and accommodate their needs.\nOther than this one issue, this is well-designed and well-executed. Thank you for acknowledging me for feedback on the GIS analyses, even though I do not recall providing any such feedback. At this point I feel you have more to offer me on this topic than I can offer you.\nValidity of the findings\nThe Discussion might mention that the road permeability calculation almost certainly overestimates road permeability because the denominator (# of crossings of transects 10-900m from highway, mean 175 m from highway) does not reflect the numbers of animals who avoided the road at an average distance > 175 m.\nIn the abstract, I suggest deleting \u201clikely resulting in population fragmentation\u201d unless you present evidence this is true. Puma data are too sparse to support inference about fragmentation. For other species, perhaps 10% passage rates are enough to avert fragmentation.\nAdditional comments\nThank you for the opportunity to review this fine contribution.\nCite this review as\nBeier P (2013) Peer Review #2 of \"Using multi-scale distribution and movement effects along a montane highway to identify optimal crossing locations for a large-bodied mammal community (v0.1)\". PeerJ https://doi.org/10.7287/peerj.189v0.1/reviews/2\nReview 5: Reviewer 3 \u00b7 Aug 14, 2013\nBasic reporting\nThe subject, methods, and results for this paper are presented clearly. The paper is very well written with only a few typographical errors. Some components of the methods, in particular sample design, could be more clearly explained. I make specific comments below for the authors.\nExperimental design\nI had no concerns with the substantive components of the paper. The data and methods fit the objectives very nicely. Results were followed by a lucid discussion of the findings. My primary concern is the choice of method for model selection and validation.\n\nThe authors conduct an exhaustive analysis of the 4 types of count distribution. Interesting, but often these details are subsumed in the unreported elements of an analysis.\n\nThe hybrid information theoretic(IT)/p-value approach for identifying the most \u2018parsimonious\u2019 model was overly complex, obtuse and technically flawed. The authors selected/combined models using a combination of best subsets, screening based on the p-values of the covariates, and finally model averaging. The best subsets method can result in models that are particular to a set of data \u2013 this is useful for exploring data, but not developing models that generalise to other study areas or time periods. Choosing individual variables based on their p-values assumes that the coefficient and SE are independent of other variables in the model. This is not the case \u2013 each covariate is dependent on the contribution of other covariates. Furthermore, this mixes two model selection philosophies: IT and hypothesis testing. I don\u2019t know if a simpler model selection process (i.e., identify 15-20 models based on existing literature and theory and select best model using AICc delta) would change the results, but the current approach is certainly awkward if not statistically incorrect.\n\nAIC provides a relative not absolute measure of model fit. Thus, we have some idea of what is the best model of the set, but not if that model has any ecological or predictive validity (i.e., even the best model could be a very poor predictor). The best model should be tested against a set of withheld data; this might involve testing the observed versus predicted probability for each count or by looking at the simple residuals (observed minus predicted counts for the withheld data). This is especially important when considering that these models will be used to predict counts on the highway and guide the location of mitigation strategies.\nValidity of the findings\nAlthough the model selection process was odd and perhaps incorrect (see previous section), the sampling and data are valid and the general statistical treatment was correct. The results support the objectives and conclusions of this work.\nAdditional comments\nGeneral Comments\nThe subject, methods, and results for this paper are presented clearly. The paper is very well written with only a few typographical errors. Some components of the methods, in particular sample design, could be more clearly explained. I make specific comments below for the authors.\n\nI had no concerns with the substantive components of the paper. The data and methods fit the objectives very nicely. Results were followed by a lucid discussion of the findings. My primary concern is the choice of method for model selection and validation.\n\nThe authors conduct an exhaustive analysis of the 4 types of count distribution. Interesting, but often these details are subsumed in the unreported elements of an analysis.\n\nThe hybrid information theoretic/p-value approach for identifying the most \u2018parsimonious\u2019 model was overly complex, obtuse and technically flawed. The authors selected/combined models using a combination of best subsets, screening based on the p-values of the covariates, and finally model averaging. The best subsets method can result in models that are particular to a set of data \u2013 this is useful for exploring data, but not developing models that generalise to other study areas or time periods. Choosing individual variables based on their p-values assumes that the coefficient and SE are independent of other variables in the model. This is not the case \u2013 each covariate is dependent on the contribution of other covariates. Furthermore, this mixes two model selection philosophies: IT and hypothesis testing. I don\u2019t know if a simpler model selection process (i.e., identify 15-20 models based on existing literature and theory and select best model using AICc delta) would change the results, but the current approach is certainly awkward if not statistically incorrect.\n\nAIC provides a relative not absolute measure of model fit. Thus, we have some idea of what is the best model of the set, but not if that model has any ecological or predictive validity (i.e., even the best model could be a very poor predictor). The best model should be tested against a set of withheld data; this might involve testing the observed versus predicted probability for each count or by looking at the simple residuals (observed minus predicted counts for the withheld data). This is especially important when considering that these models will be used to predict counts on the highway and guide the location of mitigation strategies.\n\nDetailed Comments\nAbstract\nThe abstract is well written and provides a balanced summary of the key objectives and results of the paper. However, some additional description of results (1 additional sentence) would be useful.\nL6: comma should follow \u201cCanada\u201d\n\nIntroduction\nL57-59: This sentence is somewhat confusing and should be simplified \u2013 doesn\u2019t modelling animal movement result in predictive models?\nL60: Unclear why there is an emphasis on migratory birds and reptiles; there is much published evidence of multi-scale habitat selection and movement by mammals as well.\nL77-84: The authors provide a very nice statement of the study objectives.\n\nMethods\nL104: The number and type of sampling locations should be a simple idea, but it is not clearly explained. In particular, why were transects surveyed versus the edge of a road and how do these 10 transects relate to the 9 transects on L128? Why is there a reference to a right-of-way \u2013 is this the highway or another type of linear feature? Did the authors survey highways and off-road/gravel roads?\nL108-109: The methods for sampling tracks/transects is key to understanding the results and should be reported not referenced.\nL202-208: Not necessary to derive the Vuong test; a citation will suffice.\nL217-231: This is an overly complex approach for model selection that some would classify as data dredging. Using a classic IT approach, each model should serve as a hypothesis. The approach used by the authors might result in a very predictive model, but the best subsets approach will likely capitalise on unique correlations that limit the generalisation of the \u2018best model\u2019. Furthermore, the method of selecting variables based on individual p-values is unusual and fails to consider the covariation of the covariates (i.e., the magnitude and variance of any one covariate is a function of other covariates in the model, thus, they cannot be considered as unique entities).\nL237: AIC provides a relative not absolute measure of model fit. See comment above.\nL239: What is meant by \u201cexhaustive model tests\u201d?\nL248: The identity/source of the polygon is unclear; is a polygon a pixel? If so, then just refer to pixels.\nL253: Are the \u201cmammal group estimates\u201d, estimates of the counts of crossings?\n\nResults\nL265-266: I recommend presenting tables in parentheses only; no reason to waste text to introduce a table.\nL279: Did the authors screen for excessively high multicollinearity using VIF/tolerance scores? The inclusion of the same variable at the three scales suggests that this might be a problem.\n\nDiscussion\nThe discussion was well formulated. However, there is much reference to mitigation strategies, but no mention of what those strategies might be. Two or three sentences describing mitigation would be of value to those readers not familiar with highway planning.\n\nTables and Figures\nTable 5: Not clear what is meant by \u201cselection frequencies\u201d; are these the averaged coefficients from the best models?\nThe figure captions are too brief \u2013 they should state where and what, relative to the objectives of the study.\n\nFigure 2: how does one differentiate between the predictive scores for the transect and the highway?\nCite this review as\nAnonymous Reviewer (2013) Peer Review #3 of \"Using multi-scale distribution and movement effects along a montane highway to identify optimal crossing locations for a large-bodied mammal community (v0.1)\". PeerJ https://doi.org/10.7287/peerj.189v0.1/reviews/3\nReview 6: \n",
+    "LLM_extracted_review": "[\"Review 1: The paper is a great addition to the literature.\\n\\nReview 2: The work is a useful contribution, but responding to the reviewers' comments will result in improvements. Clarify the approach for model selection and include discussion of uncertainties surrounding model choice. Highlight dangers of using p-values or stepwise approaches. Run a web of science search for existing meta-analyses on the topic.\\n\\nReview 3: The article meets basic reporting standards and is well written. Methodology is uncertain; validation of tracking methods is needed. Consider linking to habitat fragmentation and connectivity theory. Additional statistics on road impermeability and natural mitigation opportunities should be included. Present total number of tracks per species and apply rarefaction curves for species accumulation.\\n\\nReview 4: The paper demonstrates the superiority of free, remotely-sensed data and multi-scale models. The introduction states a major objective not mentioned in the abstract. Improve table captions and clarify variable names. Discuss limitations of averaging predictions across species. The road permeability calculation may overestimate actual permeability. Delete unsupported claims about population fragmentation.\\n\\nReview 5: The paper is well written but some methods could be clearer. The model selection process is overly complex and potentially flawed. A simpler model selection process is recommended. AIC provides a relative measure of model fit; the best model should be tested against withheld data. Additional description in the abstract would be useful. Clarify sampling methods and the relationship between transects. The discussion should include specific mitigation strategies.\\n\\nReview 6: [No content provided for Review 6.]\"]"
+}

peerj_json_files/PeerJ_Json_14.json ADDED Viewed

The diff for this file is too large to render. See raw diff

peerj_json_files/PeerJ_Json_140.json ADDED Viewed

The diff for this file is too large to render. See raw diff

peerj_json_files/PeerJ_Json_141.json ADDED Viewed

	@@ -0,0 +1,18 @@

+{
+    "v1_Abstract": "BioNames is a web database of taxonomic names for animals, linked to the primary literature and, wherever possible, to phylogenetic trees. It aims to provide a taxonomic \"dashboard\" where at a glance we can see a summary of the taxonomic and phylogenetic information we have for a given taxon and hence provide a quick answer to the basic question \"what is this taxon?\" BioNames combines classifications from the Global Biodiversity Information Facility (GBIF) and GenBank, images from the Encyclopedia of Life (EOL), animal names from the Index of Organism Names (ION), and bibliographic data from multiple sources including the Biodiversity Heritage Library (BHL) and CrossRef. The user interface includes display of full text articles, interactive timelines of taxonomic publications, and zoomable phylogenies. It is available at http://bionames.org.",
+    "v2_Abstract": "BioNames is a web database of taxonomic names for animals, linked to the primary literature and, wherever possible, to phylogenetic trees. It aims to provide a taxonomic \"dashboard\" where at a glance we can see a summary of the taxonomic and phylogenetic information we have for a given taxon and hence provide a quick answer to the basic question \"what is this taxon?\" BioNames combines classifications from the Global Biodiversity Information Facility (GBIF) and GenBank, imagery from the Encyclopedia of Life (EOL), animal names from the Index of Organism Names (ION), and bibliographic data from multiple sources including the Biodiversity Heritage Library (BHL) and CrossRef. The user interface includes display of full text articles, interactive timelines of taxonomic publications, and zoomable phylogenies. It is available at http://bionames.org.",
+    "v1_text": "materials & methods : BioNames integrates data on taxonomic names and classifications, literature, and phylogenies from a variety of sources. Given the inevitable differences in how different databases treat the same data (as well as internal inconsistencies within individual databases), considerable effort must be spent cleaning and reconciling data. Much of this process involves mapping \"strings\" to \"things\" (Bollacker et al. 2008), or more precisely, mapping strings to identifiers for things. results : BioNames comprises a CouchDB database and a web interface. Key features of the interface are outlined below. 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t acknowledgements : I thank Ryan Schenk for his work on the BioNames, and Cyndy Parr (EOL) for managing the EOL Computational Challenge and providing helpful feedback on the development of BioNames. Mark Holder and an anonymous reviewer provided detailed and helpful comments on the 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t manuscript. Some of the ideas in this manuscriptpresented here were first explored in a talk at the \"Anchoring Biodiversity Information: From Sherborn to the 21st century and beyond\" symposium held at The Natural History Museum, London, October 28th 2011. I thank Ellinor Michel for the invitation to speak at that meeting. discussion : The EOL Computational Data Challenge imposed a deadline on the first release of BioNames, however development of both the database and web interface is ongoing. Below I discuss some potential applications and future directions. impact of taxonomic literature : The taxonomic community has long felt disadvantaged by the role of citation-based \"impact factor\" in assessing the importance of taxonomic research (Garfield 2001; Krell 2000; Werner 2006) especially as much of the taxonomic literature appears in relatively low-impact journals. A common proposal is to include citations to the taxonomic authority for every name mentioned in a scientific paper (W\u00e4gele et al. 2011). Regardless of the merits of this idea, in practice these citations are often hard to locate, which is another motivation for BioNames. There is additional value in surfacing identifiers for the taxonomic literature. In addition to helping construct citation networks, global identifiers can facilitate computing other measures of the value of a taxonomic paper. There is a growing interest in additional measures of postpublication impact of a publication in terms of activity such as social bookmarking, and commentary on web sites (\"alt-metrics\") (Yan and Gerstein 2011). Gathering these metrics is greatly facilitated by using standard bibliographic identifiers (otherwise, how do we know whether two commentators are discussing the same article or not?). If taxonomic literature is be part of this burgeoning conversation then it needs to be able to be identified unambiguously. taxon names : At present the taxonomic scope of BioNames is restricted to names covered by the International Code of Zoological Nomenclature (animals and those eukaryotes not covered by the International Code of Nomenclature for algae, fungi, and plants). Taxonomic names were obtained from the Index of Organism Names (ION; http://www.organismnames.com). Each name in ION has a Life Science Identifier (LSID) (Martin et al. 2005) which uniquely identifies that name. LSIDs can be dereferenced to return metadata in Resource Description Framework format (RDF) (Page 2008b). I used the TDWG LSID resolver (http://lsid.tdwg.org) to obtain the metadata for each LSID. ION LSIDs provide basic information on a taxonomic name using the TDWG Taxon Name LSID Ontology (http://rs.tdwg.org/ontology/voc/TaxonName), in many cases including bibliographic details for the publication where the name first appeared (Fig. 2). The publication in which the name first appeared is listed in the contents of the \"PublishedIn\" property. In the example in Figure 2 this is the string \"Description of a new species of Pinnotheres, and redescription of P. novaezelandiae (Brachyura: Pinnotheridae). New Zealand Journal of Zoology, 10(2) 1983: 151-162. 158 (Zoological Record Volume 120)\". I used regular 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t expressions to parse citation strings into their component parts (e.g., article title, journal, volume, pagination), and then attempted to locate the corresponding reference in an external database (see below). database : Once aggregated, cleaned, and reconciled, the data was converted to JSON (JavaScript Object Notation) and stored in a CouchDB database. CouchDB is a \"NoSQL\" document database that stores objects in JSON format. Unlike typical SQL databases, CouchDB does not have a database schema and does not support ad hoc queries. Instead CouchDB accepts semi-structured documents, and the user developer defines fixed queries or \"views\" (Anderson et al. 2010). phylogeny viewer : Screenshot of phylogeny from PhyLoTA as displayed in BioNames. The user can zoom in and out and pan, as well as change the layout of the tree. PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t Figure 11 Relative importance of different publishers of taxonomic literature Bubble chart showing relative numbers of taxonomic articles made available online by different publishers. PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t availability : BioNames is accessible at http://bionames.org. The source code used to build the web site is available on GitHub http://github.com/rdmpage/bionames. Scripts used to fetch, clean, and reconcile the data are archived in http://github.com/rdmpage/bionames-data eol challenge : In response to the Encyclopedia of Life (EOL) Computational Data Challenge (http://eol.org/info/323) I constructed BioNames (http://bionames.org) (Page 2012). Its goal is to create a database of taxonomic names for animals linked to the primary literature and, wherever possible, to phylogenetic trees. Using existing globally unique identifiers for taxonomic names, concepts, publications, and sequences rather than cryptic text strings (for example, abbreviated bibliographic citations) simplifies the task of linking \u2014 we can rely on exact matching of identifiers rather than approximate matching between names for what may or may not be the same entity. This is particularly relevant once we start to aggregate information from different databases, where the same information (e.g., a publication) may be represented by different strings. Furthermore, if we use existing identifiers we increase the potential to connect to other databases (Page 2008a). This paper outlines how BioNames was built, describes the user interface, and discusses future plans. 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t bibliographic identifiers : When populating BioNames every effort has been made to map each bibliographic string to a corresponding identifier, such as a Digital Object identifier (DOI). For the example in Figure 2, the citation string \"Description of a new species of Pinnotheres, and redescription of P. novaezelandiae (Brachyura: Pinnotheridae). New Zealand Journal of Zoology, 10(2) 1983: 151- 162. 158 (Zoological Record Volume 120)\" corresponds to the article with the DOI 10.1080/03014223.1983.10423904. Once we have a DOI, we can then use services such as those provided by CrossRef (http://www.crossref.org) to retrieve author and publisher information for an article (see Fig. 11 below for one use of publisher information). While DOIs are the best-known bibliographic identifier, there are several others that are relevant to the taxonomic literature (Page 2009). DOIs are themselves based on Handles (http://hdl.handle.net), an identifier widely used by digital repositories such as DSpace (Smith et al. 2003). A number of journals, such as the Bulletins and Novitates of the American Museum of Natural History are available in DSpace repositories and consequently have Handles. Other major archives such as JSTOR (http://www.jstor.org/) and the Japanese National Institute of Informatics (CiNii; http://ci.nii.ac.jp/) have their own unique identifiers (typically integer numbers that are part of a URL). 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t Having a variety of identifiers can complicate the task of finding existing identifiers for a particular publication. Whereas for some identifiers, such as DOIs, and CiNii NAIDs (National Institute of Informatics Article IDs), and BioStor reference ids there are search tools (e.g., http://search.crossref.org) or OpenURL resolvers for this task (Van de Sompel & Beit-Arie 2001) (e.g., http://biostor.org/openurl), for other identifiers there may be no obvious way to find the identifier other than by using a search engine. Another strategy is to build a local database of bibliographic data and match citations strings to that database. I used Mendeley (http://www.mendeley.com) to store bibliographic data harvested from journal or taxon-specific web pages in publicly accessible \u201cgroups\u201d, and then queried the local copy of the Mendeley Desktop database to search for references that matched the citation strings. For the example in Figure 2, the citation string \"Description of a new species of Pinnotheres, and redescription of P. novaezelandiae (Brachyura: Pinnotheridae). New Zealand Journal of Zoology, 10(2) 1983: 151-162. 158 (Zoological Record Volume 120)\" corresponds to the article with the DOI 10.1080/03014223.1983.10423904. Once we have a DOI, we can then use services such as those provided by CrossRef (http://www.crossref.org) to retrieve author and publisher information for an article (see Fig. 11 below for one use of publisher information). Identifiers also exist for aggregations of publications, such as journals. The historical practice of abbreviating journal titles in citations has led to a plethora of ways to refer to the same journal. For example, the BioStor database (http://biostor.org; Page 2011b) has accumulated more than ten variations on the name of the journal Bulletin of Zoological Nomenclature (such as \"Bull Zool Nomen\", \"Bull Zool Nom.\", \"Bull. Zool. Nomencl.\", etc.). This practice, presumably motivated by the desire to conserve space on the printed page, complicates efforts to match citations to identifiers. One approach to tackling this problem is to map abbreviations to journal-level 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t globally unique identifiers, such as International Standard Serial Numbers (ISSNs) (for the Bulletin of Zoological Nomenclature the ISSN is 0007-5167). In addition to reducing ambiguity, there are web services such as that provided by WorldCat (http://www.worldcat.org) that take ISSNs and return the history of name changes for a journal, which in turn can help clarify the (often complicated) history of long-lived journals.. Where possible each journal in BioNames was associated with its corresponding ISSN. If an ISSN is not available for a journal, then the corresponding OCLC Control Number was used as the identifier for the journal. documents : Taxonomic publications are available under a variety of licenses, ranging from explicitly open access licenses (MacCallum 2007) to articles that are \"free\", to articles that are behind a paywall. Archives such as JSTOR and CiNii have a mixture of free and subscription-based content. Many smaller journals, often published by scientific societies, are providing their content online for free, if not explicitly under an open license. The Biodiversity Heritage Library (the single largest source of taxonomic articles in BioNames, Fig. 11) makes its content available under a Creative Commons license. Where PDFs were available online either \"for free\" or under open access, these were downloaded and locally cached. Pages were extracted and converted into bitmap images for subsequent display in a web browser. Closed-access publications that are available online are linked to by their identifier (e.g., DOI). Access to some of these publications may be available for short-term \"rent\" by services such as DeepDyve (http://www.deepdyve.com): where possible BioNames includes a link those services. 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t clustering taxonomic names : Taxonomic names comprise a \"canonical\" name and the name's authorship, for example Homo sapiens Linnaeus comprises the canonical name \"Homo sapiens\" and the authorship string \"Linnaeus\". Names in taxonomic databases such as ION display numerous variations in spelling of authors and/or variation in the year of publication, and instances of the same canonical name published by different authors (e.g., homonyms), so the names were clustered before populating BioNames. For each set of taxon names with the same canonical name the authorship was compared. If one name lacked an author and the other had an author, the names were automatically merged into a cluster. Given more than two names a graph was constructed where the nodes are the authorship strings, and a pair of nodes is connected if their corresponding strings were sufficiently similar. String similarity was computed by converting the strings to a \"finger print\" comprising lower case letters with all accented characters replaced by non-accented equivalents, and all punctuation removed, then finding the longest common subsequence of the two strings. By definition the characters in a common subsequence do not need to be consecutive, so the method allows for insertion and deletion of characters. If the length of the subsequence relative to the each of the two input strings was longer than a specified threshold (by default, 0.8, where identical strings have a similarity of 1.0) then the two author strings were connected by an edge in the graph. The components of the graph correspond to clusters of names with similar authorship strings, and were treated as being the same name. Figure 3 shows a graph for the different names that all have \"Rhacophorus\" as the canonical name. Mapping names to taxa 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t BioNames includes two taxonomic classifications, sourced from GBIF (http://uat.gbif.org/dataset/d7dddbf4-2cf0-4f39-9b2a-bb099caae36c) and NCBI (ftp://ftp.ncbi.nih.gov/pub/taxonomy), respectively. These provide the user with a way to navigate through taxonomic names, as well as view data associated with each classification (e.g., phylogenies). These classifications also provide an explicit definition of the scope of a taxon (i.e., the \u201ctaxon concept\u201d). A higher taxon comprises the set of taxa below that taxon in the classification. A terminal taxon (the lowest taxon in a classification) in GBIF can be defined as the set of occurrences linked to that taxon, a terminal taxon in NCBI can be defined as the set of sequences linked to that taxon. Ideally there would be a one-to-one mapping between a taxonomic name and a taxon, but complications often arise. In addition to the well-known problems of synonymy (more than one name for the same taxon) and homonymy (the same name used for different taxa), name and taxon databases may store slightly different representations of the same name. For example, ION has four records for the name \"Nystactes\" (each name is followed by its LSID): Nystactes urn:lsid:organismnames.com:name:2787598 Nystactes Bohlke urn:lsid:organismnames.com:name:2735131 Nystactes Gloger 1827 urn:lsid:organismnames.com:name:4888093 Nystactes Kaup 1829 urn:lsid:organismnames.com:name:4888094 GBIF has three taxa with this name (the number is the GBIF species id): Nystactes B\u00f6hlke, 1957 2403398 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t Nystactes Gloger, 1827 2475109 Nystactes Kaup, 1829 3239722 Note the differences in the name string (\"o\" versus \"\u00f6\" in \"B\u00f6hlke\", presence or absence of years and commas). To automate the mapping of names to concepts in cases like this I constructed a bipartite graph where the nodes are taxon names, divided into two sets based upon which database they came from (e.g., one set of names from ION, the other from GBIF). I then connect the nodes of the graph by edges whose weights are the similarity of the two strings computed using the longest common subsequence that the two strings share. For example, Figure 4 shows the graph for \"Nystactes\", where the nodes corresponding to ION names are enclosed in ovals, and the names from GBIF are enclosed in rectangles. Computing the maximum weighted bipartite matching of this graph creates a map between the two sets of names. Ideally GBIF should have only one entry for Nystactes because each animal name (with a few exceptions) must be unique. If a newer name has already been published before, then it should be replaced by a new name. In this case, Nystactes (B\u00f6hlke 1957) has since been replaced by Nystactichthys (B\u00f6hlke 1958), and Nystactes (Kaup 1829) by Paramyotis (Bianchi 1916). Unfortunately these changes have not yet percolated their way from the primary literature into the GBIF taxonomy. Graph depicting similarity between different authorship strings associated with the name \"Rhacophorus\". The components of this graph correspond to the name clusters recognised by BioNames. PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t Figure 4 Matching taxonomic names to taxa Bipartite graph of string similarities between taxonomic names containing the string \"Nystactes\" in the ION and GBIF databases. Solid edges in the graph represent the maximum weighted bipartite matching, and define the mapping between ION name (ovals) and GBIF names (rectangles). PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t Figure 5 images : To help the user recognise the taxa being displayed images for as many taxa as possible were obtained using EOL's API, which provides access to both the images, and a mapping between GBIF and NCBI taxon concept identifiers and the corresponding record in EOL. 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t phylogenies : Phylogenies were obtained from the PhyLoTA database (http://phylota.net) (Sanderson et al. 2008). This database contains eukaryote phylogenies constructed from automatically assembled clusters of nucleotide sequences (loosely corresponding to \"genes\"). A MySQL data dump was downloaded (version 184, corresponding to the GenBank release of the same version number) and used to populate a local MySQL database. Metadata for the sequences in each phylogeny was obtained from the European Bioinformatics Institute (EBI; http://www.ebi.ac.uk), and used to populate the MySQL database with basic information such as taxon and locality information, as well as bibliographic details for the sources of the sequences. Phylogenies from PhyLOTA are rendered in an interactive viewer using the Scalable Vector Graphics (SVG) format. The user can zoom in and out, and change the drawing style. Terminal taxa with the same label have the same colour (Fig. 10). This makes it easier to recognise clusters of sequences from the same taxon (e.g., conspecific samples), as well as highlight possible errors (e.g., mislabelled or misidentified sequences). At present the colours are arbitrarily chosen, other schemes could be added in future (Lespinats and Fertil 2011). search : BioNames features a simple search interface that takes a scientific name and returns matching taxonomic names and concepts, together with any publications and phylogenies that contain the name. Figure 5 shows an example search result. Screenshot of the search results for a query BioNames. The results include names that match the query, taxon concepts from GBIF and NCBI with thumbnail images from EOL, phylogenies containing members of the genus, and relevant taxonomic publications. PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t Figure 6 document display : BioNames uses the DocumentCloud (https://github.com/documentcloud/document-viewer) viewer to display both PDFs, and page images from digital archives such as BioStor and Gallica (http://gallica.bnf.fr/) (Fig. 6). journals : Much of the work in populating BioNames comprises mapping citation to string to bibliographic identifiers and, where possible, linking those citations to full text. For each journal that has a ISSN, BioNames has a corresponding web page that lists all the articles from that journal that are in the database, and provides a graphical summary of how many of those articles have been located online (Fig. 7). timeline : BioNames can display timelines of the numbers of taxonomic names published in higher taxonomic groups, inspired by Taxatoy (Sarkar et al. 2008) (Fig. 8). For a given node in the 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t taxonomic hierarchy the children of that node are displayed as a treemap where the size of each cell is proportional to the log of the number of taxa in the subtree rooted on that child taxon. The number of names in that taxon published in each year is displayed as an interactive chart. Clicking on an individual year will list the corresponding publications for that year. taxa : Each GBIF or NCBI taxon in BioNames has a corresponding web page that lists the associated taxonomic names, publications linked to those names, and other relevant data (e.g., Fig. 9). dashboard : The BioNames web site features a \"dashboard\" which displays various summaries of the data it contains. For example, Fig. 11 shows a bubble chart of the number of articles different publishers have made available online. \"Publisher\" in this context is broadly defined to include digital 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t archives such as BioStor and JSTOR, repositories using DSpace, and commercial publishers such as Elsevier, Informa UK, Magnolia Press, Springer, and Wiley. 273 274 PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t links : BioNames makes extensive use of identifiers to clean and link data, but the real value of identifiers becomes apparent when they are shared, that is, when different databases use the same identifiers for the same entities, instead of minting their own. Reusing identifiers can enable unexpected connections between databases. For example, the PubMed biomedical literature database has a record (PMID:948206) for the paper \"Monograph on \u2018Lithoglyphopsis\u2019 aperta, the snail host of Mekong River Schistosomiasis\" (Davis et al. 1976). The PubMed record contains the abstract for the paper, but not a link to where the user can obtain a digital version of the paper. However, this reference is in a volume that has been scanned by the Biodiversity Heritage Library, and the article has been extracted by BioStor (http://biostor.org/reference/102054). If PubMed was linked to BHL, users of PubMed could go straight to the content of the article. But this is just the start. The Davis et al. paper also mentions museum specimens in the collection of the Academy of Natural Sciences of Drexel University, Philadelphia. Metadata for these specimens has been aggregated by GBIF, and the BioStor page for this article displays those links (http://biostor.org/reference/102054). In an ideal world we should be able seamlessly to traverse the path PubMed \u2192 BioStor \u2192 GBIF. Likewise, we should be able to traverse the path in the other direction. At present, a user of GBIF simply sees metadata for these specimens and a 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t locality map. They are unaware that these specimens have been cited in a paper (Davis et al. 1976) which demonstrates that the snails host the Mekong River schistosome. This connection would be trivial to make if the reciprocal link was made: GBIF \u2192 BioStor. Furthermore, a link BioStor \u2192 PubMed would give us access to Medical Subject Headings (MeSH) for the schistosome paper. Hence we could imagine ultimately searching a database of museum specimens (GBIF) using queries from a controlled vocabulary of biomedical terms (MeSH). Making these connections requires not only that we have digital identifiers, but also that where ever possible we reuse existing identifiers. In practice forging these links can be hard work (Page 2011a), and many links may be missing from existing databases (Miller et al. 2009). However, if we restrict ourselves to project-specific identifiers then we stymie attempts to create a network of connected biodiversity data. text mining : Much of the value of a scientific publication lies dormant unless it is accessible to text mining, which requires access to full text. Where possible BioNames stores information on the publisher of each article (Fig. 11), which could then be used to prioritise discussions with publishers on gaining access to full text (Van Noorden 2012). Fortunately, the single largest \"publisher\" of content in BioNames is BioStor (Page 2011b), which contains scans and OCR text from the Biodiversity Heritage Library. BHL makes its content available under a Creative Commons license, and so can be readily mined. Indeed, the text has already been indexed by tools that can recognise taxonomic names (Akella et al. 2012). 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t dark taxa : One of the original motivations for constructing BioNames is the rise of \"dark taxa\" in genomics databases (Page 2011c). These are taxa that have been sequenced and added to GenBank, but which lack formal Linnaean names. Typically they will have a name that comprises a genus name and some combination of letters and numbers to make the name unique within GenBank (e.g. a 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t specimen code or the first letter of the lastnames of the researchers that deposited the sequence). It is clear that some dark taxa do, in fact, have names. For example, consider the frog \"Gephyromantis aff. blanci MV-2005\" (NCBI taxonomy id 321743), which has a single DNA sequence AY848308 associated with it. This sequence was published as part of a DNA barcoding study (Vences et al. 2005). If we enter the accession number AY848308 into Google we find two documents, one the supplementary table for (Vences et al. 2005), the other a subsequent paper (Vences and Riva 2007) that describes the frog with this sequence as a new species, Gephyromantis runewsweeki. This example is relatively straightforward, but it still required significant time to track down the species description. A key question facing attempts to find names for dark taxa is whether the methods available can be scaled to handle the magnitude of the problem. Alternatively, one could argue that newer technologies such as DNA barcoding make classical taxonomy less relevant, and perhaps the effort in digitising older literature and exposing the taxonomic names it contains is misplaced. A counter argument would be that the taxonomic literature potentially contains a wealth of information on ecology, morphology and behaviour, often for taxa in areas that have been subsequently altered by human activity. Given the rarity of many taxa (Lim et al. 2011), and the uneven taxonomic and geographic distribution of taxonomic expertise (May 1998; Gaston and May 1992), for many species the only significant data on their biology may reside in the legacy literature (possibly under a different name (Solow et al. 1995)). As this legacy becomes more accessible through projects such as BHL (and services that build upon that project; Page 2011a) there will be considerable opportunities to mine that literature for basic biological data (Thessen et al. 2012). 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t publishing platform : Recently some taxonomic journals have begun to mark up taxonomic names and descriptions (Penev et al. 2010), which is a precursor to linking names and data together. But these developments leave open the problem of what these links will point to. If we have a database of all taxonomic names and the associated literature (such as BioNames aims to be for zoological names), then such a database would provide an obvious destination for those links. Indeed, ultimately, we could envisage publishing new taxonomic publications within such a database, so that each new publication becomes simply another document within the database (Gerstein and Junker 2002). In the same way, we could use automated methods to extend the process of tagging names, specimens and literature cited to the legacy literature (Page 2010), so that the entire body of taxonomic knowledge becomes a single interwoven web of names, citations, publications, and data. taberlet, p., coissac, e., pompanon, f., brochmann, c., & willerslev, e. : (2012). Towards next-generation biodiversity assessment using DNA metabarcoding. Molecular Ecology, 21(8), 2045\u20132050. doi:10.1111/j.1365-294X.2012.05470.x 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t Thessen, A. E., Cui, H., & Mozzherin, D. (2012). Applications of Natural Language Processing in Biodiversity Science. Advances in Bioinformatics, 2012, 1\u201317. doi:10.1155/2012/391574 Van de Sompel, H., & Beit-Arie, O. (2001). Open Linking in the Scholarly Information Environment Using the OpenURL Framework. D-Lib Magazine, 7(3). doi:10.1045/march2001-vandesompel Van Noorden, R. (2012). Trouble at the text mine. Nature, 483(7388), 134\u2013135. doi:10.1038/483134a Vences M, Riva IDL (2007) A new species of Gephyromantis from Ranomafana National Park, south-eastern Madagascar (Amphibia, Anura, Mantellidae). Spixiana 30(1): 135-143. Vences, M., Thomas, M., van der Meijden, A., Chiari, Y., & Vieites, D. R. (2005).Frontiers in Zoology, 2(1), 5. doi:10.1186/1742-9994-2-5 W\u00e4gele, H., Klussmann-Kolb, A., Kuhlmann, M., Haszprunar, G., Lindberg, D., Koch, A., & W\u00e4gele, J. W. (2011). The taxonomist - an endangered race. A practical proposal for its survival. Frontiers in Zoology, 8(1), 25. doi:10.1186/1742-9994-8-25 Werner, Y. L. (2006). The case of impact factor versus taxonomy: a proposal. Journal of Natural History, 40(21-22), 1285\u20131286. doi:10.1080/00222930600903660 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t Yan, K.-K., & Gerstein, M. (2011). The Spread of Scientific Information: Insights from the Web Usage Statistics in PLoS Article-Level Metrics. (A. Vespignani, Ed.)PLoS ONE, 6(5), e19917. doi:10.1371/journal.pone.0019917 Zhang, Z.-Q. (2006). The making of a mega-journal in taxonomy. Zootaxa, 1385, 67-68. 537 538 539 540 PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t Figure 1 taxonomy data model : Simplified diagram of the relationships between the core entities that make up taxonomy, such as authors, publications, taxon names, and taxa. Relationships between entities are represented by lines, those in black are the focus of BioNames. PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t Figure 2 rdf for taxon name : The RDF retrieved by dereferencing the LSID urn:lsid:organismnames.com:name:371873, which identifies the taxonomic name Pinnotheres atrinicola. PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t Figure 3 displaying an article : Screenshot of BioNames displaying a document from BioStor (Conle and Hennemann 2002). The document viewer can display page images, thumbnails, and (where available) text. PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t Figure 7 Screenshot of the page in BioNames for the journal Proceedings of the Entomological Society of Washington (ISSN 0013-8797). The centre column lists the articles in a volume selected by the user using the index on the left. The right hand column displays basic data about the journal, and a graphical display of how many articles have been mapped to a globally unique identifier. PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t Figure 8 Timeline of taxonomic names for birds Screenshot of the distribution overtime of publications of new names for birds (Aves). The treemap on the left displays taxa below Aves in the taxonomic hierarchy, the chart on the right displays the number of publications in each year that publish a new bird name. The user has clicked on \"2012\", resulting in a list of the papers published in that year appearing below the timeline. PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t Figure 9 bibliography for a taxon : Screenshot of the bibliography tab on a taxon page in BioNames. This example shows the publications relevant to the bat genus Rousettus, including those for synonyms. The user can select publications from a given time slice and/or combination of synonyms. PeerJ reviewing PDF | (v2013:08:749:1:0:NEW 2 Oct 2013) R ev ie w in g M an us cr ip t Figure 10",
+    "v2_text": "results : BioNames comprises a CouchDB database and a web interface. Key features of the interface are outlined below. acknowledgements : I thank Ryan Schenk for his work on the BioNames, and Cyndy Parr (EOL) for managing the EOL Computational Challenge and providing helpful feedback on the development of BioNames. Some of the ideas in this manuscript were first explored in a talk at the \"Anchoring Biodiversity Information: From Sherborn to the 21st century and beyond\" symposium held at The Natural History Museum, London, October 28th 2011. I thank Ellinor Michel for the invitation to speak at that meeting. 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t discussion : The EOL Computational Data Challenge imposed a deadline on the first release of BioNames, however development of both the database and web interface is ongoing. Below I discuss some potential applications and future directions. impact of taxonomic literature : The taxonomic community has long felt disadvantaged by the role of citation-based \"impact factor\" in assessing the importance of taxonomic research (Garfield 2001; Krell 2000; Werner 2006) especially as much of the taxonomic literature appears in relatively low-impact journals. A common proposal is to include citations to the taxonomic authority for every name mentioned in a scientific paper (W\u00e4gele et al. 2011). Regardless of the merits of this idea, in practice these citations are often hard to locate, which is another motivation for BioNames. There is additional value in surfacing identifiers for the taxonomic literature. In addition to helping construct citation networks, global identifiers can facilitate computing other measures of the value of a taxonomic paper. There is a growing interest in additional measures of postpublication impact of a publication in terms of activity such as social bookmarking, and commentary on web sites (\"alt-metrics\") (Yan and Gerstein 2011). Gathering these metrics is greatly facilitated by using standard bibliographic identifiers (otherwise, how do we know whether two commentators are discussing the same article or not?). If taxonomic literature is be part of this burgeoning conversation then it needs to be able to be identified unambiguously. taxon names : At present the taxonomic scope of BioNames is restricted to names covered by the International Code of Zoological Nomenclature (animals and those eukaryotes not covered by the International Code of Nomenclature for algae, fungi, and plants). Taxonomic names were obtained from the Index of Organism Names (ION; http://www.organismnames.com). Each name in ION has a Life Science Identifier (LSID) (Martin et al. 2005) which uniquely identifies that name. LSIDs can be dereferenced to return metadata in Resource Description Framework format (RDF) (Page 2008b). ION LSIDs provide basic information on a taxonomic name using the TDWG Taxon Name LSID Ontology (http://rs.tdwg.org/ontology/voc/TaxonName), in many cases including bibliographic details for the publication where the name first appeared (Fig. 2). The publication in which the name first appeared is listed in the contents of the \"PublishedIn\" property. In the example in Figure 2 this is the string \"Description of a new species of Pinnotheres, and redescription of P. novaezelandiae (Brachyura: Pinnotheridae). New Zealand Journal of Zoology, 10(2) 1983: 151-162. 158 (Zoological Record Volume 120)\". I used regular expressions to parse citation strings into their component parts (e.g., article title, journal, volume, pagination), and then attempted to locate the corresponding reference in an external database (see below). 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t database : Once aggregated, cleaned, and reconciled, the data was converted to JSON (JavaScript Object Notation) and stored in a CouchDB database. CouchDB is a \"NoSQL\" document database that stores objects in JSON format. Unlike typical SQL databases, CouchDB does not have a database schema and does not support ad hoc queries. Instead CouchDB accepts semi-structured documents, and the user defines fixed queries or \"views\" (Anderson et al. 2010). phylogeny viewer : Screenshot of phylogeny from PhyLoTA as displayed in BioNames. The user can zoom in and out and pan, as well as change the layout of the tree. PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t Figure 11 Relative importance of different publishers of taxonomic literature Bubble chart showing relative numbers of taxonomic articles made available online by different publishers. PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t availability : BioNames is accessible at http://bionames.org. The source code used to build the web site is available on GitHub http://github.com/rdmpage/bionames. Scripts used to fetch, clean, and reconcile the data are archived in http://github.com/rdmpage/bionames-data eol challenge : In response to the Encyclopedia of Life (EOL) Computational Data Challenge (http://eol.org/info/323) I constructed BioNames (http://bionames.org) (Page 2012). Its goal is to create a database of taxonomic names linked to the primary literature and, wherever possible, to phylogenetic trees. Using existing globally unique identifiers for taxonomic names, concepts, publications, and sequences rather than cryptic text strings (for example, abbreviated bibliographic citations) simplifies the task of linking \u2014 we can rely on exact matching of identifiers rather than approximate matching between names for what may or may not be the same entity. This is particularly relevant once we start to aggregate information from different databases, where the same information (e.g., a publication) may be represented by different strings. Furthermore, if we use existing identifiers we increase the potential to connect to other databases (Page 2008a). This paper outlines how BioNames was built, describes the user interface, and discusses future plans. Materials & Methods 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t BioNames integrates data on taxonomic names and classifications, literature, and phylogenies from a variety of sources. Given the inevitable differences in how different databases treat the same data (as well as internal inconsistencies within individual databases), considerable effort must be spent cleaning and reconciling data. Much of this process involves mapping \"strings\" to \"things\" (Bollacker et al. 2008), or more precisely, mapping strings to identifiers for things. bibliographic identifiers : When populating BioNames every effort has been made to map each bibliographic string to a corresponding identifier, such as a Digital Object identifier (DOI). While DOIs are the bestknown bibliographic identifier, there are several others that are relevant to the taxonomic literature (Page 2009). DOIs are themselves based on Handles (http://hdl.handle.net), an identifier widely used by digital repositories such as DSpace (Smith et al. 2003). A number of journals, such as the Bulletins and Novitates of the American Museum of Natural History are available in DSpace repositories and consequently have Handles. Other major archives such as JSTOR (http://www.jstor.org/) and the Japanese National Institute of Informatics (CiNii; http://ci.nii.ac.jp/) have their own unique identifiers (typically integer numbers that are part of a URL). Having a variety of identifiers can complicate the task of finding existing identifiers for a particular publication. Whereas for some identifiers, such as DOIs and CiNii NAIDs (National Institute of Informatics Article IDs) there are OpenURL resolvers for this task (Van de Sompel & Beit-Arie 2001), for other identifiers there may be no obvious way to find the identifier other than by using a search engine. For the example in Figure 2, the citation string \"Description of a new species of Pinnotheres, and redescription of P. novaezelandiae (Brachyura: Pinnotheridae). New Zealand Journal of Zoology, 10(2) 1983: 151-162. 158 (Zoological Record Volume 120)\" corresponds to the article with the DOI 10.1080/03014223.1983.10423904. Once we have a DOI, we can then use services such as those provided by CrossRef (http://www.crossref.org) to retrieve author and publisher information for an article (see Fig. 11 below for one use of publisher information). 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t Identifiers also exist for aggregations of publications, such as journals. The historical practice of abbreviating journal titles in citations has led to a plethora of ways to refer to the same journal. For example, the BioStor database (http://biostor.org; Page 2011b) has accumulated more than ten variations on the name of the journal Bulletin of Zoological Nomenclature (such as \"Bull Zool Nomen\", \"Bull Zool Nom.\", \"Bull. Zool. Nomencl.\", etc.). This practice, presumably motivated by the desire to conserve space on the printed page, complicates efforts to match citations to identifiers. One approach to tackling this problem is to map abbreviations to journal-level globally unique identifiers, such as International Standard Serial Numbers (ISSNs) (for the Bulletin of Zoological Nomenclature the ISSN is 0007-5167). In addition to reducing ambiguity, there are web services such as that provided by WorldCat (http://www.worldcat.org) that take ISSNs and return the history of name changes for a journal, which in turn can help clarify the (often complicated) history of long-lived journals. documents : Taxonomic publications are available under a variety of licenses, ranging from explicitly open access licenses (MacCallum 2007) to articles that are \"free\", to articles that are behind a paywall. Archives such as JSTOR and CiNii have a mixture of free and subscription-based content. Many smaller journals, often published by scientific societies, are providing their content online for free, if not explicitly under an open license. The Biodiversity Heritage Library (the single largest source of taxonomic articles in BioNames, Fig. 11) makes its content available under a Creative Commons license. Where PDFs were available online either \"for free\" or under open access, these were downloaded and locally cached. Pages were extracted and converted into bitmap images for subsequent display in a web browser. 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t Closed-access publications that are available online are linked to by their identifier (e.g., DOI). Access to some of these publications may be available for short-term \"rent\" by services such as DeepDyve (http://www.deepdyve.com): where possible BioNames includes a link those services. clustering taxonomic names : Taxonomic names comprise a \"canonical\" name and the name's authorship, for example Homo sapiens Linnaeus comprises the canonical name \"Homo sapiens\" and the authorship string \"Linnaeus\". Names in taxonomic databases such as ION display numerous variations in spelling of authors, and instances of the same canonical name published by different authors (e.g., homonyms), so the names were clustered before populating BioNames. For each set of taxon names with the same canonical name the authorship was compared. If one name lacked an author and the other had an author, the names were automatically merged into a cluster. Given more than two names a graph was constructed where the nodes are the authorship strings, and a pair of nodes is connected if their corresponding strings were sufficiently similar. String similarity was computed by converting the strings to a \"finger print\" comprising lower case letters with all accented characters replaced by non-accented equivalents, and all punctuation removed, then finding the longest common subsequence of the two strings. If the length of the subsequence relative to the input strings was longer than a specified threshold (by default, 0.8, where identical strings have a similarity of 1.0) then the two author strings were connected by an edge in the graph. The components of the graph correspond to clusters of names with similar authorship strings, and were treated as being the same name. Figure 3 shows a graph for the different names that all have \"Rhacophorus\" as the canonical name. 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t Graph depicting similarity between different authorship strings associated with the name \"Rhacophorus\". The components of this graph correspond to the name clusters recognised by BioNames. PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t Figure 4 Matching taxonomic names to taxa Bipartite graph of string similarities between taxonomic names containing the string \"Nystactes\" in the ION and GBIF databases. Solid edges in the graph represent the maximum weighted bipartite matching, and define the mapping between ION and GBIF names. PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t Figure 5 mapping names to taxa : BioNames includes two taxonomic classifications, sourced from GBIF (http://uat.gbif.org/dataset/d7dddbf4-2cf0-4f39-9b2a-bb099caae36c) and NCBI (ftp://ftp.ncbi.nih.gov/pub/taxonomy), respectively. These provide the user with a way to navigate through taxonomic names, as well as view data associated with each classification (e.g., phylogenies). Ideally there would be a one-to-one mapping between a taxonomic name and a taxon, but complications often arise. In addition to the well-known problems of synonymy (more than one name for the same taxon) and homonymy (the same name used for different taxa), name and taxon databases may store slightly different representations of the same name. For example, ION has four records for the name \"Nystactes\" (each name is followed by its LSID): Nystactes urn:lsid:organismnames.com:name:2787598 Nystactes Bohlke urn:lsid:organismnames.com:name:2735131 Nystactes Gloger 1827 urn:lsid:organismnames.com:name:4888093 Nystactes Kaup 1829 urn:lsid:organismnames.com:name:4888094 GBIF has three taxa with this name (the number is the GBIF species id): Nystactes B\u00f6hlke, 1957 2403398 Nystactes Gloger, 1827 2475109 Nystactes Kaup, 1829 3239722 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t Note the differences in the name string (\"o\" versus \"\u00f6\" in \"B\u00f6hlke\", presence or absence of years and commas). To automate the mapping of names to concepts in cases like this I constructed a bipartite graph where the nodes are taxon names, divided into two sets based upon which database they came from (e.g., one set of names from ION, the other from GBIF). I then connect the nodes of the graph by edges whose weights are the similarity of the two strings computed using the longest common subsequence that the two strings share. For example, Figure 4 shows the graph for \"Nystactes\". Computing the maximum weighted bipartite matching of this graph creates a map between the two sets of names. Ideally GBIF should have only one entry for Nystactes because each animal name (with a few exceptions) must be unique. If a newer name has already been published before, then it should be replaced by a new name. In this case, Nystactes (B\u00f6hlke 1957) has since been replaced by Nystactichthys (B\u00f6hlke 1958), and Nystactes (Kaup 1829) by Paramyotis (Bianchi 1916). Unfortunately these changes have not yet percolated their way from the primary literature into the GBIF taxonomy. images : To help the user recognise the taxa being displayed images for as many taxa as possible were obtained using EOL's API, which provides access to both the images, and a mapping between GBIF and NCBI taxon concept identifiers and the corresponding record in EOL. phylogenies : Phylogenies were obtained from the PhyLoTA database (http://phylota.net) (Sanderson et al. 2008). This database contains eukaryote phylogenies constructed from automatically assembled clusters of nucleotide sequences (loosely corresponding to \"genes\"). A MySQL data dump was 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t downloaded (version 184, corresponding to the GenBank release of the same version number) and used to populate a local MySQL database. Metadata for the sequences in each phylogeny was obtained from the European Bioinformatics Institute (EBI; http://www.ebi.ac.uk), and used to populate the MySQL database with basic information such as taxon and locality information, as well as bibliographic details for the sources of the sequences. Phylogenies from PhyLOTA are rendered in an interactive viewer using the Scalable Vector Graphics (SVG) format. The user can zoom in and out, and change the drawing style. Terminal taxa with the same label have the same colour (Fig. 10). This makes it easier to recognise clusters of sequences from the same taxon (e.g., conspecific samples), as well as highlight possible errors (e.g., mislabelled or misidentified sequences). At present the colours are arbitrarily chosen, other schemes could be added in future (Lespinats and Fertil 2011). search : BioNames features a simple search interface that takes a scientific name and returns matching taxonomic names and concepts, together with any publications and phylogenies that contain the name. Figure 5 shows an example search result. 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t Screenshot of the search results for a query BioNames. The results include names that match the query, taxon concepts from GBIF and NCBI with thumbnail images from EOL, phylogenies containing members of the genus, and relevant taxonomic publications. PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t Figure 6 document display : BioNames uses the DocumentCloud (https://github.com/documentcloud/document-viewer) viewer to display both PDFs, and page images from digital archives such as BioStor and Gallica (http://gallica.bnf.fr/) (Fig. 6). journals : Much of the work in populating BioNames comprises mapping citation to string to bibliographic identifiers and, where possible, linking those citations to full text. For each journal that has a ISSN, BioNames has a corresponding web page that lists all the articles from that journal that are in the database, and provides a graphical summary of how many of those articles have been located online (Fig. 7). timeline : BioNames can display timelines of the numbers of taxonomic names published in higher taxonomic groups, inspired by Taxatoy (Sarkar et al. 2008) (Fig. 8). For a given node in the taxonomic hierarchy the children of that node are displayed as a treemap where the size of each cell is proportional to the log of the number of taxa in the subtree rooted on that child taxon. The number of names in that taxon published in each year is displayed as an interactive chart. Clicking on an individual year will list the corresponding publications for that year. 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t taxa : Each GBIF or NCBI taxon in BioNames has a corresponding web page that lists the associated taxonomic names, publications linked to those names, and other relevant data (e.g., Fig. 9). dashboard : The BioNames web site features a \"dashboard\" which displays various summaries of the data it contains. For example, Fig. 11 shows a bubble chart of the number of articles different publishers have made available online. \"Publisher\" in this context is broadly defined to include digital archives such as BioStor and JSTOR, repositories using DSpace, and commercial publishers such as Elsevier, Informa UK, Magnolia Press, Springer, and Wiley. 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t links : BioNames makes extensive use of identifiers to clean and link data, but the real value of identifiers becomes apparent when they are shared, that is, when different databases use the same identifiers for the same entities, instead of minting their own. Reusing identifiers can enable unexpected connections between databases. For example, the PubMed biomedical literature database has a record (PMID:948206) for the paper \"Monograph on \u2018Lithoglyphopsis\u2019 aperta, the snail host of Mekong River Schistosomiasis\" (Davis et al. 1976). The PubMed record contains the abstract for the paper, but not a link to where the user can obtain a digital version of the paper. However, this reference is in a volume that has been scanned by the Biodiversity Heritage Library, and the article has been extracted by BioStor (http://biostor.org/reference/102054). If PubMed was linked to BHL, users of PubMed could go straight to the content of the article. But this is just the start. The Davis et al. paper also mentions museum specimens in the collection of the Academy of Natural Sciences of Drexel University, Philadelphia. Metadata for these specimens has been aggregated by GBIF, and the BioStor page for this article displays those links (http://biostor.org/reference/102054). In an ideal world we should be able seamlessly to traverse the path PubMed \u2192 BioStor \u2192 GBIF. Likewise, we should be able to traverse the path in the other direction. At present, a user of GBIF simply sees metadata for these specimens and a 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t locality map. They are unaware that these specimens have been cited in a paper (Davis et al. 1976) which demonstrates that the snails host the Mekong River schistosome. This connection would be trivial to make if the reciprocal link was made: GBIF \u2192 BioStor. Furthermore, a link BioStor \u2192 PubMed would give us access to Medical Subject Headings (MeSH) for the schistosome paper. Hence we could imagine ultimately searching a database of museum specimens (GBIF) using queries from a controlled vocabulary of biomedical terms (MeSH). Making these connections requires not only that we have digital identifiers, but also that where ever possible we reuse existing identifiers. In practice forging these links can be hard work (Page 2011a), and many links may be missing from existing databases (Miller et al. 2009). However, if we restrict ourselves to project-specific identifiers then we stymie attempts to create a network of connected biodiversity data. text mining : Much of the value of a scientific publication lies dormant unless it is accessible to text mining, which requires access to full text. Where possible BioNames stores information on the publisher of each article (Fig. 11), which could then be used to prioritise discussions with publishers on gaining access to full text (Van Noorden 2012). Fortunately, the single largest \"publisher\" of content in BioNames is BioStor (Page 2011b), which contains scans and OCR text from the Biodiversity Heritage Library. BHL makes its content available under a Creative Commons license, and so can be readily mined. Indeed, the text has already been indexed by tools that can recognise taxonomic names (Akella et al. 2012). 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t dark taxa : One of the original motivations for constructing BioNames is the rise of \"dark taxa\" in genomics databases (Page 2011c). It is clear that some dark taxa do, in fact, have names. For example, consider the frog \"Gephyromantis aff. blanci MV-2005\" (NCBI taxonomy id 321743), which has a single DNA sequence AY848308 associated with it. This sequence was published as part of a 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t DNA barcoding study (Vences et al. 2005). If we enter the accession number AY848308 into Google we find two documents, one the supplementary table for (Vences et al. 2005), the other a subsequent paper (Vences and Riva 2007) that describes the frog with this sequence as a new species, Gephyromantis runewsweeki. This example is relatively straightforward, but it still required significant time to track down the species description. A key question facing attempts to find names for dark taxa is whether the methods available can be scaled to handle the magnitude of the problem. Alternatively, one could argue that newer technologies such as DNA barcoding make classical taxonomy less relevant, and perhaps the effort in digitising older literature and exposing the taxonomic names it contains is misplaced. A counter argument would be that the taxonomic literature potentially contains a wealth of information on ecology, morphology and behaviour, often for taxa in areas that have been subsequently altered by human activity. Given the rarity of many taxa (Lim et al. 2011), and the uneven taxonomic and geographic distribution of taxonomic expertise (May 1998; Gaston and May 1992), for many species the only significant data on their biology may reside in the legacy literature (possibly under a different name (Solow et al. 1995)). As this legacy becomes more accessible through projects such as BHL (and services that build upon that project; Page 2011a) there will be considerable opportunities to mine that literature for basic biological data (Thessen et al. 2012). publishing platform : Recently some taxonomic journals have begun to mark up taxonomic names and descriptions (Penev et al. 2010), which is a precursor to linking names and data together. But these developments leave open the problem of what these links will point to. If we have a database of 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t all taxonomic names and the associated literature (such as BioNames aims to be for zoological names), then such a database would provide an obvious destination for those links. Indeed, ultimately, we could envisage publishing new taxonomic publications within such a database, so that each new publication becomes simply another document within the database (Gerstein and Junker 2002). In the same way, we could use automated methods to extend the process of tagging names, specimens and literature cited to the legacy literature (Page 2010), so that the entire body of taxonomic knowledge becomes a single interwoven web of names, citations, publications, and data. taberlet, p., coissac, e., pompanon, f., brochmann, c., & willerslev, e. : (2012). Towards next-generation biodiversity assessment using DNA metabarcoding. Molecular Ecology, 21(8), 2045\u20132050. doi:10.1111/j.1365-294X.2012.05470.x Thessen, A. E., Cui, H., & Mozzherin, D. (2012). Applications of Natural Language Processing in Biodiversity Science. Advances in Bioinformatics, 2012, 1\u201317. doi:10.1155/2012/391574 Van de Sompel, H., & Beit-Arie, O. (2001). Open Linking in the Scholarly Information Environment Using the OpenURL Framework. D-Lib Magazine, 7(3). doi:10.1045/march2001-vandesompel Van Noorden, R. (2012). Trouble at the text mine. Nature, 483(7388), 134\u2013135. doi:10.1038/483134a Vences M, Riva IDL (2007) A new species of Gephyromantis from Ranomafana National Park, south-eastern Madagascar (Amphibia, Anura, Mantellidae). Spixiana 30(1): 135-143. Vences, M., Thomas, M., van der Meijden, A., Chiari, Y., & Vieites, D. R. (2005).Frontiers in Zoology, 2(1), 5. doi:10.1186/1742-9994-2-5 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t W\u00e4gele, H., Klussmann-Kolb, A., Kuhlmann, M., Haszprunar, G., Lindberg, D., Koch, A., & W\u00e4gele, J. W. (2011). The taxonomist - an endangered race. A practical proposal for its survival. Frontiers in Zoology, 8(1), 25. doi:10.1186/1742-9994-8-25 Werner, Y. L. (2006). The case of impact factor versus taxonomy: a proposal. Journal of Natural History, 40(21-22), 1285\u20131286. doi:10.1080/00222930600903660 Yan, K.-K., & Gerstein, M. (2011). The Spread of Scientific Information: Insights from the Web Usage Statistics in PLoS Article-Level Metrics. (A. Vespignani, Ed.)PLoS ONE, 6(5), e19917. doi:10.1371/journal.pone.0019917 498 499 500 501 502 503 504 505 PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t Figure 1 taxonomy data model : Simplified diagram of the relationships between the core entities that make up taxonomy, such as authors, publications, taxon names, and taxa. Relationships between entities are represented by lines, those in black are the focus of BioNames. PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t Figure 2 rdf for taxon name : The RDF retrieved by dereferencing the LSID urn:lsid:organismnames.com:name:371873, which identifies the taxonomic name Pinnotheres atrinicola. PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t Figure 3 displaying an article : Screenshot of BioNames displaying a document from BioStor (Conle and Hennemann 2002). The document viewer can display page images, thumbnails, and (where available) text. PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t Figure 7 Screenshot of the page in BioNames for the journal Proceedings of the Entomological Society of Washington (ISSN 0013-8797). The centre column lists the articles in a volume selected by the user using the index on the left. The right hand column displays basic data about the journal, and a graphical display of how many articles have been mapped to a globally unique identifier. PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t Figure 8 Timeline of taxonomic names for birds Screenshot of the distribution overtime of publications of new names for birds (Aves). The treemap on the left displays taxa below Aves in the taxonomic hierarchy, the chart on the right displays the number of publications in each year that publish a new bird name. The user has clicked on \"2012\", resulting in a list of the papers published in that year appearing below the timeline. PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t Figure 9 bibliography for a taxon : Screenshot of the bibliography tab on a taxon page in BioNames. This example shows the publications relevant to the bat genus Rousettus, including those for synonyms. The user can select publications from a given time slice and/or combination of synonyms. PeerJ reviewing PDF | (v2013:08:749:0:0:REVIEW 28 Aug 2013) R ev ie w in g M an us cr ip t Figure 10",
+    "url": "https://peerj.com/articles/191/reviews/",
+    "review_1": "William Jungers \u00b7 Oct 8, 2013 \u00b7 Academic Editor\nACCEPT\nThank you for your cooperation in the review process. I think the reviewer suggestions and your corresponding emendations have served to improve your study substantially. The diversity of North American archaic primates is fascinating.",
+    "review_2": "William Jungers \u00b7 Sep 12, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nIf theses are the best photographs the author can produce, then I think they are adequate if not \"perfect\" -- as the re-review notes.\n\nUnless you disagree, in your final version please make reference to \"the diagnostic trait for this genus.\"\nPlease try to clarify or modify text with respect to\n\"flanged\" and \"reduced\" (small, smaller, smallest?)\n\nPlease correct the statement about your characters since some lower dental features are part of the phylogenetic analysis. I urge you to reflect upon (and perhaps comment on) your decision to limit your characters to those observed only on Palaeocene species.",
+    "review_3": "Mary Silcox \u00b7 Sep 11, 2013\nBasic reporting\nThe author has submitted somewhat improved versions of the images of the teeth. I think it is up to the editor to determine whether these are adequate. It is now possible to see most of the relevant morphology, so my personal assessment is that they are borderline publishable. However, I would not chose to publish an image that has residue partially coating the occlusal surface of the tooth.\nExperimental design\nI'm experiencing some confusion about the phylogenetic analysis. In his rebuttal letter, the author states that \" The character matrix is composed of only upper dental traits\", but this is clearly not true (e.g. characters 19-22 all pertain to the lower i1). In any case, it would be a terrible idea to so restrict an analysis. Authors sometimes do this on the mistaken belief that only characters that can be coded for a fossil will affect that fossil's position in a tree. But the common practice of molecular backboning shows why this idea is fallacious. Even though no fossil can be coded for molecular characters, by influencing the relationships of all the OTHER taxa in an analysis, those data will end up influencing the position of the fossil in question. This is also true for morphological data, both with respect to character partitions AND taxon sampling, which is why I think it makes no sense to restrict the analysis to species alive in the Paleocene. However, the goals of this paper are limited in scope, and the results of the analysis are not unreasonable, so this could meet the standard of publishable on those bases.\n\nThe character list is generally improved over the preceding revision, although a few characters are still not clear:\n\nch. 1: adding \"flanged\" does not help, since this is also not standard anatomical terminology.\nch. 10: the author has once again used \"reduced\", which implies an evolutionary direction, and as such is inappropriate\nValidity of the findings\nI am convinced that the species is new and pertains to Zanycteris. I'm puzzled as to why the author continues not to make reference to the diagnostic trait for this genus (i.e., anteriorly protruding parastylar lobe on M1). Surely it wouldn't be difficult to add that in? Or does he disagree?\nCite this review as\nSilcox M (2013) Peer Review #1 of \"A new species of the archaic primate Zanycteris from the late Paleocene of western Colorado and the phylogenetic position of the family Picrodontidae (v0.2)\". PeerJ https://doi.org/10.7287/peerj.191v0.2/reviews/1",
+    "review_4": "William Jungers \u00b7 Jul 8, 2013 \u00b7 Academic Editor\nMAJOR REVISIONS\nTwo reviewers see real merit in your paper on these rare fossil primates, but both request much better photographs (especially since you are creating a new type specimen). Both reviewers also raise substantive issues with your phylogenetic analysis that need to be addressed (and which may require new analyses). Additional recommendations are mostly constructive, and attention to these suggestions will serve to improve the diagnosis and the discussion of its significance for primate paleobiology. Careful editing is needed to eliminate various typos and spelling errors.",
+    "pdf_1": "https://peerj.com/articles/191v0.3/submission",
+    "pdf_2": "https://peerj.com/articles/191v0.2/submission",
+    "review_5": "Gregg F Gunnell \u00b7 Jun 30, 2013\nBasic reporting\nThis paper describes a new species of Zanycteris from the Paleocene of Colorado....it is well written and well organized and I had only a few minor comments which I detail below. In general I find no problem with publishing this paper once the minor fixes are made.\nExperimental design\nNot applicable\nValidity of the findings\nThere seems no doubt that Z. honeyi is a new species of Zanycteris - I found the ecological analysis a little wanting - I agree that the teeth of Z. honeyi are of a frugivorous animal, similar to Ariteus (often mis-spelled in the mansucript as Artiteus) - but I'm not sure why thet \"restricts the stratographic range\" of picrodontids to the middle and late Paleocene only - surely there were plenty of fruits available in the Eocene as well....all of those real primates must have been eating something in the Eocene\nAdditional comments\nI find this paper to be more or less good as it is...there are few places where it needs a little editing or rewording....\n\nThe only major complaint I have is that Figure 1 showing the holotype specimen really is unacceptable for publication - it is blurry and out of focus - the buccal view is nearly impossible to interpret at all - in this day and age of microCT scanners and microscopic imaging with stacking software surely much better images than these can be provided. Also, it looks like there is still adhering matrix on the occlusal surfaces of the molars - this should be cleaned off beofre re-imaging.\n\nAlso I have a list of otrher comments:\n\nAbstract - Ariteus is mis-spelled in fourth line from bottom\n\nMethods and Material section - \"diverse fauna of fossil mammals\" is cited in line 39 and repeated again in line 42 citing Table 1 (I didn't see Table 1 anywhere) - no need to repeat this same information twice\n\nComparisons of Z. honeyi - lines 104-105 the authors notes differnces with Picrodus - he should cite the Scott and Fox, 2005 publication here\n\nDiscussion - line 146 - I think I'd be a little less dogmatic here and say that the \"molar teeth were probably specialized for cutting....\"\n\nLine 15 - Ariteus mis-spelled again\n\nLine 155 - paracone and metacone cusps (delete the \"s\" after paracone and metacone)\n\nLines 162-165 - I see no basis for saying this - yes paromomyids and picrodontids are different dentally but no actual analysis has been done to substantiate the claim that picrodontids were limited to the middle and late Paleocene by their dental specializations - very few plesiadapiforms made it past the end of the Paleocene in general - I think I'd just drop these lines.\n\nI'm not entirely sure why yet another phylogenetic analysis is being run here - the strict consensus tree differs quiet a lot from other, more rigorous trees - there are so many polytomy's that I'm not sure I believe any of what is being depicted....also the tooth images on the tree are so small as to be unrecognizable as teeth - probably should just eliminate them\n\nThat's all I have\nCite this review as\nGunnell GF (2013) Peer Review #1 of \"A new species of the archaic primate Zanycteris from the late Paleocene of western Colorado and the phylogenetic position of the family Picrodontidae (v0.1)\". PeerJ https://doi.org/10.7287/peerj.191v0.1/reviews/1",
+    "pdf_3": "https://peerj.com/articles/191v0.1/submission",
+    "review 6": "Mary Silcox \u00b7 Jun 17, 2013\nBasic reporting\nPicrodontids are incredibly rare primates, so this paper is certainly worthy of publication. However, there are a few significant issues with the paper that need to be addressed before it can be published. Most serious, the argument for the attribution to Zanycteris is not clearly made. As discussed by both Silcox and Gunnell (2008) and Scott and Fox (2005) the key diagnostic difference between this genus and Picrodus is an anteriorly protruding parastylar lobe in the former, and yet the author suggests that the new specimen belongs in Zanycteris on the basis of a \u201cshorten [sic] parastyle\u201d. Looking at the image, I think that the specimen is actually Zanycteris and does have an anteriorly protruding parastylar lobe, but this certainly isn\u2019t clear in the discussion. What\u2019s more, the image of the specimen is very poor, which makes it difficult to determine if the description as presented is correct. For example, I really can\u2019t tell if the M2 is complete or broken buccally\u2014this is critically important since the relative size of M1 vs. M2 is the key diagnostic trait for the new species. In the buccal view it is impossible to tell tooth from bone\u2014I actually think it would be okay to just cut this figure. For upper teeth, a buccal view isn\u2019t really needed. In the occlusal view it is clear that the specimen has been terribly over-whitened, so that there is a thick layer of whitening material coating the whole specimen, with beaded up material obscuring a good part of the morphology. My guess is that moisture either on the specimen, or in the bulb of the whitening tube, caused the powder to bead up. The image is also somewhat out of focus. Simply put, Figure 1 is not publishable as is. The specimen needs to be cleaned, and a much lighter layer of whitening used. Alternatively, an SEM or microCT image could be substituted. A few more minor details: Abstract, line 5: should be \u201cBY a single specimen\u201d\nAbstract line 6: should be \u201cshortenED parastyle\u201d\nAbstract line 9: should be \u201csimilarLY sized upper second molar\u201d\nAbstract line 12: should be cristid obliqua, not cristid oblique Line 20: \u201cmore recently studies\u2026\u201d requires a reference \u2013the key paper here is Szalay 1968\nLine 25: should be \u201cis strengthenED by the\u2026\u201d\nLine 28: \u201cThe enlarged incisor is considered a synapomorphy of Plesiadapiformes\u2026\u201d Clemens did suggest this, but Plesiadapiformes has never come out as a monophyletic group (in a holophyletic sense, i.e., excluding euprimates or Dermoptera) in any cladistic analysis. And it doesn\u2019t come out as demonstrably monophyletic in your own results\u2014Purgatorius is part of an unresolved polytomy with the rest of the plesiadapiforms and Plagiomenidae, Apatemyidae etc. For this reason, it is inappropriate to talk about this character as a synapomorphy, at least without a qualification.\nLine 31: refs are needed for the time span of picrodontids (i.e., after \u201cin North America\u201d)\nLine 69: It is standard to include several more layers in the systematic paleontology listing (e.g., at a minimum Order and Family)\nLine 94: It is unclear what \u201cthe broad protocone spreads into the postprotocrista\u201d means. Clarify.\nLines 163, 164: when used informally, taxonomic names should not be capitalized (e.g., should be paromomyids and picrodontids)\nFigure 2: \u201cElpidophorus\u201d and \u201cstonleyi\u201d are spelled incorrectly\nExperimental design\nWith respect to the phylogenetic analysis, it seems strange to me to limit the sample to Paleocene mammals. Because of the vagaries of fossil preservation, it is not necessarily the case that younger forms are irrelevant in assessing the ancestry of older forms. This is particularly an issue with respect to the discussion about the possible ancestors of different plesiadapiform groups (lines 133-135). Without including Premnoides douglassi, the putative ancestor of Paromomyidae (see Gunnell, 1989), it really isn\u2019t appropriate to discuss the possible descendants of the mentioned palaechthonids. I\u2019m not necessarily suggesting that the author needs to expand his sample, since the systematic questions he is asking here are fairly limited in scope. However, he should resist the temptation to over-interpret the results. There are also a number of issues of clarity with the character list that need to be addressed: Character 1: not clear what is meant by a \u201cwinged postcingulum\u201d\u2014this is not standard dental terminology. Please define and clarify. Character 4: how close to they need to be in size for it to be \u201cnearly\u201d? Define more precisely. Characters 5, 6, 12: These characters are essentially written as double negatives (i.e., the \u201cpresent\u201d state implies the absence of something), which is confusing, and non-standard. Re-write (e.g., \u201cPostprotocrista on M2 Present = 0, Absent = 1) Character 7: \u201creduced\u201d implies an evolutionary direction. Define based on observed morphology (e.g., short) Characters 8, 13, 14: It is inappropriate and imprecise to make reference to a particular genus in the definition of a character. Describe what\u2019s there. Character 10: It is a bad idea to combine two potentially independent elements of morphology in a single character (e.g., size and position in this case) since you have no way of knowing a priori if they always covary. Character 14: groove in what position? Character 112: Not clear what is meant by \u201csmall\u201d and \u201cenlarged\u201d here. Enlarged relative to what?\nValidity of the findings\nThere is an inaccuracy in how the results are reported. On lines 127-128 the author states that \u201cThe strict consensus tree shows Zanycteris honeyi as the sister-taxa [sic] to Zanycteris paleocenus\u2026\u201d No, it doesn\u2019t\u2014the two taxa form a polytomy with Picrodus. The nature of their relationships is unresolved on the tree as presented. This must be clarified. Also, \u201ctaxa\u201d is plural\u2014it would be the sister taxON.\nAdditional comments\nI as glad to see that you named this for Jim Honey. He was a very nice man--I was sorry to hear of his passing.\nCite this review as\nSilcox M (2013) Peer Review #2 of \"A new species of the archaic primate Zanycteris from the late Paleocene of western Colorado and the phylogenetic position of the family Picrodontidae (v0.1)\". PeerJ https://doi.org/10.7287/peerj.191v0.1/reviews/2",
+    "all_reviews": "Review 1: William Jungers \u00b7 Oct 8, 2013 \u00b7 Academic Editor\nACCEPT\nThank you for your cooperation in the review process. I think the reviewer suggestions and your corresponding emendations have served to improve your study substantially. The diversity of North American archaic primates is fascinating.\nReview 2: William Jungers \u00b7 Sep 12, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nIf theses are the best photographs the author can produce, then I think they are adequate if not \"perfect\" -- as the re-review notes.\n\nUnless you disagree, in your final version please make reference to \"the diagnostic trait for this genus.\"\nPlease try to clarify or modify text with respect to\n\"flanged\" and \"reduced\" (small, smaller, smallest?)\n\nPlease correct the statement about your characters since some lower dental features are part of the phylogenetic analysis. I urge you to reflect upon (and perhaps comment on) your decision to limit your characters to those observed only on Palaeocene species.\nReview 3: Mary Silcox \u00b7 Sep 11, 2013\nBasic reporting\nThe author has submitted somewhat improved versions of the images of the teeth. I think it is up to the editor to determine whether these are adequate. It is now possible to see most of the relevant morphology, so my personal assessment is that they are borderline publishable. However, I would not chose to publish an image that has residue partially coating the occlusal surface of the tooth.\nExperimental design\nI'm experiencing some confusion about the phylogenetic analysis. In his rebuttal letter, the author states that \" The character matrix is composed of only upper dental traits\", but this is clearly not true (e.g. characters 19-22 all pertain to the lower i1). In any case, it would be a terrible idea to so restrict an analysis. Authors sometimes do this on the mistaken belief that only characters that can be coded for a fossil will affect that fossil's position in a tree. But the common practice of molecular backboning shows why this idea is fallacious. Even though no fossil can be coded for molecular characters, by influencing the relationships of all the OTHER taxa in an analysis, those data will end up influencing the position of the fossil in question. This is also true for morphological data, both with respect to character partitions AND taxon sampling, which is why I think it makes no sense to restrict the analysis to species alive in the Paleocene. However, the goals of this paper are limited in scope, and the results of the analysis are not unreasonable, so this could meet the standard of publishable on those bases.\n\nThe character list is generally improved over the preceding revision, although a few characters are still not clear:\n\nch. 1: adding \"flanged\" does not help, since this is also not standard anatomical terminology.\nch. 10: the author has once again used \"reduced\", which implies an evolutionary direction, and as such is inappropriate\nValidity of the findings\nI am convinced that the species is new and pertains to Zanycteris. I'm puzzled as to why the author continues not to make reference to the diagnostic trait for this genus (i.e., anteriorly protruding parastylar lobe on M1). Surely it wouldn't be difficult to add that in? Or does he disagree?\nCite this review as\nSilcox M (2013) Peer Review #1 of \"A new species of the archaic primate Zanycteris from the late Paleocene of western Colorado and the phylogenetic position of the family Picrodontidae (v0.2)\". PeerJ https://doi.org/10.7287/peerj.191v0.2/reviews/1\nReview 4: William Jungers \u00b7 Jul 8, 2013 \u00b7 Academic Editor\nMAJOR REVISIONS\nTwo reviewers see real merit in your paper on these rare fossil primates, but both request much better photographs (especially since you are creating a new type specimen). Both reviewers also raise substantive issues with your phylogenetic analysis that need to be addressed (and which may require new analyses). Additional recommendations are mostly constructive, and attention to these suggestions will serve to improve the diagnosis and the discussion of its significance for primate paleobiology. Careful editing is needed to eliminate various typos and spelling errors.\nReview 5: Gregg F Gunnell \u00b7 Jun 30, 2013\nBasic reporting\nThis paper describes a new species of Zanycteris from the Paleocene of Colorado....it is well written and well organized and I had only a few minor comments which I detail below. In general I find no problem with publishing this paper once the minor fixes are made.\nExperimental design\nNot applicable\nValidity of the findings\nThere seems no doubt that Z. honeyi is a new species of Zanycteris - I found the ecological analysis a little wanting - I agree that the teeth of Z. honeyi are of a frugivorous animal, similar to Ariteus (often mis-spelled in the mansucript as Artiteus) - but I'm not sure why thet \"restricts the stratographic range\" of picrodontids to the middle and late Paleocene only - surely there were plenty of fruits available in the Eocene as well....all of those real primates must have been eating something in the Eocene\nAdditional comments\nI find this paper to be more or less good as it is...there are few places where it needs a little editing or rewording....\n\nThe only major complaint I have is that Figure 1 showing the holotype specimen really is unacceptable for publication - it is blurry and out of focus - the buccal view is nearly impossible to interpret at all - in this day and age of microCT scanners and microscopic imaging with stacking software surely much better images than these can be provided. Also, it looks like there is still adhering matrix on the occlusal surfaces of the molars - this should be cleaned off beofre re-imaging.\n\nAlso I have a list of otrher comments:\n\nAbstract - Ariteus is mis-spelled in fourth line from bottom\n\nMethods and Material section - \"diverse fauna of fossil mammals\" is cited in line 39 and repeated again in line 42 citing Table 1 (I didn't see Table 1 anywhere) - no need to repeat this same information twice\n\nComparisons of Z. honeyi - lines 104-105 the authors notes differnces with Picrodus - he should cite the Scott and Fox, 2005 publication here\n\nDiscussion - line 146 - I think I'd be a little less dogmatic here and say that the \"molar teeth were probably specialized for cutting....\"\n\nLine 15 - Ariteus mis-spelled again\n\nLine 155 - paracone and metacone cusps (delete the \"s\" after paracone and metacone)\n\nLines 162-165 - I see no basis for saying this - yes paromomyids and picrodontids are different dentally but no actual analysis has been done to substantiate the claim that picrodontids were limited to the middle and late Paleocene by their dental specializations - very few plesiadapiforms made it past the end of the Paleocene in general - I think I'd just drop these lines.\n\nI'm not entirely sure why yet another phylogenetic analysis is being run here - the strict consensus tree differs quiet a lot from other, more rigorous trees - there are so many polytomy's that I'm not sure I believe any of what is being depicted....also the tooth images on the tree are so small as to be unrecognizable as teeth - probably should just eliminate them\n\nThat's all I have\nCite this review as\nGunnell GF (2013) Peer Review #1 of \"A new species of the archaic primate Zanycteris from the late Paleocene of western Colorado and the phylogenetic position of the family Picrodontidae (v0.1)\". PeerJ https://doi.org/10.7287/peerj.191v0.1/reviews/1\nReview 6: Mary Silcox \u00b7 Jun 17, 2013\nBasic reporting\nPicrodontids are incredibly rare primates, so this paper is certainly worthy of publication. However, there are a few significant issues with the paper that need to be addressed before it can be published. Most serious, the argument for the attribution to Zanycteris is not clearly made. As discussed by both Silcox and Gunnell (2008) and Scott and Fox (2005) the key diagnostic difference between this genus and Picrodus is an anteriorly protruding parastylar lobe in the former, and yet the author suggests that the new specimen belongs in Zanycteris on the basis of a \u201cshorten [sic] parastyle\u201d. Looking at the image, I think that the specimen is actually Zanycteris and does have an anteriorly protruding parastylar lobe, but this certainly isn\u2019t clear in the discussion. What\u2019s more, the image of the specimen is very poor, which makes it difficult to determine if the description as presented is correct. For example, I really can\u2019t tell if the M2 is complete or broken buccally\u2014this is critically important since the relative size of M1 vs. M2 is the key diagnostic trait for the new species. In the buccal view it is impossible to tell tooth from bone\u2014I actually think it would be okay to just cut this figure. For upper teeth, a buccal view isn\u2019t really needed. In the occlusal view it is clear that the specimen has been terribly over-whitened, so that there is a thick layer of whitening material coating the whole specimen, with beaded up material obscuring a good part of the morphology. My guess is that moisture either on the specimen, or in the bulb of the whitening tube, caused the powder to bead up. The image is also somewhat out of focus. Simply put, Figure 1 is not publishable as is. The specimen needs to be cleaned, and a much lighter layer of whitening used. Alternatively, an SEM or microCT image could be substituted. A few more minor details: Abstract, line 5: should be \u201cBY a single specimen\u201d\nAbstract line 6: should be \u201cshortenED parastyle\u201d\nAbstract line 9: should be \u201csimilarLY sized upper second molar\u201d\nAbstract line 12: should be cristid obliqua, not cristid oblique Line 20: \u201cmore recently studies\u2026\u201d requires a reference \u2013the key paper here is Szalay 1968\nLine 25: should be \u201cis strengthenED by the\u2026\u201d\nLine 28: \u201cThe enlarged incisor is considered a synapomorphy of Plesiadapiformes\u2026\u201d Clemens did suggest this, but Plesiadapiformes has never come out as a monophyletic group (in a holophyletic sense, i.e., excluding euprimates or Dermoptera) in any cladistic analysis. And it doesn\u2019t come out as demonstrably monophyletic in your own results\u2014Purgatorius is part of an unresolved polytomy with the rest of the plesiadapiforms and Plagiomenidae, Apatemyidae etc. For this reason, it is inappropriate to talk about this character as a synapomorphy, at least without a qualification.\nLine 31: refs are needed for the time span of picrodontids (i.e., after \u201cin North America\u201d)\nLine 69: It is standard to include several more layers in the systematic paleontology listing (e.g., at a minimum Order and Family)\nLine 94: It is unclear what \u201cthe broad protocone spreads into the postprotocrista\u201d means. Clarify.\nLines 163, 164: when used informally, taxonomic names should not be capitalized (e.g., should be paromomyids and picrodontids)\nFigure 2: \u201cElpidophorus\u201d and \u201cstonleyi\u201d are spelled incorrectly\nExperimental design\nWith respect to the phylogenetic analysis, it seems strange to me to limit the sample to Paleocene mammals. Because of the vagaries of fossil preservation, it is not necessarily the case that younger forms are irrelevant in assessing the ancestry of older forms. This is particularly an issue with respect to the discussion about the possible ancestors of different plesiadapiform groups (lines 133-135). Without including Premnoides douglassi, the putative ancestor of Paromomyidae (see Gunnell, 1989), it really isn\u2019t appropriate to discuss the possible descendants of the mentioned palaechthonids. I\u2019m not necessarily suggesting that the author needs to expand his sample, since the systematic questions he is asking here are fairly limited in scope. However, he should resist the temptation to over-interpret the results. There are also a number of issues of clarity with the character list that need to be addressed: Character 1: not clear what is meant by a \u201cwinged postcingulum\u201d\u2014this is not standard dental terminology. Please define and clarify. Character 4: how close to they need to be in size for it to be \u201cnearly\u201d? Define more precisely. Characters 5, 6, 12: These characters are essentially written as double negatives (i.e., the \u201cpresent\u201d state implies the absence of something), which is confusing, and non-standard. Re-write (e.g., \u201cPostprotocrista on M2 Present = 0, Absent = 1) Character 7: \u201creduced\u201d implies an evolutionary direction. Define based on observed morphology (e.g., short) Characters 8, 13, 14: It is inappropriate and imprecise to make reference to a particular genus in the definition of a character. Describe what\u2019s there. Character 10: It is a bad idea to combine two potentially independent elements of morphology in a single character (e.g., size and position in this case) since you have no way of knowing a priori if they always covary. Character 14: groove in what position? Character 112: Not clear what is meant by \u201csmall\u201d and \u201cenlarged\u201d here. Enlarged relative to what?\nValidity of the findings\nThere is an inaccuracy in how the results are reported. On lines 127-128 the author states that \u201cThe strict consensus tree shows Zanycteris honeyi as the sister-taxa [sic] to Zanycteris paleocenus\u2026\u201d No, it doesn\u2019t\u2014the two taxa form a polytomy with Picrodus. The nature of their relationships is unresolved on the tree as presented. This must be clarified. Also, \u201ctaxa\u201d is plural\u2014it would be the sister taxON.\nAdditional comments\nI as glad to see that you named this for Jim Honey. He was a very nice man--I was sorry to hear of his passing.\nCite this review as\nSilcox M (2013) Peer Review #2 of \"A new species of the archaic primate Zanycteris from the late Paleocene of western Colorado and the phylogenetic position of the family Picrodontidae (v0.1)\". PeerJ https://doi.org/10.7287/peerj.191v0.1/reviews/2\n",
+    "LLM_extracted_review": "['**Review 1:** The diversity of North American archaic primates is fascinating.\\n\\n**Review 2:** The photographs are adequate but not perfect. Reference to \"the diagnostic trait for this genus\" should be included. Clarification is needed regarding the terms \"flanged\" and \"reduced.\" The statement about characters should be corrected, as some lower dental features are part of the phylogenetic analysis. The decision to limit characters to those observed only on Palaeocene species should be reflected upon.\\n\\n**Review 3:** The images of the teeth have improved but are borderline publishable. Residue on the occlusal surface of the tooth is unacceptable. There is confusion regarding the phylogenetic analysis, particularly about the character matrix. The character list has improved, but some terms are still unclear or inappropriate. The author should reference the diagnostic trait for Zanycteris.\\n\\n**Review 4:** Better photographs are needed, especially for a new type specimen. Substantive issues with the phylogenetic analysis need to be addressed. Attention to constructive suggestions will improve the diagnosis and discussion. Careful editing is required to eliminate typos and spelling errors.\\n\\n**Review 5:** The paper is well written and organized, with only a few minor comments. The ecological analysis is lacking, particularly regarding the stratigraphic range of picrodontids. Figure 1 is unacceptable for publication due to blurriness and residue on the molars. Several minor edits and clarifications are needed throughout the paper.\\n\\n**Review 6:** The paper is worthy of publication, but significant issues need to be addressed. The argument for attribution to Zanycteris is unclear, and the images are poor, making it difficult to assess the specimen. The phylogenetic analysis should not be limited to Paleocene mammals, and clarity is needed in the character list. There are inaccuracies in how results are reported, particularly regarding the relationships of taxa in the consensus tree.']"
+}

peerj_json_files/PeerJ_Json_142.json ADDED Viewed

	@@ -0,0 +1,23 @@

+{
+    "v1_Abstract": "A new species of an archaic primate (Pleisadapiformes) is described based on a maxilla containing the first and second upper molars from the Fort Union Formation, Atwell Gulch Member in northwestern Colorado. The preserved teeth show the unusual dental characteristics of members of the rare and poorly documented Picrodontidae family, including an elongated centrocrista and wide occlusal surface. The new species is placed within the genus Zanycteris (represented by a single specimen from southern Colorado). This placement is based on similarities in regard to the parastyle, curvilinear centrocrista, and wider anterior stylar shelf on the upper molars. However, the new species differs from the only known species of Zanycteris in exhibiting an upper first molar that is 30% larger in area, while retaining a similarly sized upper second molar. Phylogenetic analysis supports the separation of the Picrodontidae family from the Paromomyidae, while still recognizing picrodontids position within Pleisadapiformes. The unusual dental features of the upper molars likely functioned in life as an enhanced shearing surface between the centrocrista and cristid obliqua crests for a specialized diet of fruit. A similar arrangement is found in the living bat Ariteus (Jamaican fig-eating bat), which feeds on fleshy fruit. The new species showcases the rapid diversification of archaic primates shortly after the extinction of the dinosaurs during the Paleocene, and the unusual dental anatomy of picrodontids to exploit new dietary specializations.",
+    "v1_col_introduction": "introduction  : The family Picrodontidae consists of rare fossil mammals known only in the late Paleocene (Torrejonian and Tiffanian North American Land Mammal Ages (NALMA)) of North America. Upon the initial discovery of the Picrodontidae Zanycteris in 1917 paleontologists placed the fossil within the Order Chiroptera (Matthew, 1917; Simpson, 1935; Simpson, 1937). Indeed, there is a close resemblance between Zanycteris and some of the fruit-eating bats of the New World, such as the living genus Ariteus (Jamaican fig-eating bat). This morphological similarity is exhibited in the upper first molar which is broadly shaped and greatly expanded. The expansion of the occlusal surface of the upper first molar is likely a reflection of similar diet, rather than of any similar phylogenetic relationship to fruit eating bats, since such specializations are absent in early fossil bat lineages (Simmons and Geisler, 1998). More recently studies have positioned the enigmatic Picrodontidae as aberrant members of archaic primates (Szalay, 1968). Researchers have viewed picrodontids as stemming from a Purgatorius-like ancestor (Tomida, 1982), a Palaechthon-like ancestor (Szalay, 1968) or more derived members of the Paromomyoidea (Szalay, 1968; Silcox, 2001; Silcox & Gunnell, 2008).\nThe relationship to archaic primates is strengthened by the presence of an enlarged incisor in a lower jaw of the picrodontid Picrodus recovered from Swain Quarry of the Fort Union Formation, middle Paleocene (Torrejonian) in Carbon County, Wyoming (Szalay, 1968; Williams, 1985). The enlarged incisor is often considered a synapomorphy of Plesiadapiformes, and is also found in the early Paleocene genus Purgatorius (Clemens, 2004), although this trait is also found in groups outside of Plesiadapiformes, such as the Apatemyidae (West, 1973).\nPicrodontids remain very elusive fossil mammals, with only a handful of specimens known from a narrow span of time during the middle to late Paleocene in North America (Silcox & Gunnell, 2008); in fact this is only the second known specimen of the rare genus Zanycteris.\n10\n11\n12\n13\n14\n15\n16\n17\n18\n19\n20\n21\n22\n23\n24\n25\n26\n27\n28\n29\n30\n31\n32\n33\nPeerJ reviewing PDF | (v2013:06:580:2:0:NEW 7 Oct 2013)\nR ev ie w in g M an\nus cr ip t\nThis paper reports on the occurrence of a new species of Zanycteris discovered in the late Paleocene (Tiffanian) deposits of the Fort Union Formation, Atwell Gulch Member in northwestern Colorado, and also discusses the phylogenetic relationship of picrodontids among various groups of archaic primates living in North America during the Paleocene.",
+    "v2_Abstract": "A new species of an archaic primate (Pleisadapiformes) is described based on a maxilla containing the first and second upper molars from the Fort Union Formation, Atwell Gulch Member in northwestern Colorado. The preserved teeth show the unusual dental characteristics of members of the rare and poorly documented Picrodontidae family, including an elongated centrocrista and wide occlusal surface. The new species is placed within the genus Zanycteris (represented by a single specimen from southern Colorado). This placement is based on similarities in regard to the parastyle, curvilinear centrocrista, and wider anterior stylar shelf on the upper molars. However, the new species differs from the only known species of Zanycteris in exhibiting an upper first molar that is 30% larger in area, while retaining a similarly sized upper second molar. Phylogenetic analysis supports the separation of the Picrodontidae family from the Paromomyidae, while still recognizing picrodontids position within Pleisadapiformes. The unusual dental features of the upper molars likely functioned in life as an enhanced shearing surface between the centrocrista and cristid obliqua crests for a specialized diet of fruit. A similar arrangement is found in the living bat Ariteus (Jamaican fig-eating bat), which feeds on fleshy fruit. The new species showcases the rapid diversification of archaic primates shortly after the extinction of the dinosaurs during the Paleocene, and the unusual dental anatomy of picrodontids to exploit new dietary specializations.",
+    "v2_col_introduction": "introduction  : The family Picrodontidae consists of rare fossil mammals known only in the late Paleocene (Torrejonian and Tiffanian North American Land Mammal Ages (NALMA)) of North America. Upon the initial discovery of the Picrodontidae Zanycteris in 1917 paleontologists placed the fossil within the Order Chiroptera (Matthew, 1917; Simpson, 1935; Simpson, 1937). Indeed, there is a close resemblance between Zanycteris and some of the fruit-eating bats of the New World, such as the living genus Ariteus (Jamaican fig-eating bat). This morphological similarity is exhibited in the upper first molar which is broadly shaped and greatly expanded. The expansion of the occlusal surface of the upper first molar is likely a reflection of similar diet, rather than of any similar phylogenetic relationship to fruit eating bats, since such specializations are absent in early fossil bat lineages (Simmons and Geisler, 1998). More recently studies have positioned the enigmatic Picrodontidae as aberrant members of archaic primates (Szalay, 1968). Researchers have viewed picrodontids as stemming from a Purgatorius-like ancestor (Tomida, 1982), a Palaechthon-like ancestor (Szalay, 1968) or more derived members of the Paromomyoidea (Szalay, 1968; Silcox, 2001; Silcox & Gunnell, 2008).\nThe relationship to archaic primates is strengthened by the presence of an enlarged incisor in a lower jaw of the picrodontid Picrodus recovered from Swain Quarry of the Fort Union Formation, middle Paleocene (Torrejonian) in Carbon County, Wyoming (Szalay, 1968; Williams, 1985). The enlarged incisor is often considered a synapomorphy of Plesiadapiformes, and is also found in the early Paleocene genus Purgatorius (Clemens, 2004), although this trait is also found in groups outside of Plesiadapiformes, such as the Apatemyidae (West, 1973).\nPicrodontids remain very elusive fossil mammals, with only a handful of specimens known from a narrow span of time during the middle to late Paleocene in North America (Silcox & Gunnell, 2008); in fact this is only the second known specimen of the rare genus Zanycteris.\n10\n11\n12\n13\n14\n15\n16\n17\n18\n19\n20\n21\n22\n23\n24\n25\n26\n27\n28\n29\n30\n31\n32\n33\nPeerJ reviewing PDF | (v2013:06:580:1:0:NEW 30 Aug 2013)\nR ev ie w in g M an\nus cr ip t\nThis paper reports on the occurrence of a new species of Zanycteris discovered in the late Paleocene (Tiffanian) deposits of the Fort Union Formation, Atwell Gulch Member in northwestern Colorado, and also discusses the phylogenetic relationship of picrodontids among various groups of archaic primates living in North America during the Paleocene.",
+    "v3_Abstract": "A new species of an archaic primate (Pleisadapiformes) is described based on a maxilla containing the first and second upper molars from the Fort Union Formation, Atwell Gulch Member in northwestern Colorado. The preserved teeth show the unusual dental characteristics of members of the rare and poorly documented Picrodontidae family, including an elongated centrocrista and wide occlusal surface. The new species is placed within the genus Zanycteris (represented from a single specimen from southern Colorado). This placement is based on similarities in regard to a shorten parastyle, curvilinear centrocrista, and wider anterior stylar shelf on the upper molars. However, the new species differs from the only known species of Zanycteris in exhibiting an upper first molar that is 30% larger in area, while retaining a similar sized upper second molar. Phylogenetic analysis supports the separation of the Picrodontidae family from the Paromomyidae, while still recognizing picrodontids position within Pleisadapiformes. The unusual dental features of the upper molars likely functioned in life as an enhanced shearing surface between the centrocrista and cristid oblique crests for a specialized diet of fruit. A similar arrangement is found in the living bat Artiteus (Jamaican fig-eating bat), which feeds on fleshy fruit. The new species showcases the rapid diversification of archaic primates shortly after the extinction of the dinosaurs during the Paleocene, and the unusual dental anatomy of picrodontids to exploit new dietary specializations.",
+    "v3_col_introduction": "introduction  : The family Picrodontidae consists of rare fossil mammals known only in the late Paleocene (Torrejonian and Tiffanian North American Land Mammal Ages (NALMA)) of North America. Upon the initial discovery of the Picrodontidae Zanycteris in 1917 paleontologists placed the fossil within the Order Chiroptera (Matthew, 1917; Simpson, 1935; Simpson, 1937). Indeed, there is a close resemblance between Zanycteris and some of the fruit-eating bats of the New World, such as the living genus Ariteus (Jamaican fig-eating bat). This morphological similarity is exhibited in the upper first molar which is broadly shaped and greatly expanded. The expansion of the occlusal surface of the upper first molar is likely a reflection of similar diet, rather than of any similar phylogenetic relationship to fruit eating bats, since such specializations are absent in early fossil bat lineages (Simmons and Geisler, 1998). More recently studies have positioned the enigmatic Picrodontidae as aberrant members of archaic primates. Researchers have viewed picrodontids as stemming from a Purgatorius-like ancestor (Tomida, 1982), a Palaechthon-like ancestor (Szalay, 1968) or more derived members of the Paromomyoidea (Szalay, 1968; Silcox, 2001; Silcox & Gunnell, 2008).\nThe relationship to archaic primates is strengthen by the presence of an enlarged incisor in a lower jaw of the picrodontid Picrodus recovered from Swain Quarry of the Fort Union Formation, middle Paleocene (Torrejonian) in Carbon County, Wyoming (Szalay, 1968; Williams, 1985). The enlarged incisor is considered a synapomorphy of Plesiadapiformes, and is also found in the early Paleocene genus Purgatorius (Clemens, 2004).\n10\n11\n12\n13\n14\n15\n16\n17\n18\n19\n20\n21\n22\n23\n24\n25\n26\n27\n28\n29\nPeerJ reviewing PDF | (v2013:06:580:0:0:NEW 12 Jun 2013)\nR ev ie w in g M an\nus cr ip t\nPicrodontids remain very elusive fossil mammals, with only a handful of specimens known from a narrow span of time during the middle to late Paleocene in North America; in fact this is only the second known specimen of the rare genus Zanycteris.\nThis paper reports on the occurrence of a new species of Zanycteris discovered in the late Paleocene (Tiffanian) deposits of the Fort Union Formation, Atwell Gulch Member in northwestern Colorado, and also discusses the phylogenetic relationship of picrodontids among various groups of archaic primates living in North America during the Paleocene.",
+    "v1_text": "methods & materials : The fossil reported in this paper was recovered during geological mapping of the Citadel Plateau Quadrangle during the 1970s. The site has also produced a diverse fauna of mammals from the Fort Union Formation in western Colorado\u2019s Piceance Creek Basin (Burger & Honey, 2008). The site is referred to as University of Colorado Museum (UCM) Locality number 92177. Continued collection at the site over the years has produced additional fossil mammals (Table 1). However, the new species is represented by a single recovered specimen, despite 30 years of sporadic collection at the site. In the Piceance Creek Basin the Fort Union Formation is synonymous with the Atwell Gulch Member, which has been included as a member of the Wasatch Formation (Donnell, 1969) or DeBeque Formation (Kihm, 1984). In this paper I refer the Atwell Gulch Member, as the sole member of the Fort Union Formation, as it has been mapped elsewhere within the Piceance Creek Basin in western Colorado (Hail & Pipiringos, 1990; Hail & Smith, 1994). The Fort Union Formation (Atwell Gulch Mbr.) varies in thickness from 350 meters in the northeast to 196 meters in the south, and is divided into upper and lower informal units (Hail & Pipiringos, 1990). The fossil site UCM 92177 is 261 meters below the upper contact of the Late Paleocene Fort Union Formation (Atwell Gulch Member), with the Early Eocene Wasatch Formation (Molina Member). The lower unit of the Fort Union Formation consists of light-grey to light-brown sandstones; olive, purple, dark-reddish-brown claystone; and mudstones that are 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 PeerJ reviewing PDF | (v2013:06:580:2:0:NEW 7 Oct 2013) R ev ie w in g M an us cr ip t highly variegated. Large ribbon and sheet sandstone bodies are common in the north of the basin, where they can form massive sandstone cliffs measuring upwards of 25 meters thick, although most are 5-10 meters thick. Pebbles are exclusively composed of sedimentary rocks, including claystone and mudstones. The upper unit of the Fort Union Formation (Atwell Gulch Mbr.) is composed of carbonaceous shales; thin coals; and thin, but persistent highly calcareous sandstones. Mudstones and claystones are less common in the upper unit. Selenite is common, especially in the carbonaceous shales and coals. Fossil mammals are abundant in the lower unit of the Fort Union Formation, especially in the variegated beds found in association with UCM locality 92177. Fossil invertebrates, such as the bivalve Unio and a variety of gastropods are common in the upper unit, indicating a progression over time toward a more lacustrine environment. Large accumulations of gastropods are common in the upper unit, which have been interpreted as lakeshore accumulations (Hanley, 1974). Order PRIMATES Linnaeus, 1758 Family PICRODONITDAE Simpson, 1937 Genus ZANYCTERIS Matthew, 1917 ZANYCTERIS HONEYI new species Holotype\u2014 UCM 87378 right maxilla with upper first and second molars. Etymology\u2014 Posthumously named in honor of James G. Honey for his discovery of the holotype, type locality and for his kindness in allowing me to study this collection. 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 PeerJ reviewing PDF | (v2013:06:580:2:0:NEW 7 Oct 2013) R ev ie w in g M an us cr ip t Horizon\u2014 Fort Union Formation, Atwell Gulch Member, 261 meters below the top contact with the Wasatch Formation. Localities\u2014 Only known from UMC locality number 92177. Diagnosis\u2014 Z. honeyi exhibits an anteriorly protruding parastylar lobe on M1/. Differs from Z. paleocenus by having a 30% large M1/ area, while retaining a similar sized M2/ area to Z. paleocenus. Differs from Picrodus in lacking an extended parastyle on the M2/, having a better developed anterior stylar shelf, and postprotocrista on the M1/. Furthermore, the M1/ centrocrista is more curvilinear than Picrodus. Unlike Draconodus, crenulations occur in the trigon basin of M1/. Description\u2014 The holotype (UCM 87378) is the only specimen known from the Piceance Creek Basin. However, this specimen preserves morphology to indicate that it differs from Z. paleocenus from southwestern Colorado. The enlarged metastyle on the M1/ projects buccally from the posterior edge, and the stylar shelf bulges from the mid-point of the tooth. The paracingulum (the shelf formed by the paraconule) extends anteriorly, yet the stylar shelf encircling the paracone is poorly formed and resembles Picrodus rather than Zanycteris in shape. The strong postprotocrista of UCM 87378 extends directly toward the vestigial metaconule. This direct course of the postprotocrista results in a more enclosed trigon basin that closely resembles Z. paleocenus. The M2/ exhibits a wide stylar shelf, featuring a large parastyle. However, UCM 87378 typifies Zanycteris in lacking the greatly extended parastyle on the M2/ that is found in specimens of Picrodus. Despite wear, the molar paracone and metacone form a W-shaped crest across the midline on the M2/, while the broad protocone expands along the postprotocrista, a typical plesiadapiform trait. The M1/ measures 2.54 mm in length and 2.46 mm in width. The M2/ measures 1.26 mm in length and 1.60 in width. 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 PeerJ reviewing PDF | (v2013:06:580:2:0:NEW 7 Oct 2013) R ev ie w in g M an us cr ip t Comparison\u2014 Previous measurements of the holotype of Z. paleocenus (Simpson, 1935, Szalay, 1968) report a length of 2.05-2.20mm in length and 1.87-2.00 mm in width for the M1/, indicating that the new species (UCM 87378) is both larger and broader. However, reported measurements of the M2/ (1.25-1.30mm in length and 1.60-1.70mm in width) are similar in dimensions to UCM 87378. This indicates that while the first molar is enlarged, the second molar is of equal size between the two species of Zanycteris. This enlargement of the first molar likely served a functional role in providing a larger surface area for slicing between the centrocrista and cristid obliqua (Szalay, 1968). In some ways, UCM 87378 resembles Picrodus, such as the reduced anterior stylar shelf on the M1/ (Scott and Fox, 2005). However, other features more closely resemble Z. paleocenus, including the arrangement of the postprotocrista on the M1/ and smaller parastyle on the M2/. These features support inclusion of UCM 87378 within the genus Zanycteris rather than Picrodus. UCM 87378 and the holotype of Z. paleocenus (AMNH 17180) are the only two specimens of Zanycteris currently documented (Simpson, 1937; Scott and Fox, 2005). Zanycteris appears to be restricted to Colorado, while Picrodus has been found in Wyoming, Montana, Alberta and recently New Mexico (Simpson, 1937; Williams, 1985; Silcox & Gunnell, 2008; Scott & Fox, 2005; Silcox & Williamson, 2012). a new species of an archaic primate (pleisadapiformes) is described based on a maxilla : containing the first and second upper molars from the Fort Union Formation, Atwell Gulch Member in northwestern Colorado. The preserved teeth show the unusual dental characteristics of members of the rare and poorly documented Picrodontidae family, including an elongated centrocrista and wide occlusal surface. The new species is placed within the genus Zanycteris (represented by a single specimen from southern Colorado). This placement is based on similarities in regard to the parastyle, curvilinear centrocrista, and wider anterior stylar shelf on the upper molars. However, the new species differs from the only known species of Zanycteris in exhibiting an upper first molar that is 30% larger in area, while retaining a similarly sized upper second molar. Phylogenetic analysis supports the separation of the Picrodontidae family from the Paromomyidae, while still recognizing picrodontids position within Pleisadapiformes. The unusual dental features of the upper molars likely functioned in life as an enhanced shearing surface between the centrocrista and cristid obliqua crests for a specialized diet of fruit. A similar arrangement is found in the living bat Ariteus (Jamaican fig-eating bat), which feeds on fleshy fruit. The new species showcases the rapid diversification of archaic primates shortly after the extinction of the dinosaurs during the Paleocene, and the unusual dental anatomy of picrodontids to exploit new dietary specializations. PeerJ reviewing PDF | (v2013:06:580:2:0:NEW 7 Oct 2013) R ev ie w in g M an us cr ip t Benjamin John Burger* *Corresponding author: Department of Geology Utah State University Uintah Basin Regional Campus 320 North Aggie Blvd. Vernal, UT 84078 U.S.A. Phone 1-435-722-1778 benjamin.burger@usu.edu 1 2 3 4 5 6 7 8 9 PeerJ reviewing PDF | (v2013:06:580:2:0:NEW 7 Oct 2013) R ev ie w in g M an us cr ip t conclusions : In summary, the new species Zanycteris honeyi typifies the unique characteristics that set apart the Picrodontidae from other archaic primates known from the Paleocene of North America. Phylogenetic analysis supports the separation of the Picrodontidae family from the Paromomyidae family, while still recognizing their position within Pleisadapiformes. Further fossil discoveries, particularly cranial and postcranial remains will likely enable more confident placement of this unusual group of archaic primates among the evolutional tree during this pivotal time of primate diversification shortly after the extinction of the dinosaurs. acknowledgments : Special thanks to Jaelyn Eberle at the University of Colorado at Boulder for her support in my graduate studies which included this research. I would like to thank my committee Herbert H. Covert, Mary Kraus, Matthew Pranter, (University of Colorado) and Henry Fricke (Colorado College); Toni Culver, the collection manager at the University of Colorado Museum, as well as various members of the field crew including Alex Dutchak, Karen Lloyd, Pat Monaco, Andrea R. Bair, Ian J. Sweeney, Lou Taylor and Lea Ann Jolley. Fieldwork was conducted under BLM permit #C-60170 issued to the University of Colorado Museum. phylogenetic analysis : The acquisition of the highly specialized dentition found within Zanycteris and other members of the Picrodontidae remains a mystery. For example, how quickly did the specialized dentition evolve during the Paleocene? Among the known Plesiadapiformes, which one is most closely related to the family Picrodontidae and could possibly represent the ancestral condition for the specialized dentation exhibited by Zanycteris? To evaluate these questions and to work toward reconstructing the evolution of the specialized upper dentition of Picrodontidae, a phylogenetic analysis was undertaken using the 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 PeerJ reviewing PDF | (v2013:06:580:2:0:NEW 7 Oct 2013) R ev ie w in g M an us cr ip t morphological characteristics of the dentition of known Plesiadapiformes and outgroups (Paradectes, Cimolestes, and Leptacodon), which lived during the Paleocene in North America. The character matrix consisted of 113 dental characters, 97 of which were adopted from Silcox (2001). The analysis included 58 fossil taxa of contemporary North American Paleocene primates. A heuristic search using Mesquite version 2.75 (Maddison & Maddison, 2011) produced 6,579 most parsimonious trees (597 steps, consistency index [CI] = 0.36, retention index [RI] = 0.74). The strict consensus tree shows Zanycteris honeyi as closely related to Zanycteris paleocenus within a monophyletic clade of Picrodontidae (Picrodus, Draconodus, and Zanycteris). The family Picrodontidae was found to be within a clade consisting of Plesidapidae and Carpolestidae, rather than a placement within Paromomyoidea (Silcox & Gunnell, 2008). This phylogenetic position postulates that the expansion of the occlusal surface seen in the upper molars of both paromomyids and picrodontids is convergent, having evolved independently during the Paleocene. Possible ancestors of picrodontids are the early Paleocene taxa Plesiolestes, Torrejonia, Phoxomylus, and Talpohenach, while paromomyids appear to have arisen from the early Paleocene with a Palaechthon or Anasazia or the poorly known Premnoides like ancestor. Although there is ambiguity concerning the phylogenetic position of these primitive taxa, there is strong support for monophyletic clades of Paromomyidae, Picrodontidae, Plesiadapidae, and Carpolestidae, as well as a monophyletic clade of Plesiadapiformes, with the addition of Micromomyidae and Microsyopidae. Overall the resulting phylogenetic tree supports a position of Zanycteris honeyi within the Picrodontidae family, and that this new species is closely related to other archaic primates from the Paleocene of North America. Discussion 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 PeerJ reviewing PDF | (v2013:06:580:2:0:NEW 7 Oct 2013) R ev ie w in g M an us cr ip t The members of the Picrodontidae are exceptional in the development of a dentition that maximizes the shear forces along the long contact between the centrocrista on the upper molar (composed of a tall crest between the paracone and metacone) and the cristid obliqua that spans much of the length of the elongated talonid basin of the lower molar (Szalay, 1968). The molar teeth were thus probably specialized for cutting through hard outer husks of fruits and nuts. This was accomplished by positioning the fruit along the outer (buccal) shearing surface, which was greatly expanded (anteriorly and posteriorly) to maximize the amount of contact, much like a pair of long sharp scissors (Shaw, 1917). A similar expansion of the centrocrista is found in the upper molars of the Jamaican fig-eating bat (Ariteus), which feeds on the native Jamaican naseberry also known as sapodilla (Manilkara zapota), a fruit with a fleshy but firm texture (Sherwin & Gannon, 2005). Thus in both Ariteus and Zanycteris the major shearing surface is between the crests of the centrocrista above and the cristid obliqua below, demonstrating a similar specialized diet on fruit. This arrangement differs substantially from paromomyids, which retain distinct paracone and metacone cusps on the upper molars, with no development of a long and tall centrocrista between the two cusps. Rather, paromomyids expand the upper molars by broadening the postcingulum to form a wide talon basin on the lingual edge of the tooth. This broadening of the tooth functioned to expand the surface area particularly for holding food during mastication (Shaw, 1917). Thus paromomyids, such as Phenacolemur, broaden the upper molars to allow increased surface area for a larger and a more varied diet, while picrodontids, such as Zanycteris, expanded the upper molars to increase the shearing surface for a more specialized diet of a particular style of fleshy fruit. Further research on dental specialization within these groups might reveal why picrodontids have a limited stratigraphic range (Torrejonian to Tiffanian), when compared to the closely related, but dentally distinct paromomyids which ranged from the early Paleocene (Puercan) until the late Eocene (Duchesnean) (Silcox & Gunnell, 2008). 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 PeerJ reviewing PDF | (v2013:06:580:2:0:NEW 7 Oct 2013) R ev ie w in g M an us cr ip t image of fossil specimen. : Figure 1: Buccal and occlusal views of the holotype specimen, a maxilla containing the upper first and second molars. UCM 87378. PeerJ reviewing PDF | (v2013:06:580:2:0:NEW 7 Oct 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:580:2:0:NEW 7 Oct 2013) R ev ie w in g M an us cr ip t Figure 2 Strict consensus tree of the most parsimonious trees generated from the phylogenetic analysis. Figure 2: Strict consensus tree of the most parsimonious trees generated from the phylogenetic analysis of 113 upper dental characters scored against the 58 North American archaic primates known from the Paleocene. Temporal ranges during the Paleocene are shown for each species by blackened line. PeerJ reviewing PDF | (v2013:06:580:2:0:NEW 7 Oct 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:580:2:0:NEW 7 Oct 2013) R ev ie w in g M an us cr ip t Table 1(on next page) Mammal faunal list for fossil locality 92177. Table 1. Mammal faunal list for UCM locality 92177. PeerJ reviewing PDF | (v2013:06:580:2:0:NEW 7 Oct 2013) R ev ie w in g M an us cr ip t Table 1. Mammal Faunal List for UCM locality 92177. Mammalia Allotheria Multituberculata Ptilodus kummae Ectypodus musculus Theria Erinaceomorpha (\u201cApheliscidae\u201d) Haplaletes serior Litomylus ishami Phenacodaptes sabulosus Haplomylus simpsoni Lipotyphla Leptacodon tener Mesonychia cf. Sinonyx sp. Procreodi Thryptacodon australis Arctocyonides mumak Carnivoramorpha Protictis proteus Protictis cf. schaffi Primates Nannodectes gazini Plesiadapis fodinatus Chiromyoides gigas Zanycteris honeyi new species Ignacius frugivorus Ignacius sp. Carpodaptes cygneus Condylarthra Ectocion medituber Phenacodus grangeri Phenacodus magnus Pholidota Propalaeanodon sp. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 PeerJ reviewing PDF | (v2013:06:580:2:0:NEW 7 Oct 2013) R ev ie w in g M an us cr ip t",
+    "v2_text": "a new species of an archaic primate (pleisadapiformes) is described based on a maxilla : containing the first and second upper molars from the Fort Union Formation, Atwell Gulch Member in northwestern Colorado. The preserved teeth show the unusual dental characteristics of members of the rare and poorly documented Picrodontidae family, including an elongated centrocrista and wide occlusal surface. The new species is placed within the genus Zanycteris (represented by a single specimen from southern Colorado). This placement is based on similarities in regard to the parastyle, curvilinear centrocrista, and wider anterior stylar shelf on the upper molars. However, the new species differs from the only known species of Zanycteris in exhibiting an upper first molar that is 30% larger in area, while retaining a similarly sized upper second molar. Phylogenetic analysis supports the separation of the Picrodontidae family from the Paromomyidae, while still recognizing picrodontids position within Pleisadapiformes. The unusual dental features of the upper molars likely functioned in life as an enhanced shearing surface between the centrocrista and cristid obliqua crests for a specialized diet of fruit. A similar arrangement is found in the living bat Ariteus (Jamaican fig-eating bat), which feeds on fleshy fruit. The new species showcases the rapid diversification of archaic primates shortly after the extinction of the dinosaurs during the Paleocene, and the unusual dental anatomy of picrodontids to exploit new dietary specializations. PeerJ reviewing PDF | (v2013:06:580:1:0:NEW 30 Aug 2013) R ev ie w in g M an us cr ip t Benjamin John Burger* *Corresponding author: Department of Geology Utah State University Uintah Basin Regional Campus 320 North Aggie Blvd. Vernal, UT 84078 U.S.A. Phone 1-435-722-1778 benjamin.burger@usu.edu 1 2 3 4 5 6 7 8 9 PeerJ reviewing PDF | (v2013:06:580:1:0:NEW 30 Aug 2013) R ev ie w in g M an us cr ip t methods & materials : The fossil reported in this paper was recovered during geological mapping of the Citadel Plateau Quadrangle during the 1970s. The site has also produced a diverse fauna of mammals from the Fort Union Formation in western Colorado\u2019s Piceance Creek Basin (Burger & Honey, 2008). The site is referred to as University of Colorado Museum (UCM) Locality number 92177. Continued collection at the site over the years has produced additional fossil mammals (Table 1). However, the new species is represented by a single recovered specimen, despite 30 years of sporadic collection at the site. In the Piceance Creek Basin the Fort Union Formation is synonymous with the Atwell Gulch Member, which has been included as a member of the Wasatch Formation (Donnell, 1969) or DeBeque Formation (Kihm, 1984). In this paper I refer the Atwell Gulch Member, as the sole member of the Fort Union Formation, as it has been mapped elsewhere within the Piceance Creek Basin in western Colorado (Hail & Pipiringos, 1990; Hail & Smith, 1994). The Fort Union Formation (Atwell Gulch Mbr.) varies in thickness from 350 meters in the northeast to 196 meters in the south, and is divided into upper and lower informal units (Hail & Pipiringos, 1990). The fossil site UCM 92177 is 261 meters below the upper contact of the Late Paleocene Fort Union Formation (Atwell Gulch Member), with the Early Eocene Wasatch Formation (Molina Member). The lower unit of the Fort Union Formation consists of light-grey to light-brown sandstones; olive, purple, dark-reddish-brown claystone; and mudstones that are 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 PeerJ reviewing PDF | (v2013:06:580:1:0:NEW 30 Aug 2013) R ev ie w in g M an us cr ip t highly variegated. Large ribbon and sheet sandstone bodies are common in the north of the basin, where they can form massive sandstone cliffs measuring upwards of 25 meters thick, although most are 5-10 meters thick. Pebbles are exclusively composed of sedimentary rocks, including claystone and mudstones. The upper unit of the Fort Union Formation (Atwell Gulch Mbr.) is composed of carbonaceous shales; thin coals; and thin, but persistent highly calcareous sandstones. Mudstones and claystones are less common in the upper unit. Selenite is common, especially in the carbonaceous shales and coals. Fossil mammals are abundant in the lower unit of the Fort Union Formation, especially in the variegated beds found in association with UCM locality 92177. Fossil invertebrates, such as the bivalve Unio and a variety of gastropods are common in the upper unit, indicating a progression over time toward a more lacustrine environment. Large accumulations of gastropods are common in the upper unit, which have been interpreted as lakeshore accumulations (Hanley, 1974). Order PRIMATES Linnaeus, 1758 Family PICRODONITDAE Simpson, 1937 Genus ZANYCTERIS Matthew, 1917 ZANYCTERIS HONEYI new species Holotype\u2014 UCM 87378 right maxilla with upper first and second molars. Etymology\u2014 Posthumously named in honor of James G. Honey for his discovery of the holotype, type locality and for his kindness in allowing me to study this collection. 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 PeerJ reviewing PDF | (v2013:06:580:1:0:NEW 30 Aug 2013) R ev ie w in g M an us cr ip t Horizon\u2014 Fort Union Formation, Atwell Gulch Member, 261 meters below the top contact with the Wasatch Formation. Localities\u2014 Only known from UMC locality number 92177. Diagnosis\u2014 Differs from Z. paleocenus by having a 30% large M1/ area, while retaining a similar sized M2/ area to Z. paleocenus. Differs from Picrodus in lacking an extended parastyle on the M2/, having a better developed anterior stylar shelf, and postprotocrista on the M1/. Furthermore, the M1/ centrocrista is more curvilinear than Picrodus. Unlike Draconodus, crenulations occur in the trigon basin of M1/. Description\u2014 The holotype (UCM 87378) is the only specimen known from the Piceance Creek Basin. However, this specimen preserves morphology to indicate that it differs from Z. paleocenus from southwestern Colorado. The enlarged metastyle on the M1/ projects buccally from the posterior edge, and the stylar shelf bulges from the mid-point of the tooth. The paracingulum (the shelf formed by the paraconule) extends anteriorly, yet the stylar shelf encircling the paracone is poorly formed and resembles Picrodus rather than Zanycteris in shape. The strong postprotocrista of UCM 87378 extends directly toward the vestigial metaconule. This direct course of the postprotocrista results in a more enclosed trigon basin that closely resembles Z. paleocenus. The M2/ exhibits a wide stylar shelf, featuring a large parastyle. However, UCM 87378 typifies Zanycteris in lacking the greatly extended parastyle on the M2/ that is found in specimens of Picrodus. Despite wear, the molar paracone and metacone form a W-shaped crest across the midline on the M2/, while the broad protocone expands along the postprotocrista, a typical plesiadapiform trait. The M1/ measures 2.54 mm in length and 2.46 mm in width. The M2/ measures 1.26 mm in length and 1.60 in width. 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 PeerJ reviewing PDF | (v2013:06:580:1:0:NEW 30 Aug 2013) R ev ie w in g M an us cr ip t Comparison\u2014 Previous measurements of the holotype of Z. paleocenus (Simpson, 1935, Szalay, 1968) report a length of 2.05-2.20mm in length and 1.87-2.00 mm in width for the M1/, indicating that the new species (UCM 87378) is both larger and broader. However, reported measurements of the M2/ (1.25-1.30mm in length and 1.60-1.70mm in width) are similar in dimensions to UCM 87378. This indicates that while the first molar is enlarged, the second molar is of equal size between the two species of Zanycteris. This enlargement of the first molar likely served a functional role in providing a larger surface area for slicing between the centrocrista and cristid obliqua (Szalay, 1968). In some ways, UCM 87378 resembles Picrodus, such as the reduced anterior stylar shelf on the M1/ (Scott and Fox, 2005). However, other features more closely resemble Z. paleocenus, including the arrangement of the postprotocrista on the M1/ and smaller parastyle on the M2/. These features support inclusion of UCM 87378 within the genus Zanycteris rather than Picrodus. UCM 87378 and the holotype of Z. paleocenus (AMNH 17180) are the only two specimens of Zanycteris currently documented (Simpson, 1937; Scott and Fox, 2005). Zanycteris appears to be restricted to Colorado, while Picrodus has been found in Wyoming, Montana, Alberta and recently New Mexico (Simpson, 1937; Williams, 1985; Silcox & Gunnell, 2008; Scott & Fox, 2005; Silcox & Williamson, 2012). conclusions : In summary, the new species Zanycteris honeyi typifies the unique characteristics that set apart the Picrodontidae from other archaic primates known from the Paleocene of North America. Phylogenetic analysis supports the separation of the Picrodontidae family from the Paromomyidae family, while still recognizing their position within Pleisadapiformes. Further fossil discoveries, particularly cranial and postcranial remains will likely enable more confident placement of this unusual group of archaic primates among the evolutional tree during this pivotal time of primate diversification shortly after the extinction of the dinosaurs. acknowledgments : Special thanks to Jaelyn Eberle at the University of Colorado at Boulder for her support in my graduate studies which included this research. I would like to thank my committee Herbert H. Covert, Mary Kraus, Matthew Pranter, (University of Colorado) and Henry Fricke (Colorado College); Toni Culver, the collection manager at the University of Colorado Museum, as well as various members of the field crew including Alex Dutchak, Karen Lloyd, Pat Monaco, Andrea R. Bair, Ian J. Sweeney, Lou Taylor and Lea Ann Jolley. Fieldwork was conducted under BLM permit #C-60170 issued to the University of Colorado Museum. phylogenetic analysis : The acquisition of the highly specialized dentition found within Zanycteris and other members of the Picrodontidae remains a mystery. For example, how quickly did the specialized dentition evolve during the Paleocene? Among the known Plesiadapiformes, which one is most closely related to the family Picrodontidae and could possibly represent the ancestral condition for the specialized dentation exhibited by Zanycteris? To evaluate these questions and to work toward reconstructing the evolution of the specialized upper dentition of Picrodontidae, a phylogenetic analysis was undertaken using the 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 PeerJ reviewing PDF | (v2013:06:580:1:0:NEW 30 Aug 2013) R ev ie w in g M an us cr ip t morphological characteristics of the dentition of known Plesiadapiformes and outgroups (Paradectes, Cimolestes, and Leptacodon), which lived during the Paleocene in North America. The character matrix consisted of 113 dental characters, 97 of which were adopted from Silcox (2001). The analysis included 58 fossil taxa of contemporary North American Paleocene primates. A heuristic search using Mesquite version 2.75 (Maddison & Maddison, 2011) produced 6,579 most parsimonious trees (597 steps, consistency index [CI] = 0.36, retention index [RI] = 0.74). The strict consensus tree shows Zanycteris honeyi as closely related to Zanycteris paleocenus within a monophyletic clade of Picrodontidae (Picrodus, Draconodus, and Zanycteris). The family Picrodontidae was found to be within a clade consisting of Plesidapidae and Carpolestidae, rather than a placement within Paromomyoidea (Silcox & Gunnell, 2008). This phylogenetic position postulates that the expansion of the occlusal surface seen in the upper molars of both paromomyids and picrodontids is convergent, having evolved independently during the Paleocene. Possible ancestors of picrodontids are the early Paleocene taxa Plesiolestes, Torrejonia, Phoxomylus, and Talpohenach, while paromomyids appear to have arisen from the early Paleocene with a Palaechthon or Anasazia or the poorly known Premnoides like ancestor. Although there is ambiguity concerning the phylogenetic position of these primitive taxa, there is strong support for monophyletic clades of Paromomyidae, Picrodontidae, Plesiadapidae, and Carpolestidae, as well as a monophyletic clade of Plesiadapiformes, with the addition of Micromomyidae and Microsyopidae. Overall the resulting phylogenetic tree supports a position of Zanycteris honeyi within the Picrodontidae family, and that this new species is closely related to other archaic primates from the Paleocene of North America. Discussion 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 PeerJ reviewing PDF | (v2013:06:580:1:0:NEW 30 Aug 2013) R ev ie w in g M an us cr ip t The members of the Picrodontidae are exceptional in the development of a dentition that maximizes the shear forces along the long contact between the centrocrista on the upper molar (composed of a tall crest between the paracone and metacone) and the cristid obliqua that spans much of the length of the elongated talonid basin of the lower molar (Szalay, 1968). The molar teeth were thus probably specialized for cutting through hard outer husks of fruits and nuts. This was accomplished by positioning the fruit along the outer (buccal) shearing surface, which was greatly expanded (anteriorly and posteriorly) to maximize the amount of contact, much like a pair of long sharp scissors (Shaw, 1917). A similar expansion of the centrocrista is found in the upper molars of the Jamaican fig-eating bat (Ariteus), which feeds on the native Jamaican naseberry also known as sapodilla (Manilkara zapota), a fruit with a fleshy but firm texture (Sherwin & Gannon, 2005). Thus in both Ariteus and Zanycteris the major shearing surface is between the crests of the centrocrista above and the cristid obliqua below, demonstrating a similar specialized diet on fruit. This arrangement differs substantially from paromomyids, which retain distinct paracone and metacone cusps on the upper molars, with no development of a long and tall centrocrista between the two cusps. Rather, paromomyids expand the upper molars by broadening the postcingulum to form a wide talon basin on the lingual edge of the tooth. This broadening of the tooth functioned to expand the surface area particularly for holding food during mastication (Shaw, 1917). Thus paromomyids, such as Phenacolemur, broaden the upper molars to allow increased surface area for a larger and a more varied diet, while picrodontids, such as Zanycteris, expanded the upper molars to increase the shearing surface for a more specialized diet of a particular style of fleshy fruit. Further research on dental specialization within these groups might reveal why picrodontids have a limited stratigraphic range (Torrejonian to Tiffanian), when compared to the closely related, but dentally distinct paromomyids which ranged from the early Paleocene (Puercan) until the late Eocene (Duchesnean) (Silcox & Gunnell, 2008). 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 PeerJ reviewing PDF | (v2013:06:580:1:0:NEW 30 Aug 2013) R ev ie w in g M an us cr ip t image of fossil specimen. : Figure 1: Buccal and occlusal views of the holotype specimen, a maxilla containing the upper first and second molars. UCM 87378. PeerJ reviewing PDF | (v2013:06:580:1:0:NEW 30 Aug 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:580:1:0:NEW 30 Aug 2013) R ev ie w in g M an us cr ip t Figure 2 Strict consensus tree of the most parsimonious trees generated from the phylogenetic analysis. Figure 2: Strict consensus tree of the most parsimonious trees generated from the phylogenetic analysis of 113 upper dental characters scored against the 58 North American archaic primates known from the Paleocene. Temporal ranges during the Paleocene are shown for each species by blackened line. PeerJ reviewing PDF | (v2013:06:580:1:0:NEW 30 Aug 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:580:1:0:NEW 30 Aug 2013) R ev ie w in g M an us cr ip t Table 1(on next page) Mammal faunal list for fossil locality 92177. Table 1. Mammal faunal list for UCM locality 92177. PeerJ reviewing PDF | (v2013:06:580:1:0:NEW 30 Aug 2013) R ev ie w in g M an us cr ip t Table 1. Mammal Faunal List for UCM locality 92177. Mammalia Allotheria Multituberculata Ptilodus kummae Ectypodus musculus Theria Erinaceomorpha (\u201cApheliscidae\u201d) Haplaletes serior Litomylus ishami Phenacodaptes sabulosus Haplomylus simpsoni Lipotyphla Leptacodon tener Mesonychia cf. Sinonyx sp. Procreodi Thryptacodon australis Arctocyonides mumak Carnivoramorpha Protictis proteus Protictis cf. schaffi Primates Nannodectes gazini Plesiadapis fodinatus Chiromyoides gigas Zanycteris honeyi new species Ignacius frugivorus Ignacius sp. Carpodaptes cygneus Condylarthra Ectocion medituber Phenacodus grangeri Phenacodus magnus Pholidota Propalaeanodon sp. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 PeerJ reviewing PDF | (v2013:06:580:1:0:NEW 30 Aug 2013) R ev ie w in g M an us cr ip t",
+    "v3_text": "methods & materials : The fossil reported in this paper was recovered during geological mapping of the Citadel Plateau Quadrangle during the 1970s. The site has also produced a diverse fauna of mammals from the Fort Union Formation in western Colorado\u2019s Piceance Creek Basin (Burger & Honey, 2008). The site is referred to as University of Colorado Museum (UCM) Locality number 92177. Continued collection at the site over the years has produced a diverse fauna of fossil mammals (Table 1). However, the new species is represented by a single recovered specimen, despite 30 years of sporadic collection at the site. In the Piceance Creek Basin the Fort Union Formation is synonymous with the Atwell Gulch Member, which has been included as a member of the Wasatch Formation (Donnell, 1969) or DeBeque Formation (Kihm, 1984). In this paper I refer the Atwell Gulch Member, as the sole member of the Fort Union Formation, as it has been mapped elsewhere within the Piceance Creek Basin in western Colorado (Hail & Pipiringos, 1990; Hail & Smith, 1994). The Fort Union Formation (Atwell Gulch Mbr.) varies in thickness from 350 meters in the northeast to 196 meters in the south, and is divided into upper and lower informal units (Hail & 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 PeerJ reviewing PDF | (v2013:06:580:0:0:NEW 12 Jun 2013) R ev ie w in g M an us cr ip t Pipiringos, 1990). The fossil site UCM 92177 is 261 meters below the upper contact of the Late Paleocene Fort Union Formation (Atwell Gulch Member), with the Early Eocene Wasatch Formation (Molina Member). The lower unit of the Fort Union Formation consists of light-grey to light-brown sandstones; olive, purple, dark-reddish-brown claystone; and mudstones that are highly variegated. Large ribbon and sheet sandstone bodies are common in the north of the basin, where they can form massive sandstone cliffs measuring upwards of 25 meters thick, although most are 5-10 meters thick. Pebbles are exclusively composed of sedimentary rocks, including claystone and mudstones. The upper unit of the Fort Union Formation (Atwell Gulch Mbr.) is composed of carbonaceous shales; thin coals; and thin, but persistent highly calcareous sandstones. Mudstones and claystones are less common in the upper unit. Selenite is common, especially in the carbonaceous shales and coals. Fossil mammals are abundant in the lower unit of the Fort Union Formation, especially in the variegated beds found in association with UCM locality 92177. Fossil invertebrates, such as the bivalve Unio and a variety of gastropods are common in the upper unit, indicating a progression over time toward a more lacustrine environment. Large accumulations of gastropods are common in the upper unit, which have been interpreted as lakeshore accumulations (Hanley, 1974) Genus ZANYCTERIS Matthew, 1917 ZANYCTERIS HONEYI new species Holotype\u2014 UCM 87378 right maxilla with upper first and second molars. 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 PeerJ reviewing PDF | (v2013:06:580:0:0:NEW 12 Jun 2013) R ev ie w in g M an us cr ip t Etymology\u2014 Posthumously named in honor of James G. Honey for his discovery of the holotype, type locality and for his kindness in allowing me to study this collection. Horizon\u2014 Fort Union Formation, Atwell Gulch Member, 261 meters below the top contact with the Wasatch Formation. Localities\u2014 Only known from UMC locality number 92177. Diagnosis\u2014 Differs from Z. paleocenus by having a 30% large M1/ area, while retaining a similar sized M2/ area to Z. paleocenus. Differs from Picrodus in lacking an extended parastyle on the M2/, having a better developed anterior stylar shelf, and postprotocrista on the M1/. Furthermore, the M1/ centrocrista is more curvilinear than Picrodus. Unlike Draconodus, crenulations occur in the trigon basin of M1/. Description\u2014 The holotype (UCM 87378) is the only specimen known from the Piceance Creek Basin. However, this specimen preserves morphology to indicate that it differs from Z. paleocenus from southwestern Colorado. The enlarged metastyle on the M1/ projects buccally from the posterior edge, and the stylar shelf bulges from the mid-point of the tooth. The paracingulum (the shelf formed by the paraconule) extends anteriorly, yet the stylar shelf encircling the paracone is poorly formed and resembles Picrodus rather than Zanycteris in shape. The strong postprotocrista of UCM 87378 extends directly toward the vestigial metaconule. This direct course of the postprotocrista results in a more enclosed trigon basin that closely resembles Z. paleocenus. The M2/ exhibits a wide stylar shelf, featuring a large parastyle. However, UCM 87378 typifies Zanycteris in lacking the greatly extended parastyle on the M2/ that is found in specimens of Picrodus. Despite wear, the molar paracone and metacone form a W-shaped crest across the midline on the M2/, while the broad protocone spreads into the postprotocrista, a 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 PeerJ reviewing PDF | (v2013:06:580:0:0:NEW 12 Jun 2013) R ev ie w in g M an us cr ip t typical plesiadapiform trait. The M1/ measures 2.54 mm in length and 2.46 mm in width. The M2/ measures 1.26 mm in length and 1.60 in width. Comparison\u2014 Previous measurements of the holotype of Z. paleocenus (Simpson, 1935, Szalay, 1968) report a length of 2.05-2.20mm in length and 1.87-2.00 mm in width for the M1/, indicating that the new species (UCM 87378) is both larger and broader. However, reported measurements of the M2/ (1.25-1.30mm in length and 1.60-1.70mm in width) are similar in dimensions to UCM 87378. This indicates that while the first molar is enlarged, the second molar is of equal size between the two species of Zanycteris. This enlargement of the first molar likely served a functional role in providing a larger surface area for slicing between the centrocrista and cristid obliqua (Szalay, 1968). In some ways, UCM 87378 resembles Picrodus, such as the reduced anterior stylar shelf on the M1/. However, other features more closely resemble Z. paleocenus, including the arrangement of the postprotocrista on the M1/ and smaller parastyle on the M2/. These features support inclusion of UCM 87378 within the genus Zanycteris rather than Picrodus. UCM 87378 and the holotype of Z. paleocenus (AMNH 17180) are the only two specimens of Zanycteris currently documented (Simpson, 1937; Scott and Fox, 2005). Zanycteris appears to be restricted to Colorado, while Picrodus has been found in Wyoming, Montana, Alberta and recently New Mexico (Simpson, 1937; Williams, 1985; Silcox & Gunnell, 2008; Scott & Fox, 2005; Silcox & Williamson, 2012). discussion : The members of the Picrodontidae are exceptional in the development of a dentition that maximizes the shear forces along the long contact between the centrocrista on the upper molar (composed of a tall crest between the paracone and metacone) and the cristid obliqua that spans much of the length of the elongated talonid basin of the lower molar (Szalay, 1968). The molar teeth were thus specialized for cutting through hard outer husks of fruits and nuts. This was accomplished by positioning the fruit along the outer (buccal) shearing surface, which was greatly expanded (anteriorly and posteriorly) to maximize the amount of contact, much like a pair of long sharp scissors (Shaw, 1917). A similar expansion of the centrocrista is found in the upper molars of the Jamaican fig-eating bat (Artiteus), which feeds on the native Jamaican naseberry also known as sapodilla (Manilkara zapota), a fruit with a fleshy but firm texture (Sherwin & Gannon, 2005). Thus in both Ariteus and Zanycteris the major shearing surface is between the crests of the centrocrista above and the cristid obliqua below, demonstrating a similar specialized diet on fruit. This arrangement differs substantially from paromomyids, which retain distinct paracones and metacones cusps on the upper molars, with no development of a long and tall centrocrista between the two cusps. Rather, paromomyids expand the upper molars by broadening the postcingulum to form a wide talon basin on the lingual edge of the tooth. This broadening of the tooth functioned to expand the surface area particularly for holding food during mastication (Shaw, 1917). Thus paromomyids, such as Phenacolemur, broaden the upper molars to allow increased surface area for a larger and a more varied diet, while picrodontids, such as Zanycteris, expanded the upper molars to increase the shearing surface for a more specialized diet of a particular style of fleshy fruit. This specialization may also explain the 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 PeerJ reviewing PDF | (v2013:06:580:0:0:NEW 12 Jun 2013) R ev ie w in g M an us cr ip t limited stratigraphic range (Torrejonian and Tiffanian) of Picrodontids, when compared to Paromomyids which ranged from the early Paleocene (Puercan) to late Eocene (Duchesnean) (Silcox & Gunnell, 2008). a new species of an archaic primate (pleisadapiformes) is described based on a maxilla containing the : first and second upper molars from the Fort Union Formation, Atwell Gulch Member in northwestern Colorado. The preserved teeth show the unusual dental characteristics of members of the rare and poorly documented Picrodontidae family, including an elongated centrocrista and wide occlusal surface. The new species is placed within the genus Zanycteris (represented from a single specimen from southern Colorado). This placement is based on similarities in regard to a shorten parastyle, curvilinear centrocrista, and wider anterior stylar shelf on the upper molars. However, the new species differs from the only known species of Zanycteris in exhibiting an upper first molar that is 30% larger in area, while retaining a similar sized upper second molar. Phylogenetic analysis supports the separation of the Picrodontidae family from the Paromomyidae, while still recognizing picrodontids position within Pleisadapiformes. The unusual dental features of the upper molars likely functioned in life as an enhanced shearing surface between the centrocrista and cristid oblique crests for a specialized diet of fruit. A similar arrangement is found in the living bat Artiteus (Jamaican fig-eating bat), which feeds on fleshy fruit. The new species showcases the rapid diversification of archaic primates shortly after the extinction of the dinosaurs during the Paleocene, and the unusual dental anatomy of picrodontids to exploit new dietary specializations. PeerJ reviewing PDF | (v2013:06:580:0:0:NEW 12 Jun 2013) R ev ie w in g M an us cr ip t Benjamin John Burger* *Corresponding author: Department of Geology Utah State University Uintah Basin Regional Campus 320 North Aggie Blvd. Vernal, UT 84078 U.S.A. Phone 1-435-722-1778 benjamin.burger@usu.edu 1 2 3 4 5 6 7 8 9 PeerJ reviewing PDF | (v2013:06:580:0:0:NEW 12 Jun 2013) R ev ie w in g M an us cr ip t acknowledgments : Special thanks to Jaelyn Eberle at the University of Colorado at Boulder for her support in my graduate studies which included this research. I would like to thank my committee Herbert H. Covert, Mary Kraus, Matthew Pranter, (University of Colorado) and Henry Fricke (Colorado College); Toni Culver, the collection manager at the University of Colorado Museum, as well as various members of the field crew including Alex Dutchak, Karen Lloyd, Pat Monaco, Andrea R. Bair, Ian J. Sweeney, Lou Taylor and Lea Ann Jolley. Fieldwork was conducted under BLM permit #C-60170 issued to the University of Colorado Museum. References 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 PeerJ reviewing PDF | (v2013:06:580:0:0:NEW 12 Jun 2013) R ev ie w in g M an us cr ip t Burger BJ, Honey JG. 2008. Plesiadapidae (Mammalia, Primates) from the late Paleocene Fort Union Formation of the Piceance Creek Basin, Colorado. Journal of Vertebrate Paleontology 28(3):816-825. Clemens WA. 2004. Purgatorius (Plesiadapiformes, Primates?, Mammalia), a Paleocene immigrant into northeastern Montana: stratigraphic occurrences and incisor proportions. Bulletin of Carnegie Museum of Natural History 36:3-13. Donnell JR. 1969. Paleocene and lower Eocene units in the southern part of the Piceance Creek Basin, Colorado. U.S. Geological Survey Bulletin 1274-M:1-18. Hail WJ, Pipiringos GN. 1990. Geologic map of the lower Piceance Creek area, northwestern Colorado. U.S. Geological Survey IMAP 1936 Scale 1:100,000. Hail WJ, Smith MC. 1994. Geologic map of the northern part of the Piceance Creek basin, northwestern Colorado. U.S. Geological Survey IMAP 2400 Scale 1:100,000. Hanley JH. 1974. Systematics, paleoecology, and biostratigraphy of nonmarine Mollusca from the Green River and Wasatch Formations (Eocene), southwestern Wyoming and northwestern Colorado. PhD. thesis, University of Wyoming, Laramie, Wyoming, 1-285. Kihm AJ. 1984. Early Eocene mammalian faunas of the Piceance Creek Basin, northwestern Colorado. D. Phil. Thesis. University of Colorado at Boulder, Boulder Colorado. Maddison WP, Maddison DR. 2011. Mesquite: a modular system for evolutionary analysis. Version 2.75 http://mesquiteproject.org Matthew WD. 1917. A Paleocene bat. Bulletin of the American Museum of Natural History 37:569-571. 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 PeerJ reviewing PDF | (v2013:06:580:0:0:NEW 12 Jun 2013) R ev ie w in g M an us cr ip t Scott CS, Fox RC. 2005. Windows on the evolution of Picrodus (Plesiadapiformes: Primates): Morphology and relationships of a species complex from the Paleocene of Alberta. Journal of Paleontology 79(4):635-657. Shaw DM. 1917. Form and function of teeth: A theory of \u201cMaximum Shear.\u201d Journal of Anatomy 52:97-106. Sherwin RE, Gannon WL. 2005. Ariteus flavescens. Mammalian Species. 787:1-3. Silcox MT. 2001. A phylogenetic analysis of the Plesiadapiformes and their relationship to Euprimates and other archontans. D. Phil. Thesis, Johns Hopkins University, Baltimore, Maryland. Silcox MT, Gunnell GF. 2008. Plesiadapiformes. In: Janis CM, Gunnell GF, Uhen, MD, eds. Evolution of Tertiary Mammals of North America, Volume 2: Small Mammals, Xenarthrans, and Marine Mammals. Cambridge: Cambridge University Press, 207-238. Silcox MT, Williamson TE. 2012. New discovers of early Paleocene (Torrejonian) primates from the Nacimiento Formation, San Juan Basin, New Mexico. Journal of Human Evolution 63:805-833. Simmons NB, Geisler JH. 1998. Phylogenetic relationships of Icaronycteris, Archaeonycteris, Hassianycteris, and Palaeochiropteryx to extant bat lineages, with comments on the evolution of echolocation and foraging strategies in Microchiroptera. Bulletin of the American Museum of Natural History 235:1-182. Simpson GG. 1935. The Tiffany fauna, upper Paleocene I. Multituberculata, Marsupialia, Insectivora, and ?Chiroptera. American Museum Novitates. 795:1-19. Simpson GG. 1937. The Fort Union of the Crazy Mountain Field, Montana and its mammalian faunas. Smithsonian Institution United States National Museum Bulletin 169:1-277. 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 PeerJ reviewing PDF | (v2013:06:580:0:0:NEW 12 Jun 2013) R ev ie w in g M an us cr ip t Szalay FS. 1968. The Picrodontidae, a family of early primates. American Museum Novitates. 2329:1-55. Tomida Y. 1982. A new genus of picrodontid primate from the Paleocene of Utah. Folia Primatologica. 37:37-43. Williams JA. 1985. Morphology and variation in the posterior dentition of Picrodus silberlingi (Picrodontidae). Folia Primatologica. 45: 48-58. 227 228 229 230 231 232 PeerJ reviewing PDF | (v2013:06:580:0:0:NEW 12 Jun 2013) R ev ie w in g M an us cr ip t Figure 1 Holotype specimen (UCM 87378) Buccal and occlusal views showing the maxilla containing the upper first and second molars. PeerJ reviewing PDF | (v2013:06:580:0:0:NEW 12 Jun 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:580:0:0:NEW 12 Jun 2013) R ev ie w in g M an us cr ip t Figure 2 Strict consensus tree of the most parsimonious trees generated from the phylogenetic analysis. Tree based on 113 dental characters scored against the 58 North American archaic primates known from the Paleocene. Upper dentitions are illustrated representing each fossil species examined. Temporal ranges during the Paleocene are shown for each species by blackened lines. PeerJ reviewing PDF | (v2013:06:580:0:0:NEW 12 Jun 2013) R ev ie w in g M an us cr ip t PeerJ reviewing PDF | (v2013:06:580:0:0:NEW 12 Jun 2013) R ev ie w in g M an us cr ip t phylogenetic analysis : The acquisition of the highly specialized dentition found within Zanycteris and other members of the Picrodontidae remains a mystery. For example, how quickly did the specialized dentition evolve during the Paleocene? Among the known Plesiadapiformes, which one is most closely 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 PeerJ reviewing PDF | (v2013:06:580:0:0:NEW 12 Jun 2013) R ev ie w in g M an us cr ip t related to the family Picrodontidae and could possibly represent the ancestral condition for the specialized dentation exhibited by Zanycteris? To evaluate these questions and to work toward reconstructing the evolution of the specialized upper dentition of Picrodontidae, a phylogenetic analysis was undertaken using the morphological characteristics of the dentition of known Plesiadapiformes and outgroups (Paradectes, Cimolestes, and Leptacodon), which lived during the Paleocene in North America. The character matrix consisted of 113 dental characters, 97 of which were adopted from Silcox (2001). The analysis included 58 fossil taxa of contemporary North American Paleocene primates. A heuristic search using Mesquite version 2.75 (Maddison & Maddison, 2011) produced 6,579 most parsimonious trees (597 steps, consistency index [CI] = 0.36, retention index [RI] = 0.74). The strict consensus tree shows Zanycteris honeyi as the sister-taxa to Zanycteris paleocenus within a monophyletic clade of Picrodontidae (Picrodus, Draconodus, and Zanycteris). The family Picrodontidae was found to be within a clade consisting of Plesidapidae and Carpolestidae, rather than the previous placement within Paromomyoidea (Silcox & Gunnell, 2008). This phylogenetic position implies that the expansion of the occlusal surface seen in the upper molars of both paromomyids and picrodontids is convergent, having evolved independently during the Paleocene. Possible ancestors of picrodontids are the early Paleocene taxa Plesiolestes, Torrejonia, Phoxomylus, and Talpohenach, while paromomyids appear to have arisen from the early Paleocene Palaechthon or Anasazia. Although there is some ambiguity concerning the phylogenetic position of these primitive taxa, there is support for a monophyletic clade of Paromomyidae, Picrodontidae, Plesiadapidae, and Carpolestidae, as well as a monophyletic clade of Plesiadapiformes, with the addition of Micromomyidae and Microsyopidae. 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 PeerJ reviewing PDF | (v2013:06:580:0:0:NEW 12 Jun 2013) R ev ie w in g M an us cr ip t conclusions : In summary, the new species Zanycteris honeyi typifies the unique characteristics that set apart the Picrodontidae from other archaic primates known from the Paleocene of North America. Phylogenetic analysis supports the separation of the Picrodontidae family from the Paromomyidae family, while still recognizing their position within Pleisadapiformes. Further fossil discoveries, particularly cranial and postcranial remains will likely enable more confident placement of this unusual group of archaic primates among the evolutional tree during this pivotal time of primate diversification shortly after the extinction of the dinosaurs.",
+    "url": "https://peerj.com/articles/192/reviews/",
+    "review_1": "Desmond Tobin \u00b7 Oct 8, 2013 \u00b7 Academic Editor\nACCEPT\nSome of the last remaining points have been addressed satisfactorially, and the paper overall now makes a nice contribution to the literature in this area.",
+    "review_2": "Desmond Tobin \u00b7 Sep 30, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nThe authors are to be commended on a much improved manuscript, with the majority of reviewers\u2019 comments addressed. I believe final polishing of this manuscript could be undertaken by attending to a small number of minor issues before formal acceptance of the paper.\n\nWhile the biochemical nature of the protein factor, and of the tendon and bone cell cofactors, has been detailed in this manuscript, there appears to be no formal proof per se that either the protein/cofactor or cofactor alone control cell density. Thus, it would be prudent to alter the statement \u201cfurther proof of its role\u201d in the Abstract to something like \u201cin support of this proposed role\u2026\u201d.\n\nSimilarly, could the authors look again at those parts of the manuscript describing the biological role of the protein and cofactors, as the writing here needs to be consistent with the proposed model of controlling cell density. As current written this appears to state explicitly that cell density signaling is controlled by a protein plus cell-specific cofactor.\n\nAs the reviewer writes directly, greater prudence is required when ascribing alkaline phosphatase expression to bone production per se, rather than as a phenotypic marker of bone cells.\n\nPlease provide a clear reference from the literature in the your manuscript text of the origin, characterization and use of the U2OS cells. The current supporting statement is rather weak i.e.,\u201d from frozen stocks that have been in the laboratory for over a dozen years\".",
+    "review_3": "Reviewer 1 \u00b7 Sep 29, 2013\nBasic reporting\nThe paper has been improved. The previous version had mislabeled figures, which very much complicated the previous review. The writing is still in places vague, but it has improved from the previous version, which was quite poor. Even the rebuttal letter is odd - here is a section of the response to Reviewer 2:\n\"One must remember that PAT cells were less than a week earlier in an embryo. They will not be polite like U2OS cells and wait for things to develop.\"\nPolite? How can a cell be polite? To be honest, in many many years of reviewing papers, I have never read anything quite as odd.\nMany of the points previously raised have not been answered. Here is just one example: I previously asked the simple question \"give a ref for U2OS\". This means to give a paper reference for these cells, such as \"Smith et al, 1967\", and then in the References give the authors, journal name, year, volume, and page numbers. The reference still in the paper is\n\"(from frozen stocks that have been in the laboratory for over a dozen years)\".\nExperimental design\nno comments\nValidity of the findings\nno comments\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Cells determine cell density using a small protein bound to a unique tissue-specific phospholipid (v0.2)\". PeerJ https://doi.org/10.7287/peerj.192v0.2/reviews/1",
+    "review_4": "Reviewer 2 \u00b7 Sep 25, 2013\nBasic reporting\nNo comments\nExperimental design\nNo comments\nValidity of the findings\nThe majority of this reviewer\u2019s minor comments have been addressed in this revised manuscript. However, some general considerations remain. The biochemical natures of the protein factor, and of the tendon and bone cell cofactors, have been detailed in this manuscript; these findings are valid. Yet, there is no proof that either protein/cofactor or cofactor alone control cell density. Thus, statements such \u201cfurther proof of its role\u201d in the abstract should be modified to something like \u201cin support of this proposed role\u2026\u201d. Overall the parts of the manuscript describing the biological role of the protein and cofactors need to be rewritten to indicate the results are consistent with the proposed model of controlling cell density, rather than stating explicitly that cell density signaling is controlled by a protein plus cell-specific cofactor.\nAdditional comments\nLines 287-288: alkaline phosphatase is a marker of bone cells, but does not mean that bone is being produced. The sentence could read \u201cthe ridges produced alkaline phosphatase on the cell surface, indicating increased expression of the osteoblast phenotype.\u201d\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"Cells determine cell density using a small protein bound to a unique tissue-specific phospholipid (v0.2)\". PeerJ https://doi.org/10.7287/peerj.192v0.2/reviews/2",
+    "pdf_1": "https://peerj.com/articles/192v0.3/submission",
+    "pdf_2": "https://peerj.com/articles/192v0.2/submission",
+    "review_5": "Desmond Tobin \u00b7 Jul 26, 2013 \u00b7 Academic Editor\nMAJOR REVISIONS\nPlease pay particular attention to addressing the very significant concerns of Reviewer 1, especially those pertaining to appropriate controls, activity assays and on how you determined the significance of your results.",
+    "pdf_3": "https://peerj.com/articles/192v0.1/submission",
+    "review 6": "Reviewer 1 \u00b7 Jul 18, 2013\nBasic reporting\nNo comments\nExperimental design\nnot rigorous - key controls missing, no description of repeating experiments\nmethods missing sufficient detail\nValidity of the findings\nno statistics\nno clear conclusion, other than the experiment didn't work\nAdditional comments\nThis manuscript examines a potentially interesting topic \u2013 how cells determine their density. Unfortunately, key experiments have not been done, key controls have not been done, and the manuscript appears to have many logic errors. In addition, the writing is very opaque, and much of the writing doesn\u2019t seem to make sense. Major points\nThere is no control for the immunofluorescence, and the antibody was used at 1:40. At this high concentration, any antibody will stain any cell or tissue. You need a preimmune control. Figure 1 shows that two different anti-myc antibodies cannot detect a myc-tagged version of \u201cthe protein\u201d made in E. coli. This suggests that you have a frameshift mutation somewhere in the cDNA, and are not making the correct fusion protein. The polyclonal can barely detect the protein used as an antigen- also indicating that something is wrong here. A variety of purification / modification steps are shown for some sort of protein. However, the critical thing when purifying/ modifying a factor is to show whether it has activity. There are no activity assays done for the various treatments of the factor. You need to do test the activity of a series of concentrations of the \u2018factor\u2019, or treated factor, since many signals show activity in a concentration range that has lower and upper limits. In the mass spectrometry, the lipase buffer control should look pretty much like the untreated material, whereas it actually looks almost exactly like the lipase treatment. Something is clearly wrong here. Two tiny peaks that are different between lipase and lipase buffer control were chosen for analysis, but there are mmany other peaks that could have been similarly chosen. Also, why is only a small region of the mass spectrum shown? Most importantly, there is no clear demonstration that you have purified or identified the factor affecting tendon cells. Minor points\nAbstract-\n\u201cCell density is the critical parameter controlling tendon morphogenesis\u201d \u2013 I imagine many other parameters are also just as critical. Also, what is the evidence for this opening statement? \u201cwith an affinity for the cell layer\u201d what cell layer? A cell layer in the tendon? \u201cthe band that best correlated with a cell proliferation assay\u201d How can a band (what kind of band?) \u2018correlate\u2019 with an assay? \u201cTo function as a SNZR would require that the\nfull length protein be cleaved to a smaller protein, then secreted\u201d Why is this a requirement? \u201cthe chicken cDNA\u201d The cDNA for what? \u201cto test whether the recombinant protein would exhibit the expected activity\u201d What activity? \u201cOutside the cell, a small band was detected\u201d What kind of band? Protein? How was it detected? \u201cSignal transduction is postulated to occur by an increased ordering\nof the plasma membrane\u201d Where does this come from? Aren\u2019t most signals sensed by receptors? Page, line\n2,28 \u201cbecause induction is slow from a single copy gene.\u201d What is the evidence for this? Lots of single copy genes can show a fast induction.\n2, 32 \u201cSo manipulating procollagen mRNA levels\nis not feasible when the cells are required to make high levels of procollagen\u201d This doesn\u2019t make sense.\n3,14 \u201ccaused a dramatic change\u201d increase? Decrease?\n3,23 \u201cThe one described and a second that is not diffusible but interacts with\nthe first and changes the cellular response(Schwarz 1996).\u201d This sentence doesn\u2019t make any sense.\n3,29 \u201cThis classic approach is complicated by the\nfinding that the cell type producing this protein binds a unique tissue-specific lipid cofactor\nand this composite molecule imparts a tissue-specific response\u201d This sentence is too vague\n4,7 give a ref for U2OS, also, why do we need to know \u201c(from frozen stocks that have been in the laboratory for over a dozen years)\u201d?\n4,12 if you put 12 ml into a 225 cm2 flask, there will be so little liquid therer the cells will dry up, yes?\n4,26 give the primers used\n\u201cThis yielded an expressed protein\nthat was 7 AA smaller in the linker between the protein and the tags\u201d where does this come from?\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Cells determine cell density using a small protein bound to a unique tissue-specific phospholipid (v0.1)\". PeerJ https://doi.org/10.7287/peerj.192v0.1/reviews/1 Reviewer 2 \u00b7 Jul 10, 2013\nBasic reporting\nSome of the introduction is too detailed and not relevant to the current research. For example, the discussion of collagen production and regulation of PAT cells could be consolidated.\nThe introduction should include a discussion of what is known about the gene from which the cell density factor is derived, such as its known cellular functions, cellular location, and other known cleavage forms if they exist.\nThe nature of the Schwartz (2002) reference should be specified (U.S. Patent?).\nExperimental design\nThe results section could be improved by simply stating what experiments had been conducted and why, rather than introducing each with an extensive discourse on the rationale.\nIt is not clear exactly what was expressed in E. coli, i.e. which portion of the gene was tagged and expressed. Therefore it is not convincing evidence that the observed larger-than-expected U2OS-expressed protein is aggregated, rather than a longer form with additional amino acid sequences.\nIt should be clarified why U2OS cells were used as the heterologous expression system, rather than a more standard eukaryotic expression system. Assuming the authors are correct that different cell types express different cofactors, a less committed/differentiated cell type may have been more appropriate.\nValidity of the findings\nThe reader deduces that the straightforward experiment of taking the U2OS conditioned medium, either straight or partially purified, to PAT cells failed to affect their growth. These experiments and results should be more completely described.\nAdditional comments\nThis manuscript describes a complex system which is involved in the cellular sensing of cell density. The factor involves a relatively small protein associated with a lipid-based cofactor, a unique situation which has greatly complicated the characterization of this activity.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"Cells determine cell density using a small protein bound to a unique tissue-specific phospholipid (v0.1)\". PeerJ https://doi.org/10.7287/peerj.192v0.1/reviews/2",
+    "all_reviews": "Review 1: Desmond Tobin \u00b7 Oct 8, 2013 \u00b7 Academic Editor\nACCEPT\nSome of the last remaining points have been addressed satisfactorially, and the paper overall now makes a nice contribution to the literature in this area.\nReview 2: Desmond Tobin \u00b7 Sep 30, 2013 \u00b7 Academic Editor\nMINOR REVISIONS\nThe authors are to be commended on a much improved manuscript, with the majority of reviewers\u2019 comments addressed. I believe final polishing of this manuscript could be undertaken by attending to a small number of minor issues before formal acceptance of the paper.\n\nWhile the biochemical nature of the protein factor, and of the tendon and bone cell cofactors, has been detailed in this manuscript, there appears to be no formal proof per se that either the protein/cofactor or cofactor alone control cell density. Thus, it would be prudent to alter the statement \u201cfurther proof of its role\u201d in the Abstract to something like \u201cin support of this proposed role\u2026\u201d.\n\nSimilarly, could the authors look again at those parts of the manuscript describing the biological role of the protein and cofactors, as the writing here needs to be consistent with the proposed model of controlling cell density. As current written this appears to state explicitly that cell density signaling is controlled by a protein plus cell-specific cofactor.\n\nAs the reviewer writes directly, greater prudence is required when ascribing alkaline phosphatase expression to bone production per se, rather than as a phenotypic marker of bone cells.\n\nPlease provide a clear reference from the literature in the your manuscript text of the origin, characterization and use of the U2OS cells. The current supporting statement is rather weak i.e.,\u201d from frozen stocks that have been in the laboratory for over a dozen years\".\nReview 3: Reviewer 1 \u00b7 Sep 29, 2013\nBasic reporting\nThe paper has been improved. The previous version had mislabeled figures, which very much complicated the previous review. The writing is still in places vague, but it has improved from the previous version, which was quite poor. Even the rebuttal letter is odd - here is a section of the response to Reviewer 2:\n\"One must remember that PAT cells were less than a week earlier in an embryo. They will not be polite like U2OS cells and wait for things to develop.\"\nPolite? How can a cell be polite? To be honest, in many many years of reviewing papers, I have never read anything quite as odd.\nMany of the points previously raised have not been answered. Here is just one example: I previously asked the simple question \"give a ref for U2OS\". This means to give a paper reference for these cells, such as \"Smith et al, 1967\", and then in the References give the authors, journal name, year, volume, and page numbers. The reference still in the paper is\n\"(from frozen stocks that have been in the laboratory for over a dozen years)\".\nExperimental design\nno comments\nValidity of the findings\nno comments\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Cells determine cell density using a small protein bound to a unique tissue-specific phospholipid (v0.2)\". PeerJ https://doi.org/10.7287/peerj.192v0.2/reviews/1\nReview 4: Reviewer 2 \u00b7 Sep 25, 2013\nBasic reporting\nNo comments\nExperimental design\nNo comments\nValidity of the findings\nThe majority of this reviewer\u2019s minor comments have been addressed in this revised manuscript. However, some general considerations remain. The biochemical natures of the protein factor, and of the tendon and bone cell cofactors, have been detailed in this manuscript; these findings are valid. Yet, there is no proof that either protein/cofactor or cofactor alone control cell density. Thus, statements such \u201cfurther proof of its role\u201d in the abstract should be modified to something like \u201cin support of this proposed role\u2026\u201d. Overall the parts of the manuscript describing the biological role of the protein and cofactors need to be rewritten to indicate the results are consistent with the proposed model of controlling cell density, rather than stating explicitly that cell density signaling is controlled by a protein plus cell-specific cofactor.\nAdditional comments\nLines 287-288: alkaline phosphatase is a marker of bone cells, but does not mean that bone is being produced. The sentence could read \u201cthe ridges produced alkaline phosphatase on the cell surface, indicating increased expression of the osteoblast phenotype.\u201d\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"Cells determine cell density using a small protein bound to a unique tissue-specific phospholipid (v0.2)\". PeerJ https://doi.org/10.7287/peerj.192v0.2/reviews/2\nReview 5: Desmond Tobin \u00b7 Jul 26, 2013 \u00b7 Academic Editor\nMAJOR REVISIONS\nPlease pay particular attention to addressing the very significant concerns of Reviewer 1, especially those pertaining to appropriate controls, activity assays and on how you determined the significance of your results.\nReview 6: Reviewer 1 \u00b7 Jul 18, 2013\nBasic reporting\nNo comments\nExperimental design\nnot rigorous - key controls missing, no description of repeating experiments\nmethods missing sufficient detail\nValidity of the findings\nno statistics\nno clear conclusion, other than the experiment didn't work\nAdditional comments\nThis manuscript examines a potentially interesting topic \u2013 how cells determine their density. Unfortunately, key experiments have not been done, key controls have not been done, and the manuscript appears to have many logic errors. In addition, the writing is very opaque, and much of the writing doesn\u2019t seem to make sense. Major points\nThere is no control for the immunofluorescence, and the antibody was used at 1:40. At this high concentration, any antibody will stain any cell or tissue. You need a preimmune control. Figure 1 shows that two different anti-myc antibodies cannot detect a myc-tagged version of \u201cthe protein\u201d made in E. coli. This suggests that you have a frameshift mutation somewhere in the cDNA, and are not making the correct fusion protein. The polyclonal can barely detect the protein used as an antigen- also indicating that something is wrong here. A variety of purification / modification steps are shown for some sort of protein. However, the critical thing when purifying/ modifying a factor is to show whether it has activity. There are no activity assays done for the various treatments of the factor. You need to do test the activity of a series of concentrations of the \u2018factor\u2019, or treated factor, since many signals show activity in a concentration range that has lower and upper limits. In the mass spectrometry, the lipase buffer control should look pretty much like the untreated material, whereas it actually looks almost exactly like the lipase treatment. Something is clearly wrong here. Two tiny peaks that are different between lipase and lipase buffer control were chosen for analysis, but there are mmany other peaks that could have been similarly chosen. Also, why is only a small region of the mass spectrum shown? Most importantly, there is no clear demonstration that you have purified or identified the factor affecting tendon cells. Minor points\nAbstract-\n\u201cCell density is the critical parameter controlling tendon morphogenesis\u201d \u2013 I imagine many other parameters are also just as critical. Also, what is the evidence for this opening statement? \u201cwith an affinity for the cell layer\u201d what cell layer? A cell layer in the tendon? \u201cthe band that best correlated with a cell proliferation assay\u201d How can a band (what kind of band?) \u2018correlate\u2019 with an assay? \u201cTo function as a SNZR would require that the\nfull length protein be cleaved to a smaller protein, then secreted\u201d Why is this a requirement? \u201cthe chicken cDNA\u201d The cDNA for what? \u201cto test whether the recombinant protein would exhibit the expected activity\u201d What activity? \u201cOutside the cell, a small band was detected\u201d What kind of band? Protein? How was it detected? \u201cSignal transduction is postulated to occur by an increased ordering\nof the plasma membrane\u201d Where does this come from? Aren\u2019t most signals sensed by receptors? Page, line\n2,28 \u201cbecause induction is slow from a single copy gene.\u201d What is the evidence for this? Lots of single copy genes can show a fast induction.\n2, 32 \u201cSo manipulating procollagen mRNA levels\nis not feasible when the cells are required to make high levels of procollagen\u201d This doesn\u2019t make sense.\n3,14 \u201ccaused a dramatic change\u201d increase? Decrease?\n3,23 \u201cThe one described and a second that is not diffusible but interacts with\nthe first and changes the cellular response(Schwarz 1996).\u201d This sentence doesn\u2019t make any sense.\n3,29 \u201cThis classic approach is complicated by the\nfinding that the cell type producing this protein binds a unique tissue-specific lipid cofactor\nand this composite molecule imparts a tissue-specific response\u201d This sentence is too vague\n4,7 give a ref for U2OS, also, why do we need to know \u201c(from frozen stocks that have been in the laboratory for over a dozen years)\u201d?\n4,12 if you put 12 ml into a 225 cm2 flask, there will be so little liquid therer the cells will dry up, yes?\n4,26 give the primers used\n\u201cThis yielded an expressed protein\nthat was 7 AA smaller in the linker between the protein and the tags\u201d where does this come from?\nCite this review as\nAnonymous Reviewer (2013) Peer Review #1 of \"Cells determine cell density using a small protein bound to a unique tissue-specific phospholipid (v0.1)\". PeerJ https://doi.org/10.7287/peerj.192v0.1/reviews/1 Reviewer 2 \u00b7 Jul 10, 2013\nBasic reporting\nSome of the introduction is too detailed and not relevant to the current research. For example, the discussion of collagen production and regulation of PAT cells could be consolidated.\nThe introduction should include a discussion of what is known about the gene from which the cell density factor is derived, such as its known cellular functions, cellular location, and other known cleavage forms if they exist.\nThe nature of the Schwartz (2002) reference should be specified (U.S. Patent?).\nExperimental design\nThe results section could be improved by simply stating what experiments had been conducted and why, rather than introducing each with an extensive discourse on the rationale.\nIt is not clear exactly what was expressed in E. coli, i.e. which portion of the gene was tagged and expressed. Therefore it is not convincing evidence that the observed larger-than-expected U2OS-expressed protein is aggregated, rather than a longer form with additional amino acid sequences.\nIt should be clarified why U2OS cells were used as the heterologous expression system, rather than a more standard eukaryotic expression system. Assuming the authors are correct that different cell types express different cofactors, a less committed/differentiated cell type may have been more appropriate.\nValidity of the findings\nThe reader deduces that the straightforward experiment of taking the U2OS conditioned medium, either straight or partially purified, to PAT cells failed to affect their growth. These experiments and results should be more completely described.\nAdditional comments\nThis manuscript describes a complex system which is involved in the cellular sensing of cell density. The factor involves a relatively small protein associated with a lipid-based cofactor, a unique situation which has greatly complicated the characterization of this activity.\nCite this review as\nAnonymous Reviewer (2013) Peer Review #2 of \"Cells determine cell density using a small protein bound to a unique tissue-specific phospholipid (v0.1)\". PeerJ https://doi.org/10.7287/peerj.192v0.1/reviews/2\n",
+    "LLM_extracted_review": "['Review 1: Some of the last remaining points have been addressed satisfactorily, and the paper overall now makes a nice contribution to the literature in this area.\\n\\nReview 2: The authors are to be commended on a much improved manuscript, with the majority of reviewers\u2019 comments addressed. However, there appears to be no formal proof that either the protein/cofactor or cofactor alone control cell density. It would be prudent to alter the statement \u201cfurther proof of its role\u201d in the Abstract to something like \u201cin support of this proposed role\u2026\u201d. The writing describing the biological role of the protein and cofactors needs to be consistent with the proposed model of controlling cell density. Greater prudence is required when ascribing alkaline phosphatase expression to bone production per se, rather than as a phenotypic marker of bone cells. A clear reference from the literature regarding the origin, characterization, and use of the U2OS cells is needed, as the current supporting statement is weak.\\n\\nReview 3: The paper has been improved, but the writing is still vague in places. Many points previously raised have not been answered, such as providing a reference for U2OS cells. The response to Reviewer 2 contains odd phrasing, such as referring to cells as \"polite.\"\\n\\nReview 4: The majority of this reviewer\u2019s minor comments have been addressed, but there are still general considerations. There is no proof that either protein/cofactor or cofactor alone control cell density. Statements like \u201cfurther proof of its role\u201d should be modified to \u201cin support of this proposed role\u2026\u201d. The manuscript needs to be rewritten to indicate that the results are consistent with the proposed model of controlling cell density. Alkaline phosphatase is a marker of bone cells, but does not mean that bone is being produced.\\n\\nReview 5: Please pay particular attention to addressing the very significant concerns of Reviewer 1, especially those pertaining to appropriate controls, activity assays, and how you determined the significance of your results.\\n\\nReview 6: Key controls are missing, and there is no description of repeating experiments. The methods lack sufficient detail, and there are no statistics provided. The manuscript appears to have many logic errors, and the writing is very opaque. Key experiments have not been done, and there is no clear demonstration that you have purified or identified the factor affecting tendon cells. The introduction contains too much irrelevant detail, and the results section could be improved by stating what experiments were conducted and why. It is unclear what was expressed in E. coli, and the choice of U2OS cells as the expression system should be clarified. The straightforward experiment of taking the U2OS conditioned medium to PAT cells failed to affect their growth, and these results should be more completely described.']"
+}

peerj_json_files/PeerJ_Json_143.json ADDED Viewed

The diff for this file is too large to render. See raw diff